Dear reader,

Thanks for tuning in. This is part 2 of my deep dive on Yes Asia.

To get caught up to speed, please read part 1.

The past couple weeks have been extremely busy, so I apologize for the wait as some of you have been eager to read the full write-up. I’ve received many messages from you guys regarding write-ups, investments, and ideas — I absolutely love talking about stocks with you so always feel free to just message me!

Subscribe now

Above: Joshua Lau (CEO) and Priscilla Chu (Chairwoman and VP of Ops)

Why does the opportunity exists?

Yes Asia is an illiquid microcap listed in the SEHK. The company trades at <6x 2025 EV/EBITDA despite quadrupling top-line in just 3 years.

I think there are some things that raise eyebrows at first glance:

1. Due to Yes Asia’s significant expansion and warehouse investments, despite a major increase in sales, free cash flow was extremely volatile though is positive coming from a period of negative cash flow during the pandemic.

2. The greater mix shift to the B2B business also came at a lower gross margin.

3. I think there is a valid risk of inventory obsolescence, inventory makes up >40% of assets and has grown significantly in recent years. However, as I noted in my last post, I think this is necessary in order to be able to stock locally which improves the logistics process for both B2B and B2C clients. You generally need more inventory to prepare for future growth, this is a distribution business after all. Inventory write-downs have even in recent periods been <0.1%.

4. The Iran War has made freight costs wildly more expensive, or at the very least far less predictable. As such, I think investors are expecting lower profitability at least in the short-mid term.

Nevertheless, I think the price is too good to ignore for a strong business with a global growth runway that I believe will allow Yes Asia to grow at high rates for a long period of time.

To harken back to my use of benchmarks, I think Shein makes a great case study for the use of customer data to optimize demand and supply chain in a way that cannot be replicated by competitors.

The Shein Playbook

Being mobile native gave Shein exclusive ownership of their customer data with nearly 3/4 of traffic coming from the Shein app, today Shein has >250M monthly active users. This allowed them unique access to spot trends and consumer preferences better and earlier.

When Shein thought a design could go big, they would sell in small quantities, only buying more if there is demand. Backed by an enormous proprietary consumer dataset, this model allowed Shein to optimize inventory against demand better than less scaled peers.

6,000 small factories in China, struggling to find their place, now had a chance to sell to high-value markets in the West. Shien worked closely with its suppliers, becoming such a large part of their business many worked with Shein almost exclusively. This also allowed Shein to release new designs in days.

Then once a design was set to scale, Shein hammered down on social media marketing. As Professor Deighton of Harvard succinctly noted:

It [Shein] leverages deep skills in digital marketing to place stylish fast-fashion merchandise in front of target customers on social media sites such as Instagram, Facebook, and TikTok, as well as on its own website and mobile apps. Additionally, it has used email marketing, influencer marketing, and paid advertising to reach its target demographics, first in the US and later throughout much of the world.

In spite of generally lengthy shipping times and poorer quality, the sheer value in their product driven by Shein’s SKU variety and cost savings allowed them to circumvent incumbent fast fashion players like Zara or H&M.

The result is a behemoth fast fashion business that in less than 20 years was generating >$40 billion in annual sales and >250M in monthly active users. Shein now holds close to 20% of the global fast fashion market.

We see parallels of this model in Yes Style.

About 50% of revenue was generated through the Yes Style mobile app. Since 2022, Yes Style was downloaded >4M times annually. In March alone, Yes Style had nearly 12 million visitors, about 74% higher than closest B2C competitor Stylevana.

This data allows Yes Style to forecast trends. Yes Style’s team will identify a trend, for example the “bloom skin” trend which favors brands that contain salmon sperm. If they find new brands that fit these trends, they will first buy small amounts. If there is a lot of demand, then they will structure a larger partnership with the brand. In total, this allows them to stock roughly ~10% of their total SKU availability without much write-downs or customer experience issues.

Yes Asia works directly with their brands, who usually find them through word of mouth or Yes Asia’s open houses + conferences. Yes Asia may also reach out to brands they think have immense potential. While Yes Asia tries to have the best brand selection, they also make it a point to cut out underperforming brands. Nevertheless, the total brand count continues to grow, now at 475+ partner brands from 430 in 2022 for Yes Style. For reference, this is still over 3x Style Korean’s brand count. Asian Beauty on the other hand has 800+, compared to Silicon2’s 600.

Yes Style then curates these brands to a largeplatform of K-beauty influencers around the world. Over 500,000 partner influencers work with Yes Style either via direct partnerships or through the self-serve promotion platform. Currently about 30% of Yes Style sales are generated through the influencer platform.

But Yes Style also has its own large following: 1.9M followers on Instagram and 1.2M on TikTok, a bit over 2x the follower count for Stylevana and StyleKorean on both platforms.

Yes Style is essentially a bridge between the East and the West. Where Shein connects Chinese clothing manufacturers to Western consumers, Yes Style connects Korean (and Japanese, Taiwanese, Chinese) beauty brands to Western consumers.

Nonetheless, the market for Asian beauty products is growing so fast it naturally attracts many strong competitors.

The two primary competitors by far I’ve gather are in the B2C business, Stylevana, and in the B2B business, Silicon2. Let’s go over each.

Stylevana

Stylevana was launched in 2014 in Hong Kong, growing rapidly during the pandemic through massive adspend campaigns and influencer partnerships.

Above: Google search traffic, Yes Style (red), Stylevana (blue)

Search up any big K-beauty brand and the first thing you will see is a ton of Stylevana ads.

They were quite aggressive with pricing, driving prices to similar levels of Yes Style. They’ve also been expanding geographic footprint very fast, launching in North America, Europe, Latin America in just a couple years. By 2025 they reached >300M in revenue, getting close to Yes Style.

But when I was speaking with management, I did not get the sense they regarded Stylevana as a serious threat. And truly there are reasons to think the same:

• Sources suggest Stylevana does not work directly with many of their brands. One brand owner for example even said they never spoke with Stylevana but Stylevana would put their brand on their website. They would just buy their items in bulk and then resell on their own website.

• Stylevana has by far the most customer complaints. Customers frequently mention they’re not getting their products, delivery is taking forever, customer service is unresponsive. This is because as I mentioned Stylevana does not have authorized distributorship with many brands so they don’t stock any in-house, leading to huge logistics issues.

Here are some of the many customer reviews for reference:

You also have these reviews for all shops including Yes Style but they were far less frequent and were not as severe.

• Finally, Stylevana lacks the logistics network compared to Yes Asia and Style Korean. This means they are sensitive to trade disruptions and explains why their delivery times are so long. Most Yes Style delivery times are about a few days to a few weeks while Stylevana has seen anything as early as 1 week to over 3 months. Standard shipping fees for Stylevana are free at >$48 USD while Yes Style shipping is free at only >$35 USD

In summary, though Stylevana is growing very fast with pricing matching Yes Style, Yes Style undoubtedly has the better customer experience and actually works directly with their brands on top of a global logistics network. On the other hand, Stylevana is really a glorified dropshipper peddling online ads to pump up their sales.

Silicon2

I believe Silicon2 (owner of StyleKorean) a much larger competitive threat to Yes Asia, especially on the B2B size.

Silicon2’s B2B business is 4x larger than Asian Beauty. While StyleKorean has <200 brands available, Silicon2 has a total portfolio of over 600.

Silicon2 was founded by Sungwoon Kim in 2002 in South Korea as a semiconductor distributor, making a name for itself selling DRAM overseas. In 2012, seeing the potential of K-beauty distribution, the company pivoted to B2B K-beauty wholesale, primarily in China.

Silicon2 built out the logistics network early on, establishing 2 warehouses in the US by 2017 to support StyleKorean which was launched just a couple years earlier, then entering Indonesia via JV in 2018. Following the pandemic, Silicon2 also spent heavily in expanding into Europe, building a center in the Netherlands and Poland in 2023, as well as a center in Vietnam and Dubai in 2024-2025.

Silicon2 has about 85% of its business in B2B, with total gross margins of around 30%. On the other hand, while Yes Asia has similar gross margins, Yes Style has about 35% while Asian Beauty has only 17%, although note that Silicon2 uses FIFO method for inventory calculation whereas Yes Asia uses weighted average cost, which means becuase K-beauty product prices have generally increased overtime, Silicon2’s gross margins relative to Yes Asia’s are tad overstated.

Nevertheless, this implies Asian Beauty either has poorer route density leading to greater COGS-related transportation expense or Yes Asia is not getting the same bulk discounts overall as Silicon2 contrary to what I’ve heard from management or Asian Beauty is selling to its customers at a lower price than Silicon2.

My best guess right now is that the first 2 are most likely.

Yes Asia only recently started scaling this brand, and it is not yet at the level of route density required where one can claim they’ve reached steady state gross margins. But while H12025 Asian Beauty gross margins went up 1% point, full year remained the same due to aggressive growth into the offline B2B space which had lower markups vs online. (this is intuitive as these clients, i.e. Costco, Target, have greater negotiating leverage than fragmented online resellers).

But one thing Asian beauty has the clear upper hand in by a wide margin, is the data advantage. A proprietary consumer dataset requires a lot of…well, consumers!

End consumers, people who are actually buying and using the product.

Yes Asia started with a strong B2C business whereas Silicon2 started with a strong B2B business. Yes Style is about 15x larger than StyleKorean in revenue! To that end Yes Style has the best dataset in terms of mobile, web, and social media.

This is data Yes Asia has provided in combination to their wholesale offering which has allowed them to win over new customers, even those that were already working with Silicon2. Costco, Sephora, Ulta Beauty all were customers of Silicon2 first and now are customers of both Silicon2 and Asian Beauty.

Suffice it to say I do not look at the B2B business as excitedly as I do the B2C business. There is a long way to go in terms of scale and in there will be lower blended gross margins for Yes Asia while Asian Beauty achieves greater route density.

Nevertheless, without doubt the B2B business unlocks another layer of runway and was the catalyst in developing Yes Asia’s warehouse logistics and purchasing power. For now, the market is large enough, and growing so much so such that both companies will be able to continue expanding relentlessly for a generous period of time.

Management

Employees describe Joshua and Priscilla as highly open and approachable. Effort was made to setup offices around the world to attract global talent. Senior level turnover has been very low, many stay because of the growth opportunities and more flexible hours + WFH options relative to the typical Hong Kong company.

They like to hire a lot of young people, especially those with a passion for K-beauty. Indeed, a lot of people who work for Yes Asia had experience running their own brand or were simply enthusiasts themselves.

Despite Joshua and Priscilla being a couple, they focus on their own expertise: Joshua on product and brand, Priscilla on operations, logistics, and customer service. Thus I believe Priscilla should get a TON of credit for the growth of Yes Asia as a lot of it came form deliberate decisions to optimize and expand the warehouse network. They together own 37% of the company.

For the B2B business, Song Howon was brought in as CEO in January 2025. He had previously scaled Unleashia in Korea and to over 150+ countries. After selling the company, he then decided he wanted to help scale Asian Beauty Wholesale. From his wide network in Korea, he has, according to management, singlehandedly connected Yes Asia with 100+ brands! He’s only 39 by the way.

The key management team has proven they have the ability to launch and scale successful e-commerce businesses: Lau and Chu being owner-operators who founded Yes Asia of course and Song who had founded scaled Unleashia.

One concern though I have with management is their option program. Their program currently has an issuable limit that could lead to a maximum of 9% dilution. Nevertheless, the strike prices for all grantees is at least $5.8 HKD/share and many were above $40 HKD/share. Though Song Howon, who received over half the vested options in 2025, is at $49.6 HKD/share.

Valuation

Yes Asia trades at a 162M USD market cap, 5.6x 2025 operating profits.

Let’s talk about Yes Style alone: Projected sales growth of 24% over 4 years and operating profit margins of 12%, which by the way has been going up steadily over the past couple years, you end up with close to $100M USD by 2029, which at 5x EV/EBIT gets you to almost $500M USD in enterprise value, about 3x today’s valuation for the entire company.

Then you have Asian Beauty. Though currently at a 6% operating loss, I assume exit operating margin of 5%, far lower than Silicon2’s 18% which I believe is at steady state profitability now. At 5x EV/EBIT, Asian Beauty could be worth $165M USD.

In total without even counting the Yes Asia legacy CD business, the entire company, at very conservative exit multiples, could be worth $665M USD, and that would still be about half of my 2029E annual sales.

This implies an IRR of >40%.

Not confident about Asian Beauty? Maybe Asian Beauty won’t reach our expected profitability?

Fine, let’s remove it all together.

If we were talking about Yes Style alone we would still cap >30% IRR.

Again, I’m using a conservative 5-6x multiple. Personally I think this should be at least 10x steady state, after all, Silicon2 trades at 14x 2025 EV/EBIT.

Overall, there is a lot of uncertainty, especially with global macro tensions. However, I think the sheer runway and competitive advantages make this an absolute steal with a margin of safety even at conservative assumptions!

Note: Won’t include the dividend yield as there as there is some dilution for the option program (3% div yield on 2% est. share based dilution).

Risks

• Yes Asia fails to forecast demand or future growth properly which leads to the overpuchase of inventory which could lead to write-downs and negatively impact cash flow conversion from the sales growth.

• Decline in K-beauty demand worldwide will damage future growth prospects. We’re seeing the K-beauty trend mature in the US and sales in the US has decelerating with declines in US Yes Style sales, we should expect to see a similar trend in Europe and other areas overtime as well.

  

  Source: Yes Asia 2025 Annual Presentation

• Competition: though I firmly believe Yes Style has the best product, the competition is no doubt intense and there isn’t any barriers to someone opening up an online shop and reselling their own K-beauty brands. We saw how quickly Stylevana could grow even without nearly the same level of logistics CapEx. On the B2B side, it is worth closely monitoring Silicon2 customer acquisition vs Asian Beauty. While both are growing fast, Asian Beauty clearly has not yet reached inflection point in terms of route density and scale.

• Potential dilution from the option program

In conclusion, Yes Asia is a great business with pretty great competition. But it’s cheap, so cheap for the level of growth it’s achieving that even should the business not reach our expectations, should still make an excellent investment.

Thank you for reading!

Would appreciate any thoughts, corrections, suggestions, either in the comment section down below or just feel free to DM me.

I’m proud to share we have surpassed 300 subscribers. I still sit in awe that there is this many people willing to listen to a 22 year old talking about random unknown businesses in the far-flung corners of the Earth.

I started investing in my first year of university, 2021. Since then I have made countless mistakes. These were mistakes that cost me hundreds to a couple thousand dollars, during that time that was all the money I had!

There were times when I thought about giving up. When I thought that maybe this isn’t what I should be doing. I tried starting my own company. Failed not only once but twice! But in those moments I never stopped investing. Late at night whether I was in a good place or not there was always something so fascinating about stocks that kept me coming back. I liked it so much I wanted to share this joy with others.

I’m a beginner investor who just happened to strike an audience with a few hundred awesome kickass people. I have so much to learn and am blessed to have had the chance to work with great investors who love their craft.

So thanks for sticking along, and I look forward to keep giving you awesome investment content at Dunamis Investing.

Yesful Land is Yes Asia’s first showroom opened in Seoul on Oct 2025

Thesis:

Yes Asia (SEHK: 2209) is one of the largest b2c and b2b online distributor of Asian beauty products with a long growth runway brought about by the global proliferation of Korean culture.

Yes Asia surpasses the competition primarily through the following:

1. A superior vertically integrated logistics network that drives up route density, efficient stocking, and faster delivery times.

2. A wide availability of brands which I estimate is at least 3x their largest competitor.

3. Scale advantages and cultivating long-term brand relationships give Yes Asia negotiating power over suppliers and better cost savings

4. The largest set of proprietary consumer data on Asian beauty that allows Yes Asia to forecast demand the best and provide the most value to B2B customers.

This microcap is growing earnings at a whopping 25% CAGR yet trades at just 6x 2026E after-tax earnings. The company also has little debt.

After studying customers of Yes Asia and its competitors, it’s clear to me that Yes Asia has the best offering by far even against traditional beauty incumbents including Sephora or e-commerce giants like Amazon.

In this post we will talk about the overview of the business, logistics, B2B business, and international growth runway. Part 2 will talk about competitive landscape, marketing strategy, management, risks, and valuation.

YesAsia.com was founded in 1998 by Joshua Lau and Priscilla Chu in a San Fran basement as a website to sell Asian CDs and albums in North America.

Lau was in his early 20s, a former Goldman Sachs analyst in California while Chu had recently finished her MBA at Fresno State.

Together they scaled YesAsia to $13M in annual sales within the US by 2004.

But just a couple years later, Yes Asia would make a major pivot.

YesStyle was launched in 2006 as an online shop for niche Korean and Asian fashion and cosmetic brands which were slowly gaining popularity in the West during this time.

Within the first 6 months of launch, YesStyle surpassed $500k in total sales, and by 2011, it had annual sales surpassing $20 million, rapidly surpassing the original entertainment business.

By championing a commitment to offer the widest selection of Asian beauty products, YesStyle gained a cult following among mostly K-beauty enthusiasts who were able to find niche brands that would have otherwise not been available in the West via e-commerce like Amazon or retail giants like Sephora.

Moreover, their focus on the Asian beauty niche allows them to buy in larger bulks vs traditional incumbent retailers who are spread thin or resellers who simply lack the scale, giving them superior per unit cost savings which Yes Style then passes onto the customer.

Pricing comps suggests the average Yes Style basket is roughly 20%+ cheaper than Amazon or Sephora on the same items.

But wheat started out as “niche” has now become mainstream.

The mid-2010s saw an explosion of Korean culture abroad through the mass export of K-pop and K-drama.

All of a sudden, K-beauty was no longer an “exotic” category, it was the MOST popular product on the shelves. During that period, Yes Style saw annual sales nearly quadruple to $80M.

Yes Asia Holdings officially went public in 2021 in spite of declining sales. As they only had one warehouse in Hong Kong, they could not locally stock inventory in Western countries. Pandemic-led supply chain disruptions and lower discretionary spending across the board saw an overall decline in Yes Style sales of 25% from 2020-2022.

This could not happen again as Yes Asia was then trying to scale its B2B business quickly which would require a LOT more volumes. So as a result, Lau and Chu used the financing to do two things:

1. Establish a vertically integrated logistics network internationally to reduce shipping time, increase availability of brands, and de-risk from supply chain disruptions.

2. Reduce reliance on the United States by expanding aggressively into other markets, most notably Europe.

Logistics Upgrades

To reduce shipment problems, Yes Asia partnered with CN Logistics, long-time investor in the company. CN Logistics helped Yes Asia with ocean and air freight forwarding as well as regional distribution.

Then after the pandemic, Yes Asia invested heavily into warehouses in the United States, Europe, Korea, and Hong Kong, partnering with Geek Plus to integrate robotic automation.

The Goodman Warehouse in Tsing Yi, HK for example, was built in 2022 at a cost of $6M.

In 2025, Yes Asia would also upgrade their legacy HK Mapletree warehouse.

They also established warehouses in key customer markets. In 2023, a warehouse was setup in Los Angeles while in 2024, one warehouse was set up in Hahn, Germany and one in Sheffield, UK. In 2025, Yes Asia added another warehouse in South Korea.

By building out a vertically integrated warehouse network, Yes Asia was able to stock a far greater amount of inventory and far closer to the target customer.

The results of these investments were the following:

1. labor savings of $6.6M

2. 50% increase in shipments compared to non-automated warehouses.

3. Cut in the return rate by >50%

4. Decline in freight costs/order companywide by 35% since 2022 since they could now ship more products by sea vs by air alongside route density

Current warehouse utilization is <70% so there is still plenty of room to scale in their current set-up.

In my opinion, the most important effect of these improvements was the inventory management and delivery process!

After speaking with several K-beauty enthusiasts, the biggest pain point I’ve heard over and over again is when they order something that the seller fails or takes too long to deliver.

This was often the case with Yes Style’s biggest competitor Stylevana: they order something on Stylevana, then turns out Stylevana didn’t even have it in stock, and so the customer now has to talk to customer service which is often non-respondent, coordinate a re-order, and then haggle for a refund. These stories came up very infrequently when discussing Yes Style.

For these reasons, I’ve found that the biggest selling point for stores like Yes Style is not only the price but the smoothness of the delivery process!

Drop-shippers, resellers, and even scaled e-commerce players who just lack the same network of international warehouses such as Stylevana are not able to stock the right amount of inventory in proportion to the demand. These dilute their value proposition and leads to lost customers, or more accurately, customers gained for Yes Style. In comparison, Yes Style not only has the logistics but also the proprietary data from their larger customer base to forecast demand more accurately.

The advantages Yes Style has over their competitors would be the same ones that would lay the foundation on which their B2B business could be built…

Enter Asian Beauty Wholesale

Started in 2018, Asian Beauty Wholesale was Yes Asia’s B2B arm where, taking full advantage of their extensive brand partnerships, would distribute Asian beauty products to smaller resellers and brick and mortar stores.

Below: revenue breakdown Yes Asia

Later on, they would start supplying major retailers like giants such as Ulta Beauty, Costco, Sephora, and Asian beauty specialist retailers such as the quickly growing Kiokii in Canada. We know they are moving to larger clients since average order size has more than doubled to $2.8k/order from $1.1k at the end of the pandemic.

Many of these had problems with their local distributors.

For example, Kiokii’s previous Canadian distributor had many problems. Selection availability was not strong, logistics were inefficient and items were frequently out of stock. Yes Asia filled the gap by providing world-class logistics, a full SKU and brand selection, as well as proprietary customer data to help clients track trends and brand closely.

This was supported by their new and improved warehouse network which allowed them to stock 8x more inventory compared to pre-pandemic. B2B clients also wanted Yes Asia to stock items in their country, so without the warehouse investments they likely would not have been able to secure the larges clients in the first place.

Asian Beauty now contributes 30% of overall sales, but while Asian Beauty Wholesale unlocked massive growth opportunity, wholesale naturally has a lower markup — much larger earnings but at the expense of a lower blended gross margin.

Below: gross profit breakdown Yes Asia

Online reseller gross margins are about 6% points higher than offline/physical store clients. This makes sense as the average account for the offline channel orders a LOT more volumes.

Now in 2024, Asian Beauty Wholesale gross margins went from 23% to 17%. But what’s interesting is in 2025, Yes Asia booked many major clients in the offline channel, yet gross margins increased to 19%. In spite of the offline segment growing to 28% of B2B sales, we saw a 3% point increase in online gross margins. This was primarily due to scale.

AsianBeautyWholesale’s key competitor is Silicon2 owned by StyleKorean. Silicon2 is currently the largest B2B global distributor of K-beauty though Yes Asia is catching up fast according to management.

The main differentiator between AsianBeauty and Silicon2 is the data. StyleKorean started with a strong B2B base and then entered B2C. Yes Asia did the opposite, staritng with a strong B2C base and then entering B2B. Unfortunately for Silicon2, Yes Asia has a far bigger end-consumer data set. This gives them a dangerously powerful value proposition for B2B clients who need strong proprietary data to optimize stocking. I will talk more about Silicon2 in the next post.

So, to summarize, since the pandemic, management executed these 2 changes:

1. Yes Asia is no longer just a reseller but a global logistics powerhouse in Asian beauty distribution

2. Yes Asia now serves both B2C and B2B both to online resellers and brick and mortar retail giants

The result is far better distribution and scale advantages which Yes Asia would need to fuel the next phase of the masterplan: global expansion.

An International Growth Runway

2022 was a transition from a focus on North American to a more global customer base.

As the US market matured, Yes Asia focused more on markets with a faster growing K-beauty cult such as Latin America and Europe.

Source: Company filings

Asian Beauty Wholesale would end up following a similar path. Canada would become a major part of the business via the partnership with Kiokii. Asian Beauty Wholesale will be critical in stocking Kiokii’s new flagship store in American Dream, New Jersey which just opened.

Above: Kiokii, American Dream opened in February 2026

Online sales would remain more US centric as there are a large online re-sale base over there.

Source: Company filings

The demand growth of Asian culture exports in Europe mirrors a younger version of the 2010s era in the United States.

Currently a bit over half of global K-beauty sales are in the US. Europe grew from 3% of the pie to 11% in less than 3 years. China, which had historically been the largest destination of cosmetics exports, has far been surpassed by the West as rapid growth in domestic brands are competing with Korean brands.

Countries like Poland is literally seeing triple digital YoY growth in K-beauty consumption which has in recent times seen gigantic consumption of Asian products and media.

You also have markets where Yes Asia is much less penetrated such as the UAE and India whose consumption are also growing >60% YoY.

This is a global growth runway, and one that’s we can model as we’ve seen the wave take place in America from early adoption to growth to maturity.

Though different countries will require strategic nuance, I don’t see why Yes Asia can’t continue using the North America playbook again in places like Latin America, Eastern Europe, India, the Middle East, which are still in the early adoption + growth phases and have less competition.

Moreover, you also diversify yourself from FX and tariff risk.

Therefore, I believe Yes Asia should be able to grow for a long period of time. Without reliance on American sales, you will still be able to expand in new pockets with faster growing K-beauty base even though mature segments like the US are not growing anymore.

Is the growth sustainable?

There are 2 things I believe should be flagged when going over the books.

1. Low cash conversion

In spite of the high revenue growth, you will notice a far lower growth in operating cash flow. In fact, CFO as a % of net income is only 56%. This is primarily due to rapidly increasing inventory and some increase in accounts receivable.

Firstly, the warehouse buildout means a far greater capacity to stock inventory. Stocking more inventory at warehouses allows them to deliver product more efficiently and reduces risk of trans-national shipping issues.

Secondly, this is simply the kind of business that requires more inventory to grow. A wide SKU selection is important, so to continue sales growth you will see a rise in inventory stocking and then topline will follow. With the B2B business opening the floodgates for volumes, inventory would naturally expand rapidly.

Nevertheless logistics improvements led to a reduction in G&A as a % of sales by a whopping 10% points! Delivery cost per order have also come down from nearly $20/order to $12/order in just a few years. You also have ROE surging past 30% in spite of the increase in working capital.

I don’t mind increases in working capital if it is needed to continue growing at a strong pace over the next several years. (Note that Yes Asia uses weighted average cost method of calculating inventory.)

2. Rising marketing spend per acquired customer

Yes Asia has adopted quite an aggressive marketing strategy by utilizing an army of influencers to promote brands in Instagram, TikTok, etc. Influencers are typically paid an affiliate fee whenever a sale is made through their posts.

We saw an increase in CAC from $10-$12/customer in 2023-2024 to $14 in 2025. This is due to an increase in rising competitiveness in traditional digital adspend such as on Meta platforms or Google as well as an expansion into new markets which at the beginning has much higher CAC. Another reason for this could be a deceleration in US sales growth as the market matures, making it generally more competitive and thus more expensive to acquire new customers.

Management notes they are focusing more on their influencer base now. Nevertheless, I will be monitoring the CAC closely. More on this in the next post.

There’s a lot of things I would like to talk more about but this post is getting too long so figured I will save the other key research points for part 2.

I encourage you to subscribe to stay tuned to the continuation of the research.

Thanks for reading!

Above: Long Thanh Airport rendering

Coming off from my write-up on this company, I received very interesting questions that I think deserve time for discussion, thank you to those who asked.

So in this post, I will simply go over several key questions concerning the SCS thesis:

1. What is the customer concentration risk should SCS lose the Long Thanh bid?

2. How likely is SCS to win the Long Thanh bid?

3. What is the risk of heightened oil prices?

4. Thoughts on management?

If you haven’t yet, please consider subscribing, would highly appreciate your support.

Subscribe now

For those who are not caught up, I’ll give you the 60 second commercial.

As I’ve stated many times before, Vietnam is growing, fast, very fast, stemming from a boom in manufacturing and trade caused partly by offshoring from China onto Vietnam.

The Saigon Cargo Terminal (SCS) owns a piece of infrastructure in the Tan Son Nhat airport that stands as a prime beneficiary of this, as I expect both direct international longhaul and intra-Vietnam trade to increase significantly long-term.

The stock has come down A LOT since the up and coming international airport, Long Thanh, will likely lead to most international long-haul flights moving to the new airport. Investors are worried SCS won’t win a concession to operate a terminal in the new airport.

But as we discussed, this is an opportunity not a risk. Even IF SCS does not win a terminal bid in Long Thanh, you still have 1) regional and domestic flights that will remain in Tan Son Nhat, 2) a profitable office building with another one one the way, and 3) a gigantic portfolio of deposits. And you are rewarded a 10% div yield while you wait.

For the full detailed scenario analysis, please check out my initial write-up.

1. How likely is SCS to win the Long Thanh bid?

There are really only 3 major entities that have expressed interest in operating a terminal: Vietnam Airlines, Viettel Post, and Saigon Cargo Services.

After talking to some ex-TCS employees, seems that the Terminal 2 concession will likely be given to either Vietnam Airlines or Saigon Cargo.

If Vietnam Airlines wins, it will basically be the same gig as the Tan Son Nhat Terminal with SATS from Singapore as a co-investor. Given their track record with TCS, and the possibility of preventing other airlines like Vietjet from working with them, I think ACV would not want Vietnam Airlines to operate one. That said, they also have a cargo terminal at Noi Bai Airport which is the major hub for Hanoi, and as they have strong political power, I would not rule them out.

Viettel is an interesting case as they don’t have any track record operating a full airport cargo terminal. But they are a logistics giant, and have proven they can build and operate large logistics centers like a 300 sqm warehouse in Noi Bai Airport and a giant logistics hub in Da Nang.

Like Tan Son Nhat, the Noi Bai Airport has a terminal invested in by Vietnam Airlines (NCT) and another terminal invested in by ACV (ACSV). There is a third, ALS, but that’s also invested in by Vietnam Airlines and NCT. The point is if Long Thanh follows the ACV playbook, there will likely be at least one terminal operated by an ACV-backed entity.

With this in mind plus the two key competitors focusing on Terminal 2, I think there is a high chance an ACV-backed entity could be selected for Terminal 1.

The question is will it be SCS or will ACV create a new company.

SCS and ACSV are separate entities and both were established roughly at the same period, 2009-2010. But back then, ACV was divided into the Northern Airports Group and the Southern Airports Group, so it’s not a reliable case study.

As ACV owns 13% of SCS, there is an argument that ACV has an incentive to give SCS the terminal, as it would immediately increase the value of their stake, making them tens of millions of dollars in capital gains. However, ACV may also desire to crate a new entity that would allow them to have a greater stake, whether or not they continue to work with Gemadept.

But here’s the kicker.

SCS is already working with ACV and international consultants like Incheon in planning said terminal.

They are already discussing what the design will be, what technology will be used, what day-to-day will look like.

Therefore, I think the chances of SCS getting it is very high.

2. What is the customer concentration risk should SCS lose the Long Thanh bid?

As I mentioned before, SCS’s domestic airlines include pretty much everyone except Vietnam Airlines, the largest state-owned airline.

Vietjet is by far the biggest domestic customer. And based on current airline market share numbers, they would comprise 75% of the market ex-Vietnam Airlines, likely closer to 80-90% if we’re talking just HCMC. When Vietjet has issues, SCS inevitably gets affected. For example, when Vietjet created their own ground handling unit, their efficiency was temporarily reduced, contributing to a decline in throughput. However, so far this is all short-term volatility and does not meaningfully hurt the overall business.

Now if SCS loses the Long Thanh bid, would that give Vietjet too much negotiating power over SCS? I.e. if they want much lower prices and SCS says no, they move to TCS.

With 20% of international and 80% of domestic cargo volume expected to remain in Tan Son Nhat*, I expect about 65% of volume to come from Vietjet in the scenario SCS loses Long Thanh.

However. I believe the chances of them ever switching to their only competitor, TCS, is extremely low.

Why? Because TCS is owned by Vietnam Airlines, Vietjet’s main competitor. Putting a gigantic chunk of your domestic commercial cargo flows in the hands of your enemy is not a good idea.

This would mean:

1. Vietnam Airline’s cargo and schedules get priority over Vietjet’s, which would cause major conflicts of interest if Vietjet wants to expand and TCS is constrained.

2. Vietnam Airlines will be able to track Vietjet’s shipment movement through TCS data whereas Vietjet can’t track Vietnam Airline’s.

3. Vietnam Airlines could potentially deliberately sabotage Vietjet’s operating efficiency, like telling their workers to secretly work slower or something sneaky like that lol.

In other words, switching to TCS would be shooting yourself in the foot, the same goes of other domestic airline competitors. This is the whole reason no domestic airlines owned by Vietnam Airlines wanted to work with TCS.

TCS was really built to vertically integrate with Vietnam Airlines. Now as Tan Son Nhat’s land bank is pretty much used up (which is the whole reason Long Thanh was built in the first place), there is little risk of Vietjet trying to co-invest in their own cargo terminal in that specific airport like what Vietnam Airlines did at the beginning.

In my opinion the bigger risk is Vietjet going bankrupt.

3. What is the risk of heightened fuel prices?

We could have a whole separate discussion of how the Iran War will end. However, I cannot predict where fuel prices or shipment volumes will be tomorrow, next month, next year, or next decade.

The war has forced many planes to take alternative routes which has affected flight schedules and capacity of airlines like Qatar Airways and Turkish Airlines, which are key clients of SCS. As such, I expect to see lower SCS international throughput this year.

On the flip side, you’re seeing immense demand for air cargo as shipping by air acts as a relief valve due to constraints in ocean freight. As such, air cargo prices are up by 70% since the start of the war. We’re seeing goods previously shipped by sea such as generic medicines switching to air.

Chief Supply Chain Strategist Steve Blough said on this:

"Customers are shifting freight from ocean to air, however it is extremely expensive - typically 5x to 10x higher - and those costs are climbing as capacity tightens"

So there is less flights moving through Tan Son Nhat but a greater share of goods being shipped by air.

This is a net negative for SCS overall as the amount of goods shipped are capped by the airline capacity, which has gone down. Moreover, though there is more demand now, if rates and fuel prices stay too high, it may be more economical for certain goods to switch to sea freight. This goes for lower value goods like some electronics and fast fashion, whereas goods that are likely to remain on air are delicate high value goods where fast shipment is needed, such as foods, machine parts, fragile equipment, smartphones, semiconductors, vaccines, etc.

Since SCS charges more for perishable and high value goods that should offset some of the decline from low value goods shift.

This is an uncertainty we simply have to accept and is a risk IF the key air routes remain downed and fuel prices remain high in the long-term.

If they win the Long Thanh, I think the market discount will still disappear at the current state we’re in.

If they do not win the Long Thanh bid, domestic routes will still face heightened fuel prices if the Straight remains closed although there is obviously far less airline capacity constraint for domestic vs international routes, so the damage at least in the short-term as far as I know, is not as bad.

I could model out a scenario where this happens but that would require making a long-term prediction on throughput, airline capacity, and fuel prices, which to me would be pure speculation. We will likely see double digit declines in cargo throughput in the short-term though. If the conflict remains long-term, i.e. 5-6 years or more, then intrinsic value would be meaningfully lower, but I also think should this scenario happen, many many businesses and many many industries will also be meaningfully hurt.

4. Thoughts on management

As I went over in the last post, CEO Nguyen Khanh Quoc has extensive experience in logistics, serving with Gemadept for nearly 20 years before being chosen to found SCS. At Gemadept he became head of container agency which would provide the foundation for successfully running a cargo terminal later on. Khanh Quoc owns about 1.7% of the company, worth nearly $2 million USD.

Total insider ownership is nearly 6%. Not that large, though most insiders and board members, including the chairwoman, are from Gemadept. This means Gemadept, who owns a third of the company, is effectively running it. Their stake should be roughly 7% of their entire market cap. Thus Gemadept has an incentive to maximize returns.

As far as I know, ACV has one representative board member, Truong Minh Sang, though Minh Sang is part of the supervisory board, so no decision-making capabilities but they ensure SCS is acting in the interest of ACV.

To add on to what I talked about in my initial write-up on capital allocation, there has been minimal dilution of 2-3% since IPO primarily related to their employee stock ownership program. Management in recent years has been focusing on maintaining a strong net cash position to be able to commit to the Long Thanh terminal and invest in the SCSC-2 office building + other investments.

Should the Long Thanh bid fail, I wouldn’t be surprised if they announced a special dividend or increase their dividend payout as management has historically offered very high dividends, on average above 70%, which has translated to a div yield on a market cap basis of 5-10%.

Whether or not the Long Thanh bid goes through, they will still have international throughput for a while so div yield will likely be 11-12% for the year or two and then should descend into the 5-8% range later on.

Source: 2025 AGM minutes

Overall, I think the risk to reward profile is highly attractive even with these uncertainties regarding the Iran War and customer concentration with Vietjet due to the 1) duopolistic nature of the cargo terminals at Tan Son Nhat, 2) the conflicts of interest TCS has since they’re owned by the biggest competitor Vietnam Airlines, 3) SCS’s strong net cash position + office portfolio opportunity, and 4) the high chances SCS will win the Long Thanh bid.

The stock now trades at 4x 2025 EV/EBIT** with a current div yield north of 10%. Please see my initial writeup for a valuation rundown.

Thanks for reading!

*Note: Though Tan Son Nhat is expected to retain 20% of international cargo and 80% of domestic cargo, I decided to use a scenario where Tan Son Nhat only retains all domestic cargo in my original valuation (this is because I don’t know which international and domestic airlines will remain thus the rates/kg changes are hard to model out). Running the numbers, you end up with 13% higher throughput volumes with 80% domestic 20% international vs my original no Long Thanh scenario. Thus my original valuation is generally more conservative. I regard the “no Long Thanh and you continue growing whatever’s left + SCSC-2 office”, as the base case scenario.

**For share count I am removing preferred shares which is why the market cap and thus the EV looks smaller on CapIQ than Vietstock or other sources.

Disclaimer: the following is not investment advice, please do your own due diligence before making any financial decisions.

I looked into Airports Corporation of Vietnam. I did not like what I saw.

The company is 95% owned by the Ministry of Finance. There were tons of issues. Entire airports at risk of being taken away. Major dilution if the state forces ACV is issue them more shares.

The business was excellent no doubt but came with big risks I simply was not willing to take.

So after discarding ACV, I stared at my screen and noodled in my own thoughts for a good 20 minutes.

“How can I own a piece of infrastructure with my downside protected?”

Then it hit me.

Though in my last post I talked about my distaste for airport-adjacent plays, i.e. ground services, cargo terminals, etc, I decided to take another look.

…and I found what I was looking for.

The company I will talk about today is one where your downside is protected and your upside is enormous should things turn out well.

This stock has been profiled before on Substack. Cao Nguyen, a prolific Vietnamese value investor I’ve been reading, has a post on this stock, and I think his breakdown covers the main points.

So today, I will try to dive deeper on the business and give my own view on this idea.

Saigon Cargo Services - market cap: $210M USD

SCS is one of the 2 cargo terminal operators that manage a cargo terminal at the Tan Son Nhat Airport, Saigon’s primary airport which takes up most air traffic in Southern Vietnam.

This is a duopoly business with very sticky customer relationships. Due to the inherent risks involving operating efficiency and cargo security that could cost carriers and logistics providers money and customers, SCS’s services are extremely mission critical and typically involve contracts ranging from 1-10+ years.

The result is inflation resistance: SCS’s revenue per ton has grown at a cagr of 3.5% on inflation averaging around 3%.

How does SCS make money?

Cargo terminals makes most of its money from logistics agencies and OEMs who want to move goods to and from the airport. Handling, storage, deconsolidation, and documentation fees are the main source of revenue but money is also made from reweighting items, printing labels, regulation checks, customs declaration, and more.

However, the demand is dependent on the airlines who transport the customer’s goods. Airlines are charged separately for equipment rentals, item reinforcement, and other services, but the airlines are who the agencies go to transport goods. Because agencies follow the airlines, when an airline switches, you lose the customers.

No airlines, no customers

And this is where the switching costs come in. Talks with industry experts suggest low churn with about a 95% rate of airline contract renewal.

The terminal business is about fixing 2 constraints: service quality and capacity. After all, those are really what airlines are looking for. Thus airlines who are happy with their terminal usually stay, it’s not worth risking your fuel efficiency and delivery schedule on a terminal you’ve never worked with.

SCS serves both international and domestic airlines with domestic cargo making up about 25% of volumes. International cargo has historically grown at a 5% cagr whereas domestic cargo has grown at 4.5% cagr. This, plus pricing growth, has led to almost 10% cargo revenue growth for nearly a decade.

Above: SCS annual cargo throughput volume in tons

Why does the opportunity exist?

While SCS has historically traded at a >10x ltm EV/EBIT and >15x P/E, the stock has cratered over the past year.

Why?

Because due to the congestion of the Tan Son Nhat Airport, the state has been building a new airport, the Long Thanh Airport, which is planned to eventually take up most of the international traffic going to Tan Son Nhat today. Investors are worried SCS will fail to win the concession to operate a terminal at Long Thanh, thereby slowly losing international cargo volume as it shifts away from Tan Son Nhat. Moreover, SCS’s tax benefits have expired and will now be taxed at the full 20% corporate tax rate (from 10%).

But this is where the opportunity lies…

The company now trades at <5x ltm ev/operating profits, < 8x ltm after-tax earnings, and about 6x free cash flow, the lowest valuation the company has ever seen since getting listed on the HOSE!

If we assume the worst-case scenario, i.e. SCS loses ALL international cargo volume which means sale and margins decline, SCS still has

1. A growing domestic cargo business with now longer runway due to higher capacity.

2. A high-demand office tower with another one expected to start construction in the near future.

3. A net cash position nearly a quarter of market cap that is invested in short-term deposits.

I estimate the remaining non-international business is worth roughly 15% more than the company sells at today. And at the current payout ratio, you also get away with a 5-10% div yield while waiting.

On the other hand. If SCS wins the Long Thanh concession, this could be a multi-bagger in just a couple years. As the Long Thanh airport is expected to begin operations mid-2026, we will know the outcome by end of year.

In other words, this is a heads I win tails I don’t lose situation. You get a small return and, should things go well, which I am inclined to believe so, you win big.

So now that we’ve got the gist of the thesis. Here’s how this deep dive will be structured.

1. The history of terminals in the Tan Son Nhat Airport and SCS’s competitive landscape

2. The real estate portfolio

3. The cash balance and high-level capital allocation

Then we will put it all together to find a sum of the parts valuation of SCS given the different scenarios that might play out.

The Cargo Duopoly

The Tan Son Nhat Airport is one of Asia’s most congested airports, commanding the lion’s share of domestic and international travel and air cargo in Vietnam.

Now the Tan Son Nhat initially had 1 terminal, built in 1997 under the operation of Tan Son Nhat Cargo Services (TCS). TCS was formed with Vietnam Airlines, state-owned and the country’s largest airline, as the majority shareholder (55%). Other investors include Saigon Airport Services, 49% owned by Vietnam Airlines, and SATS, a Singaporean firm, owning 30%.

But in 2010, another terminal was constructed for $50 million USD backed by a consortium of private and public investors, the largest being logistics giant Gemadept, who were big in seaport logistics, owning about 23%, the operator of the Tan Son Nhat Airport itself the Southern Airport Corporation owning about 25%, who would later merge with other airports to form the ACV, and military-backed firms such as Aircraft Repairing 41, owning about 24%.

Together they formed Saigon Cargo Services, which contrary to TCS was

1) Not state controlled

2) Did not have one >50% shareholder

3) Was backed by the ACV itself who runs the very airport both SCS and TCS were operating on.

Above: Initial landbank provided to SCS in 2008. TCS’s terminal 1 in blue.

The Saigon Cargo Terminal was designed to cater to international logistics operators. It received its TAPA A, ISAGO, and RA3 certification in just 4 years. And at the get go, they already had perishable goods processing centers and X-ray scanners. The terminal prior to future phase improvements was designed to handle 200,000 tons of shipment volume annually, similar to TCS’s 200,000 tons.

Above: SCS terminal outside (left) and inside (right)

Moreover, compared to TCS at the time, it was better equipped with modern elevatoring systems, cold storage, IT, and better trained staff and security.

In 2014, the World Bank reported the following concerning TCS:

In general, LSPs (logistics service providers) view the skill level of air cargo terminal operator workers as low, which results in delays, cargo damage, and increased costs for LSPs and BCOs. Security is also lax.

LSPs consider TCS to be crowded and inefficient with cargo-handling equipment that is substandard relative to other Asian countries. All air carriers utilize this terminal.

TCS tried to catch up, building a perishable goods center and installing X-ray scanners in 2011-2012 as well as achieving similar certification to SCS by 2015 (SCS would later receive CEIV Pharma in 2019 with TCS following suit in 2022).

Above: TCS Terminal 1

However, due to the age and legacy structure, renovation and equipment upgrades on terminal 1 could only add so much capacity. As their terminal started to get full, they constructed a second terminal using their remaining landbank. Terminal 2 was completed in 2012, adding 150,000 annual tons of volume capacity featuring a 25m height (SCS was 15m and TSC 1 <10m) and modern elevatoring systems, ULD, and cold storage similar to SCS.

Above: (Left) TCS AFT2. (Right) Grey building is AFT1 and yellow building right above is AFT2.

However, this terminal would also get full in just a few years. Since TCS has run out of landbank, its only means of capacity expansion is by increasing efficiency on existing fixed assets.

Source: ABS

In 2018, since SCS’s terminal also was fully utilized, SCS commenced a Phase 2 which involved upgrading and increasing handling equipment, expanding domestic cargo warehouses, investing in IT, upgrading cold storage facilities, and adding storage rooms. They also added solar panels on the rooftop in 2021 which generates 40% of its electricity needs. These upgrades added 150,000 tons in annual capacity without using up a lot of additional landbank. Utilization today is around 75% with returns on capital north of 40%.

Among domestic airlines, TCS’s customers are Vietnam Airlines and VASCO, a subsidiary of Vietnam Airlines. On the other hand, SCS’s clients are everyone else, including the big budget airlines like VietJet and Pacific, as well as new players like Bamboo Airways and Viettravel.

Source: Vietnam Investment Review, Center for Aviation

Vietnam Airlines has slowly ceded market share to new players with Vietjet achieving immense growth in just a short period of time. The space remains a very competitive space and no doubt the bankruptcy of a major customer is bad news for a cargo terminal. Vietjet likely makes up the majority of SCS’s domestic throughput, so this is a risk to monitor.

Among international players, SCS seems to have grabbed significant market share. When SCS opened, 10 international airlines switched from TCS to SCS. More recently, one of TCS’s largest international airlines, Qatar Airways, also switched to SCS citing capacity constraints in the TCS terminal.

Market share has slowly gone more and more to SCS over the years. In 2016, SCS and TCS had a roughly 30-70 split on Tan Son Nhat’s throughput. Today, this split is 50-50 as the TCS is operating at and above designed capacity.

Real estate portfolio

Saigon Cargo owns an office building called SCSC-1 built in 2013. It is located right beside the SCS terminal with 5 floors consisting of 8,650 sqm in gross floor area and 7,500 sqft in net leasable area.

In 2025, the office building brought in nearly 63 billion VND or $2.5 million in revenue which translates to $28/sqm/month. The real rents however are closer to $14/sqm/month, which is on the mid-range of office space rents I could find in the Tan Son Nhat ward.

There is exceptional demand for this office space as many airlines and logistics providers benefit immensely from keeping their admin close to areas of operation. Even if international travel is redirected to Long Thanh, thehigh demand + scarcity of office space within airport grounds make SCSC-1 a prime location.

Given HCMC offices are typically valued at an 8-8.5% cap rate. If we value this property on a 9% cap rate on about 50 billion VND in NOI, the SCSC-1 building should be worth around $21 million, roughly 10% of the market cap.

That’s not all. SCS is currently planning SCSC-2, a second office building that is expected to have a net leasable area of over 14,000 sqm, almost double SCSC-1. Assuming same rents trended on 3% rent growth, I think the building could be worth north of $50 million at stabilization on an 8% cap rate.

However, the timeline is uncertain as SCS is still waiting for approval from the Ministry of Defense as the land SCS wants to build on, though leased by SCS, is currently still being used for other airport operations.

ACV’s influence will definitely help SCS and personally I think it will happen given historically blank land used for airport operations have been converted into more productive airport assets, such as cargo terminals, offices, warehouses, and customs sites.

Nevertheless, I decided that in a worst-case scenario, SCS will not get approved for SCSC-2.

Aside from the office buildings, SCS also owns airplane parking space, warehouses, and a small football field, though I ascribe little value to these.

Capital allocation

Finally, SCS has $6.6 million in hard cash and $45 million in short-term deposits, more than half of total assets. These deposits now generate almost $3 million a year in income less interest expense.

Management has also historically issued high dividends, paying out 70-80% of net income annually. In this particular case, I am for it: If management cannot reinvest excess capital at high rates, I’d rather have them return it to shareholders who can.

As SCS is under capacity, there will inherently be less places to invest capital while waiting for demand to fill up terminal space. CapEx right now is mostly maintenance and/or upgrades rather than for expansion.

Historically, they’ve raised dividend yield as they’ve gotten larger and increased idle cash. This was the case in 2020 and 2024 where dividends were increased in response to significant cash pile up. Shareholders like Gemadept have an incentive to maximize returns as the largest investor and Nguyen Quoc Khanh, whose been managing director since founding, was an executive at Gemadept and has been with Gemadept for 14 years before leading SCS. Bui Thi Thu Huong, Chairwoman of the Board of SCS, is also the CFO of Gemadept and a member of its board. Ergo I am inclined to believe management’s goals are in line with the shareholders’ goals.

Putting it all together…

We have 2 scenarios:

Scenario #1: SCS does not win the Long Thanh bid and completes SCSC-2 office building

In 6 years SCS loses all international cargo revenue, a 70% decline in cargo sales (here I’m including airplane parking and other small revenue sources which are gross unprofitable by the way) from 2025 from $45M to $14.7M assuming continued growth only on the domestic cargo business. Operating earnings from cargo services slump to $8.5 million assuming G&A* just growing by inflation as SCS still has to maintain its facilities even if its underutilized. This is $6.8 million after tax → $102M on 15x after-tax earnings.

Then you have your office properties:

SCSC-1 trended on 3% rent growth brings in $2.3 million in net operating income*. SCSC-2, now stabilized, brings in $4.4 million. This brings a total of $6.7 million → $80 million on an 8.3% blended cap rate (9% on SCSC-1 to account for aging and 8% on SCSC-2).

Finally, you have a net cash position comprised of deposits and held-to-maturity loans bringing roughly $2.7 million a year based on current portfolio size or a 5% yield. Compounded at this yield would lead to a total net cash position of $74 million by year 6. But take out the construction cost of $13 million based on today’s HCMC office build cost per sqm, and you’re left with $61M

So…you have the Domestic cargo business (and others) worth $102M, the office business worth $80M, and a net cash position of $61M. This brings the entire intrinsic value of SCS to $243M.

This implies today’s price is at an 14% discount.

Scenario #2: SCS does not win the Long Thanh bid and does not complete SCSC-2 office building

Office business NOI of $2.3 million on 9% cap rate is $26.6 million. Do the same math, add back SCSC construction cost, and you get $203 million in intrinsic value, almost equal to today’s market cap.

If you think of it in reverse, it means that at today’s valuation, the market is already expecting the worst-case scenario.

Scenario #3: SCS wins the Long Thanh bid and does not go through with the SCSC-2 building.

Growth of sales volumes for both domestic and international remains the same. As cargo G&A* as a % of cargo revenue is about 5.6%, I’m assuming 10% of revenue is G&A, basically almost doubling, as more costs will be needed to operate the new terminal which, if SCS gets one terminal like in Tan Son, would add 225,000 annual tons in capacity.

Revenues reach $62M with after-tax profits at $35.6M. At today’s multiple, this alone is already more than the intrinsic value of SCS in scenario #1.

But say they return to 15x earnings → this would value the cargo business at $535M. Add $26.6M for the office and $74M net cash, you come to $636M.

Over 3x today’s market cap.

In short, the market is assuming SCS 1) does not win the Long Thanh bid and loses all international cargo sales, and 2) SCS never starts building the SCSC-2 project.

But if we think in terms of expected value, the baby has been violently thrown out of the bath water.

And the cherry on top?

You even get a 5-10% div yield while waiting!

(Note: 5% is the floor I calculated given payout ratio stays the same on much lower profits if SCS does not win the Long Thanh bid. 10% is current yield)

In the next post on SCS I will dive deeper on the Long Thanh airport situation and competitive landscape of SCS.

Thanks for reading!

Subscribe now

*Note: NOI for office was Revenue - COGS for office - G&A (without office building D&A) as proportional to office revenue over total revenue. Operating cost for non-office business calculated as G&A (without office building D&A) as proportional to non-office revenue over total revenue.

Firstly, just want to say thank you to all who have subscribed, we are almost at 200 subscribers which is an audience larger than I would have ever expected to be writing to. I have a very specific style of investing, and I’m surprised it has resonated with this many people.

Subscribe now

Know that what I write about can be found almost nowhere on Substack and it is for that very reason I find joy in doing what I’m doing.

I remain focused on my mission to find small businesses that can follow in the footsteps of excellent companies around the world whose success stories are underpinned by the development of durable competitive advantages. These stories serve as models to help study and discover new opportunities; such is the crux of benchmark investing.

Coming off my last post, which was on Viettel Construction, I continued looking for small-cap companies in Vietnam and went through all 1600 of them across the HOSE, HNX, and UpCoM — let’s just say I quickly passed on 99% of them. I didn’t do deep dives on each one, but I did end up with a shortlist of small companies that I think look promising.

But before we go into that, I want to first spend some time talking about how I opened a trading account in Vietnam. Thanks for planting the idea along with everyone else who asked me how I did it.

The first thing to know is that right now you need to go to Vietnam in person to open an account (Though my broker told me they will be offering online registration in the near future).

You need to first open a bank account in Vietnam and then open a trading account with a broker.

I used SSI, Vietnam’s largest brokerage, and they guided me through the entire process.

Here are the steps I took.

1. Called SSI’s English department and got their number.

2. Provided them with my information online.

3. Arrived in Vietnam and headed to their office.

4. They took me to BIDV, their partner bank, to open a bank account. There was a lot of paperwork so read it all carefully. This part took me an hour.

5. We went back to SSI and there I registered my investment account and logged in to the SSI trading app.

It only took a morning to set everything up.

To clarify, you transfer cash from your local bank account to the BIDV bank account. If you make a trade, SSI will access your BIDV account, so SSI does not hold your cash, the bank does.

If you want, DM me and I can send the contact of my broker, and he can get you started. Would recommend pairing this with a vacation, perfect if you would like to bring you or/and family and learn about the country.

Now back to the stocks…

There were several businesses I came across as I was digging through the indexes where I could connect patterns of businesses I have studied in the past to these companies.

Let me go through each category that sparked my interest:

Factory stocks

I found a TON of factory and trading stocks valued at generally cheap prices so for the deep value microcap people out there, you might have fun taking a crack at this.

Though that is not my cup of tea, there were a couple that seemed worth taking a deeper look.

I previously looked at Dream International (with a market cap back then of $400M), Asia’s largest plush toy manufacturer and supplier to Disneyland, they moved a lot of their manufacturing to Vietnam and had scale advantages due to their volumes, allowing them to increase sales on similar SG&A cost-base and giving them negotiating leverage on suppliers and customers as they were competing with smaller factories. It was 4x earnings that time, but I did not invest because I was suspicious of several insider transactions I found as well as a previous scandal where lack of management controls led to their i-banking token along with $41M HDK stolen. Nevertheless, the stock performed exceptionally well given strong dividends and decent earnings growth as well as reports coming out that Dream was making Labubus for Pop Part.

And so, from this case study, I narrowed down the framework for investing in factories to the following:

1) the factory has to have scale advantages, better if in a fragmented market

3) the factory should ideally be already working with major names to draw in other big customers that could further drive sales

4) pay more attention to price as you want to be paying cheap for a business that’s not great but pretty good.

Using this framework, I came across a couple Vietnamese factory stocks that seemed promising.

Viet Tien Garment was an interesting case. It’s a $70 million company with $400M in annual sales, making it one of the largest majority private textile manufacturers in the country.

They’ve not only secured large contracts with brands like Nike, Converse, Skechers, Calvin Klein, etc but have also been growing their own brands under the Viettien name as a low-cost producer.

Sales and profits have returned to pre-pandemic levels as of 2025 amidst cost control efforts. And at a 6% div yield and just 4x earnings, this company definitely seems interesting.

TNG is the closest comp to Viettien, also one of the largest non-state-owned textile manufacturers with contracts for Decathlon and Columbia. They’re more export oriented and earnings growth have also been quite strong. This one is at 4.5x earnings with a 10% div yield.

Rock pits

I also checked out rock pit stocks: there were a bunch.

I was interested because rock pits, by which I mean aggregate mining, are generally great businesses as it’s so costly to transport gravel that for a builder, it only makes sense to source aggregates nearby your construction site.

Ergo, if you had control of a certain area, there is no rock pit outside a 50km radius that could compete with you. And NIMBY would usually prevent a competitor from setting up their own rock pit next to yours. Vulcan is a US benchmark that has performed exceptionally well by creating local monopoly rock pits on these economics.

But guess what guys, this did not apply AT ALL in Vietnam. Most of the largest quarry sites were divided among multiple players. So, a rock pit company might have their quarries shared with several other rock pit companies → no more pricing power.

And more often than not, NIMBY worked AGAINST them. As urban development reached the mining site, residents wanted the miners OUT. Most quarries failed to renew their mining rights, forcing players to acquire other mines or open a new one, which could take years.

Was promising initially, but I quickly moved on.

Infrastructure monopolies

There were a lot of state-owned or state-sponsored construction companies building various pieces of infrastructure like toll roads and bridges and many were in charge of utilities like electricity and water supply. Some of them may be worth taking a look. For example, there were a lot of stocks which were involved in water processing and supply which had pretty good returns on capital and decent growth. Nonetheless, as I mentioned last post, because the state regulates pricing for a lot of these infrastructure-plays, the vast majority of these businesses cannot exercise pricing power.

There were some private port operators which seemed interesting at first glance, Gemadept is the best example of a multi-concession operator that was not state owned. VIP Greenport is another case.

However, the problem for both of them is they would issue huge amounts of equity to finance their CapEx, diluting the shareholders.

In fact, just a note for a lot of stocks I was looking at, many times management had no hesitation issuing equity to finance CapEx and some were even proud to display that their capital base was increased due to share issuance. Obviously, this lack of care for existing shareholders is no good, and hopefully corporate governance for public companies become better overtime.

Airport plays

Akin to the infrastructure plays, there was one category of stocks which I got excited about initially.

Businesses like Nobai Cargo Terminal Services, Saigon Cargo Services, Vietnam Airport Services, to name a few, are given essentially a monopoly over the airport they’re in respectively as subsidiaries of Airport Corporation of Vietnam (ACV).

A cargo service company may own some warehouses or buildings or equipment. Ground service businesses will own a fleet handling equipment and vehicles. But because these assets are almost fully depreciated, these companies didn’t have that much fixed assets employed and with no competitors, they generally have extremely high returns on capital…on paper.

The problem is I don’t see how these companies are going to be able to reinvest their cash at high rates of return. Their reinvestment is limited to the growth of the airport they operate in. These airports are definitely growing but their top-line growth is essentially capped unless they can secure bids with other airports outside their jurisdiction.

Taseco Air Services is another airport-adjacent play that has been enjoying massive growth in recent years. They have a number of businesses throughout multiple airports including duty-free retail, restaurants, luggage transport, airport ads, hotels, and lounges. Did I name all of them?

The company was trading at 13x earnings for what I felt was an interesting group of businesses that were benefitting from Vietnam’s overall economic and travel growth.

However, the problem was the ACV. While they previously were paying the airport fixed leases, the ACV had recently demanded profit sharing instead. And this is where the issue lies: they have zero negotiating leverage over their suppliers.

I think if you want to play airport infrastructure in Vietnam, your best bet would be to go straight to the Airports Corporation of Vietnam, which controls all of the 22 civil airports in the country. While it is state-owned, you can’t ignore how good a business this is even without doing a deep dive. You have 22 local monopolies that are mission-critical infrastructure, plus plenty of room to reinvest capital to expand airports, improve system, etc. Even low hanging fruit like the immigration line would massively up efficiency. When I came in, it was a nearly 2-hour wait time. Simply adding more stalls can increase capacity by a landslide.

It’s a $7 billion company, had a major drawdown recently and is now at 11x earnings (EDIT: due to a stock dividend that raised share count it’s technically 16x, still very cheap imo)

For comparison, Airports of Thailand is a $25 billion company, and they only have 6 airports and more importantly trade at over 40x earnings!

Despite my preference for smaller companies, I think ACV is definitely worth a closer look just based on valuation and quality of economics.

Fahasa

When I visited Vietnam, I saw a lot of Fahasa stores. Fahasa is an $18 million company and Vietnam’s largest bookstore with an estimated 70% market share in imported books.

Now you might be wondering, what’s so special about a random ahh bookstore?

Well, this is not just a bookstore, it’s more like a dollar store for books, stationery, school supplies, toys, and other nifty stuff. Many textbook publishers and school supply producers distribute their product through retailers like Fahasa.

This business model is very common in Southeast Asia. The Philippines has National Bookstore which has an 80% market share in books and stationery retail with nearly $200 million in annual revenue. Indonesia’s Gramedia is another example with majority market share of their industry. Scale advantages allow these stores to offer products cheaper than smaller competitors and are also able to provide a much wider selection of items making for a better shopping experience.

These companies have been around for decades. They’ve withstood it all and have shown no signs of slowing down despite the rise of e-commerce or the events of Covid-19. This has of course been supported by the exceptionally strong retail and mall scene in these countries.

Now because these businesses are private, it will be very hard to benchmark Fahasa to National Bookstore or Gramedia, and as such more DD will have to be done to get a better sense of where this business will be going.

But at 7x earnings, you’re getting a high return on capital low-cost retail business with bottom-line growth historically at mid-low teens. Seems like a steal on paper so I shall be digging into this further.

…

So that’s a brief summary of the Vietnamese stocks I’m interested in. There are a lot of other good investments out there, but these are the ones I’m most interested in right now and wanted to share with you guys.

Other research updates:

Aside from the Vietnam stuff, I’ve been trying to go through some Philippine companies again. There' aren’t many public companies so it shouldn’t take too long to go through the entire index.

I am also looking at a couple European industrials — oh what I would give to find another Lindbergh. If you have any micro or small cap industrials or serial acquirers you could recommend, I’d love to take a look. Mo-Bruk seemed quite interesting as a regulatory-moat play for waste processing in Poland, will be devoting some time as well into studying them.

Anyways, my watchlist is crazy long right now so will be hunkering down to do research.

Share

Thanks for reading and if you enjoyed this post, would appreciate a share!

This is a preliminary discussion on a construction company with a rapidly growing cell tower leasing business that is set to dominate Vietnam’s 5G rollout. In under 5 years, I expect the operating earnings of their cell tower business alone to be equal to the 2024 EBIT of all the company’s business segments combined.

Note that I have not initiated a position. But it’s a strong watchlist item that I will be monitoring closely.

CTR trades at 21x 2024 earnings which I think is a bit too high given the risks involved. But there a lot to like and still worth discussing!

I’ve previously wrote about the rapid growth Vietnam has experienced from the China + One trend where factories are near shoring new factory capacity in Vietnam due to the cheap labour, proximity to China, and more favorable US relations. Vietnam today is one of the fastest growing economies with FDI and trade shooting through the roof.

I’ve spent the past months searching for Vietnam businesses that can absorb this growth to add to my portfolio.

Infrastructure was the first that came to mind.

However, the one problem I’ve found with these businesses is that most of them are almost 100% owned by the state.

In a lot of cases, there is no incentive for the business to increase profits, even though they are natural monopolies, instead prioritizing public good even if it means operating at a loss.

You see, these businesses typically do not own the underlying infrastructure assets, the state does.

For example, most ports run on concessions to operate the port and do not own the port itself. If the government has a problem, they can revoke the concession.

Additionally, the government might build a new port right next to an existing one and retire the old one. As such, single-concession port companies have no guarantee they can maintain their monopoly.

Railways face a different issue.

The Railway Transport JSC (UPCom: TRV), which owns most of Vietnam’s passenger railways, have indefinite usage rights but do not own the tracks or land underneath the tracks. The state simply allows TRV to operate the cars on the tracks. TRV has no control over pricing or track expansion. As such, though they have no direct competitors, they do not get to exercise the advantages a monopoly would normally have.

But!

There was one kind of infrastructure play that I think deserves just as much attention…

If there’s anything you can learn from quickly growing markets, it’s that one of the most important transformations a country needs to make to sustain its growth is to adopt strong telecom infrastructure. I’m talking cables, transmission lines, broadcast stations, cell towers.

As business and trade grows and people start living in a modern interconnected economy, data and wifi and communication is everything.

So let me tell you a story…

It’s the 90s — The United States had recently released Vietnam’s embargo and Vietnam’s relations between China and the US had normalized. The country was in full Đổi Mới reform with the goal of creating a socialist market economy.

VNPT or VinaPhone, the 100% state-owned telecom operator, was a mandated monopoly and controlled all telecom infrastructure in Vietnam.

But 1995 would change everything.

Facing pressure to create a more competitive telecom industry, the state allowed the creation of other state-owned telcos. This led to the formation of Viettel, owned the Vietnamese Military, as well as Saigon Postel, owned by VNPT and 10 other state-owned shareholders.

Furthermore, the government opened up the sector to foreign investors. This led to the creation of several new telcos most notably Mobifone by Comvik via a cost-revenue sharing agreement with the government.

But due to restrictions placed by government, such as all players being forced to use VNPT-owned infrastructure, which maintained the state-driven dominance of VNPT, as well as the severe price-cutting competition sparked by the flow of new entrants, the economics broke down and by 2012, nearly all foreign players were gone.

Comvik got out after their agreement expired, which led VNPT to acquire Mobifone, though the state mandated VNPT to separate Mobifone as a separate company with ownership transferring to CMSC, another government body, in 2018.

Though VNPT continues to maintain control over fixed transmission line, as the mobile market would come to dominate, it allowed new entrants to build mobile network infrastructure not monopolized by VNPT.

Today, Vietnam’s mobile network is dominated by 3 operators: Viettel, VNPT, and Mobifone.

These three operators (or MNOs) make up over 90% of the entire market.

You will notice Viettel has risen as the market leader, by a significant margin.

But why?

Weren’t Viettel, VNPT, and Mobifone all wholly state-owned companies?

The answer has to do with Viettel’s connection to the military of Vietnam.

Why was Viettel so successful?

You see, while the typical state-owned business was slow, bureaucratic, and hierarchical, I’ve found that Vietnam’s military-owned businesses have displayed a culture and organization that is fast, performance-driven, and meritocratic.

This was no different for Viettel.

Performance evaluations were based on hitting KPIs. On the other hand, traditional state-owned companies were content with just giving out salaries.

Ex-employees described Viettel’s culture as one of military-like discipline. Employees were expected to push as hard as they can to reach their goals and many worked long hours.

Nevertheless, structure was not too hierarchical — innovation and ability were prized with Viettel often paying highly to reward and retain excellent talent.

When Viettel entered the telco space, they acted fast to dominate the rapidly growing mobile demand with new technology whereas VNPT had continued to depend on older systems as their monopoly and infrastructure control endowed them with high margins, margins that would erode as soon as aggressive operators began charging half off.

Secondly, Viettel knew they couldn’t win fighting VNPT in the cities, so they had to focus on the underpenetrated countryside which allowed it to extend its mobile network nationwide. VNPT and Mobifone would later be forced to play catch up.

There were many other benefits of having the military to back you up:

→ The military’s goals of strengthening the army and its intelligence incentivized Viettel to focus on R&D and foreign partnerships, supplying best-in-class equipment for the business where they could be good enough for military use.

→ The military’s power over the government enabled them to lobby against VNPT’s anti-competitive practices, which influenced the increasing liberalization and pro-competitiveness of the entire telecom industry.

→ The military’s extensive land network (bases, stations, etc) and hard working military-trained workforce allows Viettel to build faster than all other competitors and provide undisputed access to land rights to build towers and infrastructure where other competitors were still forced to rely on VNPT’s infrastructure due to high CapEx requirements.

In other words, you could make the argument that being owned by the military was itself already a moat due to a combination of the culture and access to hard-to-get political and physical resources.

This leads me to the company I want to discuss today.

Meet Viettel Construction (HOSE: CTR), a subsidiary of Viettel Group, valued at roughly $450M USD trading at 21x after-tax 2024 earnings with net cash position about 8% of market cap.

CTR’s main business is operating and maintaining Viettel’s towers, lines, and network. Nearly 40% of revenues comes from the telco operations segment and Viettel Group is really the only customer.

Then we have the construction business. This segment is involved in building, installing, and reinforcing stations, antenna poles, television poles, lightning protection grounding systems, fiber optic cable systems, and more.

They are also involved in various construction projects in real estate, civil works, and industrial parks serving a variety of non-Viettel. Construction accounts for close to 25% of revenues.

Above: 1 year 270 billion VND Suburban 5A project for CDC Real Estate completed in December 2024.

Above: 1.5 year 271 billion VND An Dien Project for Eco Pearl JSC scheduled for completion by May 2026.

About 14% of sales come from Myanmar and Cambodia related to constructing and operating Viettel’s network in those countries, so that belongs to the construction and operations segments. Additionally, CTR is also involved in other businesses ranging from solar panel installation to IoT distribution.

However, what I’m really interested in is their Cell Tower leasing segment.

These towers act as leasable passive infrastructure on which MNOs will place their telco equipment to be able to provide mobile data coverage. A tower is not simply just a piece of steel but also the fiber cable wiring, rack space, etc. CTR makes money by charging tenant MNOs monthly for the space to place their equipment on it. Towers can come in the form of larger stand-alone towers (macro or ground-based) or small cell/rooftop towers, which are more common in the city.

This segment was started in 2020 when Viettel announced a push towards sharing network infrastructure.

CTR builds and, importantly, OWNS the towers and leases them to Viettel and other MNOs.

In just 5 years, the tower co went from 0.5% of revenue to 5%, and from <4% of gross profits to over 20%!

While the core construction and telco operations segment have grown gross profits at about 5% and 19% respectively, the tower co has grown profits at a whopping >75% CAGR!

CTR’s towers are largely in areas uncovered by Viettel Group’s existing tower network or places that require densification to accommodate more traffic. In bigger cities, CTR may own up to 20% of Viettel towers.

In other words, CTR is Viettel’s primary vehicle of 1) extending coverage nationwide and 2) deploying 5G

Source: MBS Research

I believe CTR’s tower co possesses the following properties:

1) a long growth runway as Vietnam’s 5G penetration is still just 25% with government mandates of 99% by 2030. Due to 5G’s shorter wavelenghts, Vietnam will probably need somewhere in the range of 100-200,000 more towers to reach total coverage.

2) a 10% lower cost base than peers due to the vertical integration + scale advantages

3) and tower economics that could reach >30% ROIC due to the high incremental margins and high switching costs.

CTR already owns >12,000 towers and aims to build 2,500 towers a year, a 15% 5 year cagr in tower count growth. I estimate the rollout to be 50% of Viettel Group’s annual total tower expansion goals in Vietnam.

Now let’s discuss what makes cell towers such a good business?

1. High predictable recurring revenue

As telecom operators optimize their network coverage, switching to another nearby tower would require you to re-optimize your network coverage, risk potential downtime which would frustrate your subscribers, and you would also have to spend to transfer active equipment from one tower to another (note that I do not yet know the exact figure). Contracts typically last 5-10 years with price escalations. Prices may increase 5-10% at renewal.

As a result, there is little reason to switch to a nearby tower even if a competitor might offer a better price.

2. High incremental margins

Towers can accommodate multiple tenants/telecom operators and the cost of having one tenant is almost the same as having two or three → operating leverage!

Most countries typically have about 3-4 dominant telecom operators, as such in most cases a tower can at most take in 3-4 tenants. High consolidation will lead to lower tenancy and hurt margins.

Currently, most telecom operators in Vietnam operate on their own towers, though the state has recently expressed the desire to have operators share their tower infrastructure going forward.

If Vietnam’s operators decide to share their future towers, the tower co tenancy ratios will go up and so will their gross margins.

3. Access to scarce locations.

Towers are often subject to local and regulatory pushback as they don’t look pretty and residents fear side effects or externalities like radiation, lightning strikes, or large fires.

I’ve founded plenty of evidence for strong local “Not in my backyard” situations where towers would face significant backlash from residents claiming it may invite lightning strikes, cause health side effects, and more.

Nevertheless, there remains no official regulation in Vietnam preventing you from building a tower right next to an existing one. And enforcement is inconsistent with evidence to believe some towers from all players lack the proper permitting and instead use political connections to build.

The picture below shows my brief breakdown of CTR’s tower economics:

If CTR is able to add ~0.5 average tenants/tower, i.e. the tenancy reaches ~1.5, ROICs crack 20%. And adding ~1 whole average tenant/tower brings ROIC to 30%.

As you can see, tenancy is the biggest lever for this thesis to work. Right now, since most towers still only have Viettel as a tenant, i.e. tenancy is close to 1, ROIC per tower today remains quite low. But as the country develops and based on the rhetoric I’m seeing from the state, I think 1.5x is within reach, dare I say inevitable. For reference, competitor OCK Vietnam already has a tenancy of 1.4x.

Essentially the key factors influencing tenancy overtime are:

1) high CapEx requirements for expanding coverage, which will be massive due to the 5G rollout

2) regulation against tower duplication and

3) high competition which will force MNOs to lease to reduce upfront costs and replicate each other’s coverage. (But not too high as we’ve seen in India where too much competition can force many players into bankruptcy leading to high tower vacancy).

I would need to publish an entire post to talk about this in detail covering the history of other telecom industries in other countries but suffice it to say I think CTR is on the path to 1.5x and my base case assumptions assume 1.5x tenancy within 5-6 years. I think 2x possible but no guarantees.

Competition

In Vietnam, CTR is the largest tower co, followed by OCK, an international tower portfolio, and other small players.

Independent tower cos like OCK have usually opted to acquire rather than build and despite OCK historically being the largest tower co, they have in recent years seen far slower tower growth compared to CTR.

But why has CTR left OCK in the dust? From my conversation with an ex-OCK employee, I learned that OCK didn’t want to deal with the politics. Building towers meant you need connections, friends in high places, or be forced to give hefty bribes.

Since OCK did not want to deal with that, they opted for acquisition, but even then it was simply worth focusing more in countries like Malaysia where OCK had better expertise.

On the other hand, Viettel has been explicit about it’s aggressive tower plans.

Chairman and General Director of Viettel Group Tao Duc Thang has emphasized the importance of CTR’s position as the top tower co:

“Viettel Construction must continue to maintain and firmly affirm its position as the No. 1 TowerCo in Vietnam. This is not just a target of scale or market share, but a strategic requirement associated with Viettel’s responsibility in ensuring infrastructure for socio-economic development and defense and security tasks in the new period.” - Lieutenant General Tao Duc Thang at the Military Conference 2025

Below shows my quick and dirty estimates for gross profit breakdown over the next 5 years. Here, I am assuming the tower business reaches 1.15 tenancy by 2030 with 20% 2-tenant discounts.

Revenue will remain a small part of the company but a big part of the bottom-line:

On a side note: seems like CTR is depreciating tower assets at a smaller useful life than the norm. For reference, American Tower had historically depreciated tower assets on 20 years useful life, and since 2024 has extended it to 30 years.

Below is from CTR’s Q3 2025 financial statement:

We can even see CTR themselves extending useful life. Check out FY 2024:

We can assume a lower useful life due to many factors including quality of structure, material, equipment, etc. But suffice it to say I believe the high depreciation is understating tower gross profits as in this case depreciation is part of cost of sales.

Nevertheless, even though I am expecting strong earnings growth, there are risks to consider.

1. While the development of communication points to the state’s desire for tower sharing, MNOs still see their towers as strategic assets and so they want to own the infrastructure they need to grow. I do not know how long it will take for the MNOs to wake up.

2. There is a key conflict of interest. Since CTR’s biggest client is its majority shareholder, there is pressure for CTR to provide steep discounts to Viettel, preventing CTR from exercising pricing power. CTR already gives Viettel a discount, charging 7.1M VND/month instead of the 8-12M VND/month most would be charging.

3. Despite strong evidence of local NIMBY resistance to towers, without clear regulation preventing cell towers from being built near each other, then the local monopoly is much smaller than benchmarks like American Tower.

Even though Vietnam may be heading towards a more regulated shared-infrastructure market, right now, I feel like the economics are just not as good as other tower cos, and so I don’t think if CTR deserves an >25x EBIT multiple like American Tower where their towers benefit from far strict zoning requirements, NIMBY, and high tower sharing among MNOs.

At 16x 2024 EBIT, CTR remains in my watchlist.

Still, I think it has the potential to be a great business so will be posting an update if I change my mind, which may be sooner rather than later.

If you think I am wrong. Please challenge me in the comments.

Let’s learn together :)

Subscribe now

Ginebra trades at 10x 2025E after-tax earnings, 9x 2025E FCF while generating returns on invested capital above 50% (check part 1 again for my ROIC calculations). After-tax profits have grown at >30% for the past 5 years

Meanwhile Emperador, a business which is far worse as I’ve extensively explained, trades at 40 times 2025E after-tax earnings!

I believe Ginebra deserves a price to earnings of at least 20x. And you get 5% dividend yield for waiting.

Even if I had $10 billion dollars, I doubt I could cause a dent in Ginebra’s business. Both Tanduay and Emperador have launched gin brands before with grand campaigns but this did not do anything to Ginebra’s market share.

Many have tried, all have failed.

As I mentioned in Part 1, mass-market liquors like Ginebra depend heavily on sari-sari stores where most of the end-customers buy from. These are small convenience stores scattered across the country, especially in urban areas, that allow you to buy bits of everything from snacks to toothpaste to of course, alcohol.

There are about 1.3 million of these stores. For reference, the US only has 152,000 convenience stores.

This means 1) an extremely fragmented customer base which gives you and your distributors more negotiating power, but it also means 2) you need both extensive logistics capabilities and customer relationships (or lots of accounts with customer relationships). I think Ginebra, through their parent SMC, has probably the most extensive distribution network in the country due to SMC’s cross-island and nationwide warehouse and route network as well as having a large account base that are able to cater to most sari-sari stores in the country.

I will first go over the supply chain and distribution of Ginebra and then discuss their marketing strategies, such as their basketball team (yes they have a basketball team) and then thoughts on management.

Production

GSMI is like many other liquor firms in the Philippines: owning a portion of the distillation process as well as all of the bottling process, same with Emperador and Tanduay. Unlike a business like Coca-Cola, covering the bottling process requires far more CapEx. Capacity increases are inevitable as sales volumes grow. For Ginebra’s case, they own one of Asia’s largest distilleries which has no doubt allowed them to achieve immense sacle and they expanded facilities already in 2021 for $110 million. In 2024, there was about a billion pesos est spent in growth CapEx on capital projects, likely to increase capacity to facilitate volume growth.

Let’s go over the Ginebra production process from feedstock to bottle.

Ginebra sources 38% of its alcohol internally via its subsidiary Distileria Bago. Molasses are first purchased from sugar mills in the Philippines, India, and Indonesia, which are then sent to Bago’s distillery to convert to alcohol. The rest are sourced internationally, a deliberate move to diversify supplier base. They tried producing alcohol from alternatives such as cassava starch in 2001 but it seems this didn’t stick. Byproducts like wastewater and technical alcohol are sold off.

The distilled alcohol is then sent to the bottling plants. About 2/3 of bottles are sourced from their second-hand retrieval program. The other 1/3 are produced by San Miguel Yamamura, San Miguel’s packaging arm for its beverages. Flavouring is made with extra sugar purchased locally and flavouring agents imported from various Asia-Pacific traders.

The company owns 6 bottling plants mostly in Luzon and Visayas: 3 via the main company in Mandaue City, Cebu (mid-Visayas); Sta. Barbara, Pangasinan (north-Luzon) ; and Cabuyao, Laguna (center-Luzon) and 2 via East Star Pacific Bottlers in Cauayan, Isabela (north-Luzon) and in Ligao City, Albay (south-Luzon). They acquired East Star in 2012 to expand bottling capacity for $5M USD.

Having some of the largest distilling and bottling facilities in the country plus the largest sales volumes leads to supreme economies of scale. I believe these scale advantages are what allow GSMI to achieve better gross margins compared to Emperador and Tanduay despite all three brands charging almost the same price per ml.

→ This is especially the case for Emperador who not only has a significant mix of premium higher margin brands but who also pays far lower excise taxes due to the lower alcohol levels compared to a bottle of Ginebra. In fact, each Emperador bottle would have a bit less than 40% of its price in excise taxes while excise taxes make up nearly 50% for each Ginebra bottle. Excise taxes make up close to 60% of Ginebra’s cost of sales. This implies the Emperador brand has a far poorer cost structure. Emperador’s brandy segment had a GM of 21% in 2024 compared to GSMI’s 24%.

→ Tanduay achieved a GM of 15% in 2024 (note: Tanduay, Emperador, and Ginebra all use weighted average cost for inventory calculation). They have a similar ABV to Ginebra and should have a similar excise tax. Therefore, it has nothing to do with the taxes and everything to do with raw materials and economies of scale. Probably the biggest difference in cost structure from GSMI is that Tanduay (and Emperador) require aromatics and barrel aging.

→ One thing to consider is that the San Miguel Corporation produces probably one of the largest volume of beverages in Southeast Asia. San Miguel beer is the primary beer brand in the Philippines and the most consumed alcoholic drink. What this means is that San Miguel beats all others when it comes to bulk purchasing. However, I think the added cost savings in terms of raw materials aren’t huge as really the only raw material that overlaps with Ginebra and San Miguel is glass for bottling via San Miguel Yamamura.

Distribution

Lackluster infrastructure and road systems, stemming from slow development and corruption, and being a multi-island country means its it’s much harder to ship.

The barriers to selling nationwide economically in the Philippines is enormous. Even early on, tapping into the wide SMC distributor network was incredibly valuable to getting Ginebra product to all corners of the country.

Read this from Ginebra’s 2000 annual report:

SMC controls a lot of the country’s infrastructure, especially in Luzon.

They operate several port terminals, inter island carrier fleets, hundreds of warehouse depots, and toll roads. All these allow for significant reach plus discounted transport cost.

For comparison, Tanduay has about 22 warehouses while LT Group, its parent, is really a banking and tobacco conglomerate. Emperador’s parent, Alliance Group, leaned more into casinos and hospitality. It’s really SMC that has overwhelming dominance in infrastructure.

SMC is also building the New Manila International Airport, MRT-7 railway which connects Quezon City to Bulacan, which could potentially allow for further connectivity.

….

Once bottled, finished goods are sent to Ginebra’s 100 exclusive dealers via SMC’s network of truck fleet and warehouses that operate across the country. Each dealer is contracted to serve a specific region.

These dealers then sell to mostly to sari sari stores as well as wholesalers and large accounts such as Puregold, the Philippines largest supermarket.

Ginebra operates 17 sales offices which serve 158,000 accounts! One account for every 730 people.

What’s interesting is that having an extensive account network is very helpful in growing end-user accessibility. This is because many retailers buy from other retailers physically rather than having it delivered from wholesalers or distributors. One sari-sari store owner I found online said they sourced everything from the local market a few kilometers away. I found a similar situation on reddit where a store owner would buy from another nearby account and then resell, eventually finding the supplier of that seller and cutting out the middle-man. Either way, the way these products get around to the end drinker seldom involves formal contracts: further fragmenting the customer base and making sari-sari store reach more difficult.

(This might reverse if Purgold ends up consolidating the market. Puregold is often able to outprice a lot of mom & pop supermarkets and offer good membership discounts. I think as Puregold grows, slowly more and more sari-sari stores will buy directly from Puregold rather than their usual local markets. This will give more leverage to large bulk buyers like Puregold.)

…..

Ginebra sells to less than 12% of the country’s sari-sari stores. There is still room to expand to more routes and locations with minimal Ginebra presence. Within Luzon however, I think Ginebra is pretty saturated in terms of sales volumes. Visayas and Mindanao growth will mostly come from Vino Kulafu, though it’s going to be hard to beat Tanduay. I think instead, the growth runway opportunity comes from the following:

1. The B2Bar opportunity looks great if you’ve had the chance to visit the Philippines. Rapidly growing urbanization will promote a much larger commercialized night life. We are seeing low-mid-end karaoke rooms, clubs, and neighborhood bars pop up everywhere.

2. There is a ton of intra-migration towards Luzon as people seek white-collar jobs and opportunity. Many regions in Luzon outside of Manila are growing really really fast. The country overall has a fast growing population with a young median population compared to developed nations.

3. Greater disposable income across the board has led to a growing lower-middle and middle class. With higher wages, I think GSMI could price up continually. Now, there is a sweet spot, we might have people trading up to premium gins if this is the case but I think this is very unlikely.

With all these factors in mind, it is exceedingly clear that the main beneficiary of this will be GSM Blue. It is already the most used for cocktails and popular with the younger generation.

Nonetheless, the growth runway is not primarily in terms of sales volume but in pricing power, especially as the country is developing fast and incomes rise.

Marketing Strategies

Someone on VIC who wrote-up GSMI said this about their basketball team:

“GSMI owns Barangay Ginebra San Miguel, one of the leading professional basketball teams in the Philippines. Its costs are not separately disclosed. While this is not rare in emerging markets (e.g. Aefes in Turkey has a similar setup), it really is not good governance.”

I disagree with this for a couple reasons.

Firstly, having a basketball team in the Philippines is ENORMOUS for marketing. Basketball is the number 1 sport by a longshot.

Barangay Ginebra is probably the MOST popular team in the entire country. For a top PBA team to be owned by Ginebra is huge.

The San Miguel Beermen is the most successful with 30 titles to Barangay Ginebra’s 15 titles. Following is Magnolia Chicken Timplados with 14, all three indirectly owned by SMC. There are no other liquor companys with basketball teams. Tanduay used to have a team called Tanduay Rhum Masters but they disbanded twice after failing to secure any titles. Neither Emperador nor Tanduay currently have a team.

Even if Barangay Ginebra is probably operating at a loss, I think of it as an ongoing nationwide sponsorship campaign and incredibly valuable for mindshare.

Advertising is a big part of GSMI’s marketing strategy. While Emperador has long tried to position themselves as classy with actors in suits and ties, GSMI has truly leaned into appealing to the mass market, focusing on “masculinity” and casual as core themes. A big part of the brand is definitely this idea that Ginebra is a man’s drink, a rite of passage. Meanwhile GSM Blue’s marketing looks more modern and minimalist to appear to younger people and bars.

GSMI spends less than $30 million (P1.73 billion) a year on advertising and promotion, which translates to about P1.3/350ml bottle in marketing cost. For an ave. price of P70/bottle at 350ml , this gives a return on ad spend of over 54x!

Advertising and sponsorships isn’t a big requisite to grow long-term as the existing brand equity does 90% of the job. Still, it’s useful to see how Ginebra thinks of their position and how they reach their target audience.

Management

As noted by other investors, one of the major changes for GSMI over the past couple decades have been management and capital allocation. Let me explain:

Back then, the company was run by its long-time CEO Eduardo Cojuanco…

In 2010, GSMI took out P1.5 billion in long-term debt and increasing short-term debt by >30% from P3.8 to 5 billion, raising D/E from around 50% the previous year to almost 80%. Still, coverage was good, until Gran Matador was hit the following year, plunging GSMI into unprofitability exacerbated by the interest expense.

Now the short-term borrowing was drawn to add working capital to support growing volume. However, the long-term borrowing was drawn to “fund the permanent working capital requirements of the Parent Company.” This was bad because the parent company was effectively drawing its own debt from its subsidiary and it was at a high interest rate as well, (7.25-7.89%).

Back then, GSMI was also involved in a lot of different things. In 2007, they had expanded into Thailand which was never profitable. They also owned several non-liquor brands. In the 2000s, they were producing vinegar, condiments, sauces, and even jelly snacks. GSMI also acquired several brands from San Miguel Beverages as part of Cojuanco’s desire to consolidate more brands under GSMI. This included Magnolia fruit drinks, iced tea, San Miguel bottled water, and more.

So clearly GSMI had issues. With the decline of Gran Matador and subsequent profitability issues, change had to happen.

….

During the pandemic, Eduardo Cojuangco, then the longtime CEO, died, transferring the title directly under Ramon Ang. He is currently president and CEO of Ginebra. In 2024, he also became CEO and Chairman of both SMFB and SMC.

Ramon Ang was originally a car mechanic when he met Cojuanco, who became CEO of SMC after buying out a lot of the Zobel family’s stake in 1983. Now Cojuanco was known as a crony of the former president Ferdinand Marcos, famous for the martial law dictatorship.

Ang started in the cement sector of San Miguel, and later served as temporary president when Cojuangco had to flee the country alongside the Marcoses during the Edsa Revolution in the 80s. He eventually returned and tried and failed to run for president.

Later on, impressed by Ang’s management, Ang was elected vice-chairman in 1999 and Cojuangco eventually appointed him as President in 2002.

Under Ang and Cojuangco, San Miguel acquired companies with huge market share and diversified the firm into many areas, such as Southeast Asian alcohol and oil markets. This era saw San Miguel grow so much that it now accounts for around 6% of the Philippines’ total GDP!

Ramon Ang gradually assumed leadership and by 2012 had acquired majority of Cojuanco’s shares in SMC. By this time, Cojuanco was 77 and was welcomed the leadership change.

Left is Ramon Ang, Right Ed Cojuanco

Reading Cojuanco’s shareholder statements as long back as I could find, it seems Cojuanco was trying to build a bigger and bigger conglomerate which led to what I would describe a “mess”.

This involved numerous acquisitions, entry into different food and beverage industries, and adding new brands. Diversification appears to be one of Cojuanco’s goals, which led focus away from core brands under GSMI. GSMI had several random brands while SMC’s other subsidiary San Miguel Food and Beverage also had a mix of alcoholic and non-alcoholic brands, so the structure was messy.

My belief is that by 2012, Ang was really the one calling most of the shots and is thus the mastermind behind most of the GSMI reforms during the 2011 period of poor performance.

I think Ang’s direction for GSMI was really cutting out the fluff, streamlining the conglomerate, paying down debt, improving governance, and most importantly, focusing on Ginebra and Vino Kulafu.

Ang made several changes to the company that helped bring it out of unprofitability:

1. By 2013, they effectively ditched Gran Matador and redirected marketing effort and spend to Ginebra San Miguel and Vino Kulafu as well as expanding GSM Blue’s product line up.

2. Ginebra divested all non-liquor brands in 2015, making Ginebra a pure-play liquor company for the first time in many many years.

3. In 2017, COO Bernard Marquez was replaced by Emmanuel Macalalag (In 2024, Macalalag was promoted to COO of SMFB, with the GSMI COO position given to Cynthia Baroy, the previous CFO and who has been at Ginebra since 2006). Ramon Ang moved from Vice Chairman to President.

4. In 2017, SMC transferred its ownership stake in GSMI to San Miguel Food and Beverage, streamlining the group structure and aligning all food and beverage businesses under one integrated holding company.

5. In 2020, a dividend policy was released whereas before dividends would be upped and cut at management discretion if they needed cash. They also redeemed preferred shares.

6. They wrote off the loss making Thailand JV.

7. Capital allocation was redirected from business diversification to paying down debt, paying down all long-term borrowings in 2021. The next year, they placed over 20% of their cash balance into 7-year time deposits generating 6.9% in interest annually. While there are probably better ways to use the case, it sure beats letting it sit in the bank. The debt investment has since contributed to positive net interest income.

GSMI was brought out of unprofitability in 2016, thanks in part of consistently growing core brand sales amidst the same amount of operating costs.

Overall, Ramon Ang is no doubt an excellent operator. Like many Southeast Asian conglomerate founders, all his children hold high positions with SMC and SMFB. It’s a typical family-run business, though that in itself is not that big an issue. I think the concerns come down more to his net worth and age:

Ginebra San Miguel is 75.8% owned by San Miguel Food and Beverage, which is 88.8% owned by the San Miguel Corporation, which is 66.1% owned by Top Frontier Investment Holdings. Ramon Ang is the second-biggest shareholder of Top Frontier Investment Holdings, raising his stake from 26% to 35% in 2023. So Ramon Ang collectively owns 16% of Ginebra San Miguel, and he hasn’t sold any of it (side note: 52% of Top Frontier is still owned by the Zobels by the way, though Ang is the President and CEO).

This puts his GSMI stake at about $224 million USD. However, after selling his stake in Eagle Cement to SMC (which is clearly him using his own company to cash himself out bruh), he personally pocketed probably over a billion dollars. Bloomberg puts his net worth at $3.5 billion. I personally think it’s not that high because his Top Frontier is worth just $400 million dollars, even though their stake in SMC should be worth $1.8 billion. If you consider his Eagle Cement cash out and ownership in Top Frontier, his total net worth would be closer to $2-2.5 billion. Either way though, GSMI is a meaningful but not a high portion of Ramon Ang’s net worth. Moreover, the guy is 71, and I’m not sure if he has the drive a younger CEO would have.

I think he just needs to keep the flywheel running and not burden GSMI with parent company shenanigans, which to be fair he has done a good job since taking the helm. Though I believe Ginebra is a business good enough for an idiot to run.

….

Some Projections

I’m expecting around 5% annual growth in prices and volume growth. Eventually, I expect volume growth to be in the 2% range while price increases can continue to be about 5%. Remember, there hasn’t been a single year price increases have been below 7%. This should lead to after-tax earnings and free cash flow growth of 10% or more (could be lower with future growth capital projects).

With no multiple expansion and dividends reinvested, I’m estimating an annualized return of at least 15% which is already conservative compared to historical earnings growth. If P/E expands to 20x which I think it should, returns could be over 40%.

GSMI is simply an excellent business with long-term pricing power while trading at a steep discount to peers.

Thanks for reading!

If you would like to show your support, I would immensely appreciate it if you subscribed to Dunamis Investing. By subscribing, you can stay up-to-date on new write-ups, research, commentary, and portfolio changes.

A bit over 3 years ago, I was a freshman at the University of Toronto. Attempting to join our school’s value investing club, I submitted a short 15-page report on the most popular liquor brand where I come from.

Now, I did not invest in this company.

3 years ago, it was trading at 7.5x after-tax earnings while generating >30% ROIC. The company was growing sales double digit every year and margins were constantly improving, not to mention is was a market leader.

So why did I not invest?

You see, back then, I was a new investor.

Back then, I knew little about the concept of pricing power, nor have I read classic mental models of See’s Candies, Maotai, or Coca Cola.

I knew a great brand was wonderful, but I did not understand how it could produce a compounder.

Well, I’ve been kicking myself in the ass ever since…the stock I pitched has since nearly tripled!

I knew the fundamentals have not changed much, and in fact was still trading at relatively cheap multiples. Yet the feeling of missing out has prevented me from truly revisiting this company. But now I have gained greater temperance, to not let previous misgivings cloud my thoughts.

In this post I will be discussing Ginebra San Miguel, the most popular gin brand in the Philippines and the worlds largest gin producer. Ginebra trades at ~10x 2025E after tax earnings, has consistently delivered double digit growth and high ROIC, and still has a huge growth runway to continue compounding.

…

Ginebra San Miguel (PSE: GSMI) | Market cap: $1.36 billion USD or P80 billion

Ginebra San Miguel is the world’s largest producer of gin and the Philippines is the world’s largest consumer of gin, taking >40% of the world’s gin consumption.

But why is gin so popular?

Gin was first introduced to Manila during the British occupation of Manila in 1762 and was initially accessible only to the Spanish and Filipino upper class via imports. In 1834, Ayala y Compania built the first local distillery and sold gin under the brand name Ginebra San Miguel as well as other liquor products.

In 1879, chemist Anacleta Del Rosario developed a formula to purify alcohol with a stronger body and smoother taste. Ayala y Compania acquired the formula and with new industrial equipment from France, started to produce Ginebra for the mass-market.

Gin became popular among the working class because it was affordable and available. Gin was cheaper to produce vs other liquors as it required no aging and was made with molasses, which was much cheaper than local ingredients used to make liquor such as nipa palm and imported feedstock. Ginebra thus became the most affordable commercial hard liquor of the time.

In 1924, Don Carlos Palanca Sr. acquired Ayala y Compania along with the Ginebra San Miguel brand and built a new distillery in Tondo, pioneering the production of gin using molasses. After World War 2, Palanca Sr. built mechanized bottling systems and expanded distribution throughout the country. At this time today’s most popular Ginebra bottles were introduced, including the iconic kwatro kantos.

Culturally, drinking is deeply embedded. Drinking sessions, called inuman or ginuman (gin + inuman), are common where liquor is shared with friends and family. This is typically at home but is also often done during celebrations, nights out, and during karaoke sessions. As the premier clear liquor, gin is also used frequently as the base for cocktails. Gin-mixing is extremely popular, especially among younger people. And this isn’t just a bar thing, people will just mix gin at home with everything, such as gin + coffee in the morning and gin + fruit juice after dinner. A lot of new combos gained popularity on Tiktok and Facebook. The most popular are Gin Pom, Gin Mojito, and Red Alert (this would definitely kill you if you drank it everyday).

The conglomerate San Miguel Corporation, which owns San Miguel beer, bought Palanca Sr’s company in 1987. Currently, the firm is 76% owned by its subsidiary San Miguel Food and Beverage. Today, Ginebra San Miguel comprises >90% of the country’s entire gin consumption, selling over 410 million liters every year.

We estimate the Ginebra San Miguel brand accounts for at least ~90% of total sales volume and total sales. In 2024, Ginebra and the company’s second most popular drink Chinese wine brand Vino Kulafu, comprised ~99% of total sales. The other 1% of sales come from the company’s other brands, including Antonov Vodka, Primera Light Brandy, Anejo Gold Rhum, Don Enrique, and more.

—

The Ginebra mindshare

We now turn our attention to Ginebra’s source of pricing power, which I believe comes from their mindshare. This is a mix of:

1) cultural embeddedness and iconography of the Ginebra brand

2) cost leadership/affordability compared to all other clear spirits

3) accessibility across most of the country for the working class

1. Branding

Ginebra is an almost 200-year old brand with a long history. The brand has been a household name since the early 20th century.

It’s so ingrained in Filipino drinking culture that when people think of inuman, there are only a few names that come to find: Ginebra is one of them.

Fernando Amorsolo, the most famous Filipino painter and friend of the Ayalas who sponsored his studies in Madrid, designed the bottle art for Ginebra in 1917. Despite some modifications over the years, the logo art retains all the important pieces of the original drawing. This branding continues to be one of the most iconic piece of commercial art throughout the country.

Local vernacular have slang names for several of their products. For example, when someone asks for a bilog, they are referring to Ginebra’s Gin Round series. Or when asking for a kwatro kantos, they are referring to Ginebra’s Frasco series. Other slang such as gin bulag (gin-blind) and ginuman are also common and almost always refer to Ginebra’s original product line-up.

Search up anything related to gin + Philippines on Youtube and you will not find anything other than Ginebra pop up (or their basketball team).

GSM Blue is a lighter version of the original and is very popular among younger people and new drinkers. GSM Blue have variants made specifically for cocktails such as GSM Blue Pomelo and GSM Blue Mojito. They have grown substantially within the past few years and is now the world’s fastest growing gin brand according to Drinks International, demonstrating successful efforts to bring the younger generation into the Ginebra customer base. As remarked in the 2025 Millionaire’s Club report:

“The continuous growth [of GSM Blue and Ginebra] in 2024 was driven by a rising nationwide user base driven by the younger users. Gin mixing is widely popular here as it provides our consumers another new way of enjoying the gin drinking experience apart from drinking it on a per shot, with ice basis.”

I think it’s worth also talking about Vino Kulafu. Vino Kulafu is mostly consumed for the nutritional benefits associated with the Chinese herbs in its ingredients. Vino Kulafu is popular because it stems from Kulafu, which was one of the first Filipino comic book heroes back in the 30s and is now a cultural icon. Since 2017, the drink held over 60% of the market share by volume. I estimate they are currently selling 4M cases annually.

There have also been plenty of superstitions surrounding Vino Kulafu that have contributed to sales. According to some customers, drinkers believe it will improve reproductive health. The drink is most popular in the south, Mindanao and Visayas, where it’s so culturally embedded that Vino Kulafu’s slogan became “Mindanao’s number one medicinal wine.” (ironic as Mindanao is majority Muslim). Most of its competition come from foreign brands and See Hok Tong produced by Destiliera Limtuaco. We estimate Vino Kulafu has a >60% market share for Chinese wine and comprises about 7.6% of total GSMI sales volume.

2. Cost leadership

Ginebra is the most affordable gin and clear spirit in the market by far, priced a bit over $1 dollar for a 350ml bottle. (Vino Kulafu has a similar price on ave.)

Price is king in a country where the daily minimum wage is below $12 a day. The next largest gin brands such as Gordon’s and Bombay Sapphire are over 8x the price.

(GSMI’s pricing is similar to other leading spirits such as Tanduay rhum and Emperador brandy, though they are aromatic spirits. We will talk about them later.)

In terms of taste, Ginebra is smooth and almost syrupy, probably because it’s made of molasses instead of grain. Flavour profile is distinct in that Ginebra is less juniper and more anise. However, since Ginebra positions itself as a cost leader, it’s not known to have the best taste — obviously a premium gin would beat Ginebra any day. But at 40% ABV, it’s really affordable if you want to get a kick or use it as a base for cocktails, which I think is really the purpose of most clear alcohols. Now Ginebra’s longtime liquor competitor LT Group (which owns Tanduay) released gin brand Ginto in 2021. But it is slightly more expensive and is not clear since it’s barrel-aged.

3. Accessibility

Ginebra is distributed widely, especially in Metro Manila, and can be found in most sari-sari stores, these are little variety stores targeting the working class. Sari-sari stores account for 40% of all FMCG sales and are most common in urban sprawl but are also scattered around rural provinces.

In the next post, we will discuss the Ginebra supply chain and how they are so accessible.

Due to Ginebra’s mindshare, Ginebra has been able to raise prices continually while maintaining volume growth and market share. In fact, Ginebra has been able to keep prices above inflation most years. Most retailers and sari-sari stores I see have simply passed each price increase to the end-consumer, maintaining markups of 15% (stores typically markup 15-20%, so this is at the lower end).

The chart below shows GSMI’s aggregate ave. price/9L case for nearly the past couple decades.

While the Philippines average inflation rate since 2010 has been 3.4%, Ginebra has been able to raise its prices 6%.

Price increases have been even steeper since the pandemic after excise taxes on alcohol were increased: ad valorem tax went from 20% to 22% of net retail price in 2020 (Ginebra upped their prices 16%) while specific tax per proof liter has increased close to 10% per year and is now set to increase 6% per year (now P70 per proof liter). In fact, in 2024, 44% of sales went to paying excise taxes vs 36% in 2020.

Ginebra seems to be able to continually pass on these increases to their customers — gross margins have remained at 24-25% over the past decade.

Revenue growth has averaged 10% since 2007, and volume + pricing have continually gone up with the exception of 2011-2013.

All the while SG&A has been growing at a slower pace, leading to operating margins coming from 4% post-2013 recovery to nearly 16% as of Q2 2025.

This divergence between earnings growth and costs have also translated to capital employed in the business. This, and due to the huge P14 billion net cash position, has led to high returns on capital employed.

…

But what happened in 2011-2013?

You see, prior to 2011, about 23-28% of sales have been coming from Gran Matador, Ginebra’s brandy line which was released in 2003 to compete against the most popular brand Emperador.

Early on, Gran Matador gained traction and was seemingly becoming a mainstay of Ginebra’s product line. However, in 2011, Emperador released one of the best selling alcohol launches of all time: Emperador Light. The year Emperador Light was launched, the brand doubled its sales volume from close to 10M cases to 20M. The next year, it was selling 30M.

Emperador Light was sweeter and had a lower ABV of 27.5%. While Gran Matador tried to undercut prices, Emperador focused on brand positioning and won the younger generation. In 2011, Gran Matador went from selling 10M cases a year to 2M.

According to Drinks International:

“Where once San Miguel’s Gran Matador was the brandy du jour now it is Emperador. When asked, consumers told DI they switched for its rounded, sweeter taste and slightly lower abv. The spirit is mainly drunk neat or with a mixer.”

Meanwhile, Ginebra San Miguel sales volumes increased 25% est though saw a 17% decline in 2012. In response, management introduced several new GSM flavors including a Light version of GSM Blue at also 27.5% ABV and other GSM flavors.

Then, in 2013, a sudden tax hike, known as the Sin Tax Reform increased specific tax rates on distilled spirits by over 50% and ad valorem tax rates by 33%. Tax reforms also hiked excise tax on lighter alcohols like beer in a similar fashion. The hike was implemented just 12 days after announcement of signage, leading to a nationwide slowdown in wholesale purchases. This did not affect the company as bad as 2011 though and Ginebra San Miguel sales volumes were only down <4% for the year.

Nevertheless, unpredictable tax levy is one concern I have as if future tax increases go too far (or are sudden hikes like the 2013 case in comparison to planned increases for the following years giving wholesalers time to prepare), customers of all alcohol brands may have a hard time paying up for their favorite drinks.

While Ginebra San Miguel recovered fast, Gran Matador did not and would eventually be rebranded as Primera and Primera Light, making <1% of today’s GSMI portfolio. Management also sold off all non-liquor businesses which at the time comprised about 5% of sales — Ginebra San Miguel would once again become the main brand of the company with >90% of sales volumes and with sales volumes growing about ~8% each year since.

—

Liquor Game of Thrones

As Ginebra utterly dominates the gin market, the real competition is really substitute liquors. Accounting for just the Ginebra brand, we estimate they control about 44% of the Philippine liquor market in terms of sales volumes in 9L cases. The table below shows our estimates for market share over the years.

Sources: Annual reports, Millionaire’s Club liquor reports, Euromonitor

As you can see, the top three players control about 80% of the market in terms of volume. You can also see that the massive sell-out of Emperador Light was mostly at the expense of smaller brands. Gran Matador would have had approximately 17% of the market in 2009. I attribute the rise in the share of “Others” to:

1. Growing disposable income have led to an uptick in more upscale brands. The Keepers Inc. is one of the largest importer of European and American brands in the Philippines. Products include Johnny Walker, Pernod, Alfonso, etc. They have close to 10% of the market by value.

2. Vino Kulafu has reportedly grown well during the pandemic. I estimate they went from 3M cases in 2018 to nearly 6M cases in 2023 and have since subsided. I don’t know what specifically caused this spike, but Vino Kulafu likely now has about 4-6% of total spirit volumes.

There are several reasons why I think Ginebra has been able to continually grow its market share while Tanduay and Emperador have seen less success:

→ Sari-sari stores I contacted have told me Ginebra and San Miguel beer (also owned by parent SMC) outsells all other alcoholic drinks. Will expound more on the distribution network of Ginebra in the next post, but I think utilizing SMC’s massive trade network + marketing channels across the country has been a major driver of Ginebra’s popularity.

→ Because Ginebra is most popular in Metro Manila and Luzon/North (whereas Tanduay is most popular in Visayas/Mid-South), the country’s urbanization and migration of people from the province to city centers has been a huge boon for the company. I actually think Tanduay is GSMI’s biggest hurdle for future geographic expansion as Tanduay simply has better mindshare in Visayas and has a similar century plus long cultural heritage. The angle I see GSMI taking is really leaning into the clear alcohol position and taking out Tanduay White.

→ Though all three are priced almost the same, Ginebra and Tanduay have a long >100+ year history and enjoy deep cultural embeddedness while Emperador only hit the market in the 1990s. While they enjoyed huge popularity with Emperador Light, it seems it just did not stick with wildly shifting consumer preferences, the same factors that killed Gran Matador. Since the mid 2010s, Emperador has seen a major decline in sales and market share. I think one reason for this is the parent company Alliance Global deciding to diversify their liquor business internationally. From 2013-2019, they made several large acquisitions, acquiring big Western brands such as Bodega San Bruno in 2013, Whyte & Mackay for £430M in 2014, Beam Suntory’s brandy business for €275M which came with a basket of brands like Fundador, and Casa Pedro Domecq in 2019, which controlled 57% of the Mexican brandy market. This spread them thin, especially in marketing effort. Meanwhile, GSMI was in the process of concentrating on Ginebra San Miguel, selling off non-liquor assets (they owned some snack businesses before). I will say where Emperador has it good is its popularity for lighter drinkers due to lower ABV. This also means Emperador isn’t as affected by excise taxes. Nevertheless, it seems GM for Emperador’s brandy business and Ginebra has been quite similar at near 25%, which implies the actual Emperador brand drinks have a lower GM than Ginebra as a good chunk of their business is now premium Western brands.

→ As I mentioned, Ginebra’s lead as a clear spirit makes it particularly favored for drink mixing while Emperador and Tanduay are less so (except for Tanduay White, a variant which is a clear rhum). What this means is Ginebra is a mainstay for drink mixing and cocktails, adding, even if a small part of the total business, an extra layer of dominance within b2bar liquor sales.

The table below shows 9L case volume sold est.:

Realized this post is way too long so I’ll continue this in the next post. There, I will discuss the distribution network, marketing strategies, management, balance sheet, corporate history, growth runway, and thoughts on valuation.

I’ve been busy with interviews these past couple weeks so the next part will probably come out a bit later.

Subscribe and comment. Thanks.

To those who read part 1, thank you for waiting. I just graduated and spent some time with family over the past couple weeks. On top of that, I’ve been trying to start a business.

Thanks for your support as always!

—

Competition

Property Databank competes with horizontal solutions such as Excel and ERP systems like OBIC as well as a variety of property management vertical software products.

Our channel checks indicate T2TR as ‘@property’s primary property management software competitor. The main difference between T2TR and ‘@property’ is that ‘@property’ covers a wide range of property management tasks, though specing more into maintenance, while T2TR focuses more on the fund management side focusing on features such as financial data, fund reporting, and forecasts, but is lacking severely in facility management. Hence, T2TR is extremely popular among investment managers and REITs, with 80% of J-REITs and private REITs using the system. Many REITs actually use both T2TR and ‘@property’, using the former for financial metrics (e.g. NOI, FFO calculation, forecasting) and ‘@property’ for daily operations and everything else. It’s important to note that T2TR charges based on fund size/AUM while ‘@property’ charges on building count, so pricing is not exactly comparable. Both rely mostly on inbound and general market presence.

I think T2TR does not pose a serious threat as their products seem to focus on different things. Moreover, ‘@property’ is dominant in maintenance, which I think is the chokepoint while simultaneously enjoying a richer product feature set. However, T2TR will be an obstacle for ‘@property’ in getting their customers to adopt ‘@property’s asset management features.

Probably the largest competitor PDB competes with is OBIC7, OBIC’s flagship ERP solution that commands a larger market larger than the next two competitors put together. OBIC7 is customized to specific client needs, which allows it to cater to property managers which have property-related accounting and operational metrics. As OBIC has been dominant in the market for so long, there is a switching cost customers must overcome to switch completely over to ‘@property’.

Overall, it seems that the operations and real estate people prefer ‘@property’, the finance/IR people prefer T2TR, and the accounting people prefer OBIC7. I think PDB is kinda unique in the sense that as their capabilities are so vast, they don’t 100% directly compete with any one player, each choosing to focus on a different region of property management, unlike other competitive dynamics where it really is a race to have the most and best features (See Plus Alpha Consulting vs Kaonavi)

There’s plenty of whitespace open for the taking. ‘@property’s feature depth and control of the chokepoint should protect it from losing its dominance in the coming future.

—

Management

PDB has 22% insider ownership. Co-founder and Chairman Toshimasa Itaya owns 9.4% while co-founder and CEO Hideki Takashi owns 6.5%.

The largest shareholder is partner Shimizu Corporation, Takashi and Itaya’s former employers, who own close to 25% of the company. They have not changed their position since IPO.

The next largest shareholder is Ken Corporation which owns 12.5% of PDB. Ken Corporation is a property management company that handles rental, sale, management, and brokerage of high-end rental apartments, houses, and offices. They manage close to ¥260B in property. Ken Corporation was one of PDB’s first big investors, and have maintained their 12.5% stake since IPO.

The third large shareholder is investment firm Hikari Tunshin, a holding/investment company that has generated an above 15% IRR for nearly a decade. They are commonly compared to Berkshire Hathaway, and management have indicated that they are value investors. Hikari acquired 7% of PDB is 2024 and has since increased its stake to over 12%. Though due to PDB’s size, it remains a miniscule position in their trillion yen portfolio.

Overall, a healthy insider ownership and seemingly confident corporate investor base.

Management has been conservative with capital allocation like many other Japanese cos: no debt, dividends over stock repurchases, large cash balance. Cash is mostly spent on capitalized software expenses.

Feature roll-out has recently been focused on expanding the other ‘@’ products, end-to-end transactions/tenant portals, and utilizing its huge dataset to expand into appraisal and evaluation. The company is also building out AI features, such as features that allow users to extract information from receipts and lease documents, AI data analysis features, and AI customer support bots.

Management has expressed interest in doing M&A. They acquired Reivo in 2024 (8x operating income) to improve their technical talent in mobile app development. We can see concentrated effort to improve UX/UI from their outdated look. This is especially apparent in their newer and add-on products:

It’s also interesting to note that PDB hosts their own private cloud servers via internal data centers rather than outsource to AWS/Azure. This provides control over the database and reduces costs, though it presents larger operational risk as they have to manage their own infrastructure. Management plans to offer IaaS products, called Data Lake, to other system integrators working with customers in the real estate space, though I’m skeptical of how they will fare in that front.

—

There are some concerns I have with management and the business.

One is that management is overpromising regarding mid-term and long-term guidance. Management is targeting ¥7.5B in sales by 2027, close to half of which is banking on the growth of their new businesses (grey bar).

Property Data Technos was started in 2022 to provide BPO and document management services to their clients. Growth in this business has been muted, slower than what management has expected.

Property Data Science, also started in 2022, arose to provide products to their retail customer base. ‘@commerce’ was the subsidiary’s first major product, released in 2023. Now they’ve expanded into customer support, releasing Speed Answer, a software product for lead and call management.

“Other” includes newer modules such as ‘@knowledge’ and ‘@iwms’ that expand usage outside the ‘@property’ functions though they are highly interconnected, as well as acquisitions.

I sincerely doubt Property Databank will be able to achieve their mid-term guidance.

We already saw historically how project delays in the service business, accounted for under percentage-to-completion method, leads to severely volatile earnings recognition, which leads to either exceeding expectations or hugely missing them, as they did in 2023; projects which are becoming larger and larger, which can be at bigger risk of schedule delays.

Nevertheless, many software companies have successfully doubled or tripled sales simply by cross-selling new businesses to existing clients (such as Appfolio). Ergo, I am not discounting this completely.

The investor may find be turned off by this dynamic, in a Japanese market that is hyper-focused on meeting projections. However, the earnings power and growth runway of the core business is so large I do not think of this as long-term risk. Additionally, if the market was confident management would hit ¥7.5B in sales by 2026, then the stock would be trading at close to 1x forward revenue, which is downright absurd. I think the market is already cognizant of the fact that the mid-term guidance figures is a nothing-burger.

The volatility of service profit margins (btw, there is no breakdown of the segments, which has been a black box for investors, though management has confirmed the service business has lower margins than the cloud business) leads to:

1. More buying opportunities when Mr. Market unfairly discounts Property Databank simply because contract performance is delayed.

2. Overtime, the consistency and growth consistency of the cloud business will lead to a generally higher % of operating profit coming from said segment, so it will have far higher significance to the market, diminishing stock movement due to short-term uncertainty.

Again, we have a case of a focus on short and mid-term uncertainty over long-term earnings quality.

—

Valuation

At 6x EBITDA, 9x FCF, <3x sales (2024 figures), PDB is a great business at a fair price.

You can say the pandemic was a major catalyst for PDB. In 2017-2018, margins were in the teens level, service sales jumped to >40% of the total while the cloud segment was growth 5-7%. During and after the pandemic however, sales were growing above 8%, with a near 20% spike in 2020 alone while service sales continued to grow. Furthermore, margins significantly improved: operating margins jumped to 23% in 2020 and 29% in 2021, EBITDA margins reached close to 40% by 2022.

According to management, profit margin increases have been mainly driven by operating leverage as the growth in cloud sales cost relatively little increase in SG&A. Again, I’ll mention that the company has no debt and ROE has maintained 17% since the pandemic with the exception of 2023.

Overall, a teens profit growth and near 10% revenue growth in the mid-term is a reasonable assumption.

The service segment definitely warrants a lower multiple, probably 6-8x operating profit after taxes. I would have designated it lower due to the poor margins and unpredictability of earnings, but the service segment is clearly growing with bigger and bigger projects over the years. In fact, service sales have grown >20% since 2017.

On the other hand, I think the core cloud business definitely warrants a mid-high teens multiple on earnings after taxes considering a potentially gigantic terminal value coming from consistent and predictable earnings growth and immense runway — I think 16-18x makes sense.

If you assume cloud solutions make up around 80% of earnings, then 15x is fair if you assume only little contribution from their new modules and segments. Though at 9x FCF and 2.6x revenue, this is generally cheap for SaaS (remember that Appfolio trades at 40x earnings).

Assuming a stagnant multiple, stable business, and new segments fail, I’m comfortable in estimating an IRR of 15%. If indeed the cross-sold modules significantly increases ARPU and switching cost however, this would lead to not just increase profit growth but a valuation re-rate, at which point a >30% IRR can be entertained.

—

The conclusion is that Property Databank is a classic near-monopoly vertical software business. If you think that Property Databank will remain the dominant property and asset cloud solution over the long term, the sheer pricing power combined with operating leverage will no doubt reward the patient investor with immense terminal value.

—

If you’re interested, I’d love to see you on Twitter. Every like, follow, and share is extremely appreciated.

If you’d like to learn more about vertical software, check out my other post on this.

Thank you.

Subscribe now

—

Disclaimer: The following write-up is no investment advice. The author may own, buy, and sell securities mentioned in this post. Please do your own due diligence.

Property Databank ($4389), a vertical software business whose primary product, “@property”, is the leading SaaS player in the Japanese property management space targeted to large investors and facility managers.

As of writing, Databank has a market cap of ¥9.5 billion, around $66 million USD. It’s a company that’s been modestly written up in the investing space — Made in Japan and small cap investor Teddy Okuyama have done excellent research on this company. Nevertheless, I think Databank is one of the lesser known software plays in Japan.

But to me, this is more than just an illiquid microcap. This is a unique opportunity to own a local monopoly with extremely high switching costs at a very cheap 6x trailing EBITDA (15x P/E).

And it just so happens that over the past few months, I’ve been trying to build my own property management SaaS business, so I might know a thing or two about this space.

Why property management? After studying various VMS businesses, I have come to the conclusion that property management software might have some of the best inherent qualities out of any vertical.

Property management/real estate has everything there is to like about a VMS vertical:

1. Real estate cash flows tend to be stable, even during recessions. This keeps real estate businesses afloat, even small ones.

2. Real estate, especially when it comes to leases, maintenance, and fund management, tends to be highly regulated, making it more difficult for users to trust other systems if they already trust their current product

3. There are a variety of unique business activities including maintenance logging, utility management, rent collection/payments, lease generation, show hosting, trust accounting, customer support, reporting, and eviction. All these are typically entirely handled by the property manager. Developers (that don’t outsource management) have more things to worry about, including appraisals, construction contracting, pre-sale, etc.

4. Property managers have integrated touchpoints with stakeholders that can each have their own portals and transaction gateways for more efficient business activity. This mainly includes tenants, but can also include vendors and maintenance contractors.

Appfolio and Yardi are both >$7 billion companies that, despite commanding less than 25% of market share respectively, have exhibited strong returns on capital from ever-increasing ARPU and astoundingly low churn, further bolstered by a growing demand for AI add-ons and features.

I believe Property Databank has the potential to follow a similar path to these great property management SaaS businesses.

Databank is further blessed with excellent macro tailwinds as the digitalization push has sparked a wave of cloud software adoption among Japanese firms. This is no doubt also true for in real estate.

Fears concerning Japan’s declining population can also be mitigated: as the population declines, more people will be concentrated in the metropolises like Tokyo and Osaka. This means that the decline in real estate cash flow coming from population decline will be partially offset from a greater concentration of people in the largest cities, where “@property”s largest customers such as J-REITs and infrastructure tend to be present in.

Japan has an uniquely regulated property management regime. There is a ton to go over, so I will summarize:

• Japan is very document heavy and highly formalized, and most of these documents are paper-based. Prospective buyers or renters must typically provide a certificate of residence (juminhyo), a registered seal (jitsuin) along with a seal registration certificate (inkan shomei), a valid passport and/or residence card, and proof of income or employment. Seals must be registered and certified, adding an extra layer of security. Property managers are also expected to provide regular, detailed reports to property owners, including tax documentation, financial statements, and compliance records

• Compared to North America, Japan’s regulation is more uniform on a nation-wide basis. Documentation and requirements tend to be more standardized, which means SaaS systems can more easily transfer across regions. As regulation is centralized (whereas North America tends to be decentralized), the federal body typically has final say on a lot of real-estate issues, which arguably makes compliance rules stricter.

• Maintenance is subject to strict codes to ensure facilities have good structural integrity, fire safety, and meet equipment standards. Managers need to procure various confirmation, inspection, and equipment certifications for their properties. Certain buildings (e.g., large apartments, offices, hotels) are required to have regular safety inspections by licensed professionals (kenchikushi) and must be logged and reported to local authorities. Intervals range from six months to three years, depending on building type and use.

• For both Japan and North America, escrow accounts for managers who manage other people’s properties are required to separate their clients assets from their own. It seems that managers in Japan are subject to stricter disclosure requirements including explanations of master lease agreements and risks, and detailed written documentation for all significant transactions in their reports. Ergo, transaction ledgers must be recorded perfectly.

Overall, the property management regulation in Japan is characterized by greater documentation, formality, and centralized governing bodies as opposed to more dynamic documentation rules and local/municipal regulators. This not only makes the reporting and documentation side of property SaaS more important, but also increases the level of trust a customer has with their system, greatly raising the costs of switching to a platform they don’t trust.

—

Property Databank (PDB) was established in 2000 by Sadahisa Takeno. The company was created through an intra-corporate entrepreneur contest Takeno won organized by his previous employer, Shimizu Corporation, one of Japan’s leading construction and real estate firms. At the time, there was a growing need for continously adapting property management IT— this was especially apparent when asset-impairment accounting was introduced in 2006.

Shortly after its founding, PDB launched its flagship real estate cloud service, “@property,” leveraging Shimizu Corporation, which invested in PDB, to access large property developers and REITs. The company went public in 2018. By the pandemic, PDB had established 3 data centers and had expanded to fixed asset management, sales support, renovation management, and more.

Today, despite having less than a thousand customers, PDB leads the enterprise property management space, controlling >53% of J-REITs (REITs that are publicly listed) and >38% of REITs in Japan as well as majority of railroad networks. PDB facilitates (likely) hundreds of thousands of buildings.

Still, >51% of the real estate space use excel and 29% use in-house systems. There is plenty of room to acquire more customers.

—

Evidence of pricing power

PDB has consistently driven up ARPU from a mix of price increases and cross-selling new features. Monthly average subscription fees have grown at a CAGR of 8% over the past 10 years!

Moreover, over the past decade, annual churn has remained below 5%, implying an average contract length of 20 years.

As PDB operates in the enterprise space, organic customer acquisition is slow. However, they have still been able to grow the cloud business 8-10% annually, largely driven by pricing power.

This pricing power comes from sheer switching cost, which I break down into the following factors:

1. Mission criticality: PDB’s customers rely completely on their products for day-to-day operations of their business

2. Data embeddedness: the amount of data required to manage a portfolio plus the strict reporting requirements makes for an extremely entrenched customer base

3. Learning curve: the sheer extensiveness of features and high level of trust required immensely raises learning curve.

Let’s dive in:

Mission Criticality:

What would happen to a client if PDB was shut down for a week?

Transactions would grind to a halt. Payments would go unrecorded, un-invoiced, and unaccounted. All operational data will disappear, preventing firms from filing required reports to regulators and making decisions. Financials or accounting metrics are gone, which would lead to a breakdown of internal controls. Maintenance teams will lose access to their logs and schedules, slowing down repairs and building issues.

I would argue the most mission critical parts of property and asset management in Japan is maintenance. Japan has strict maintenance regulation that requires properties and assets to be in tip-top shape regarding structural integrity, fire safety, and equipment readiness. Various certifications both during construction and after are required to continue operations. Moreover, regular inspections are mandated, particularly for large buildings and especially for infrastructure assets like railroads.

PDB’s customers manage huge portfolios and assets that impact the lives of millions. PDB’s products are absolutely necessary for these firms to not just continue day-to-day operations but also to abide by regulation.

Data Embeddedness:

In property management, relevant data includes various logs including transaction history and maintenance history. This data is important not only for internal business purposes but also for external regulatory purposes. Documentation is a big deal as Japan’s property regulation is document-heavy and highly formalized, a lot of which is still paper-based.

Business integration is extremely high: payments/rent collection are connected to lease management which is then connected AR/AP accounting which then flows into reporting and portfolio analysis. Data for large portfolios have get real heavy, often extending 10+ years (again needed for both business and regulatory purposes).

To switch to another system would be a massive operation! Not only must you re-upload tens of thousands of documents and logs, you must also re-input all the tenant, maintenance, receivable, payment, and accounting data into the new system.

Furthermore, you also have to get your employees, vendors, and tenants connected on PDB to switch off. This would require further coordination, time, and communication — prone to significant operational risks and potential for worse stakeholder relations.

Learning Curve:

While “@property”a core product is in accounting and general property management, PDB has since expanded its offerings to include a variety of other property-related functions. This has led to far more expansive feature richness and variety compared to competitors, maximizing product entrenchment and learning curve.

Feature roll-out outside core property management and accounting has primarily been focused on asset management. This makes sense as large REITs and fixed asset managers would have a greater need for maintenance, reporting, construction, and contract management. Regular logged maintenance reports, lots of paperwork on property permits, and strict building requirements have led to a greater need for physical asset maintenance documentation tools compared to PDB’s North American counterparts.

Here are several asset management features that PDB has released in recent years:

A lot of construction-related modules such as a 3D building modelling modules and an earthquake simulation tool in 2021 in collaboration with Shimizu.

Several other new full standalone yet integrated products have been released expanding on their simpler counterparts in ‘@property’:

‘@knowledge’ is a standalone document management application where users can handle document management in compliance with the Electronic Bookkeeping Act and e-Document Act.

‘@commerce’ is targeted for managers handling retail stores which provides a variety of analysis tools including forecasting, competitor analysis, and trade area analysis. Firms can find foot traffic data on certain streets, profile consumers and demographics, and get support on where to place new stores.

‘@iwms’ helps managers plan spaces. Features include drag-and-drop floor plan tools to optimize office layouts, manage desk booking (hoteling, assigned seating), and maximize space efficiency. Clients can track workspace usage, identify underutilized areas, set maintenance scheduling, handle facility reservations, handle check-ins/visitor logging, and more.

‘@cmms’ expands on the construction module. Clients can not only facilitate project budgeting, cost tracking, reporting, document management, construction life cycles/scheduling, and accounting.

PDB has also been trying to utilize their massive amounts of data to provide appraisal, valuation, and analytics features.

This would be a significant “next step” in extending the learning curve of PDB for its clients.

Why the opportunity exists?

I usually talk about why the opportunity exists at the beginning. But in this case, I wanted to go over the quality of the cloud business before I jump into this so you can easily see what the market doesn’t understand about PDB.

>25% of PDB’s sales come from installation, setup, and consulting fees. While the typical SaaS company will have no more than a tenth of sales in this bucket, PDB has been a unique case due to the behemoth operational requirements of their enterprise customers. New customers take risk being a customer of PDB. One mistake, and shit ensues. As such, PDB offers a ton of services to help clients integrate their existing systems and operations onto PDB. Major asset and construction projects may require additional help and unique customization from PDB to ensure SaaS systems run smoothly.

As PDB has secured larger and larger customers, they’ve also been dealing with larger and larger projects. This has resulted in a ton of service revenue — and while more revenue is good, there is an issue: 1) the service business has inherently lower margins and more importantly, 2) the service business is extremely inconsistent.

The big issue with the service business is that due to the longer performance obligations, revenue recognition can get delayed, and with it the realization of its gross profits. This was especially severe in 2023 when longer-than-expected contract fulfillment in the service business led to a sharp decline in profits. While this should balance out as the profits are recognized in the following year, it has more often than not led to not only unpredictable earnings which distorts the quality of its core cloud business but also missed guidance and analyst expectations.

Azeus Systems had a story that rhymed, and I talk about it in my 2024 year end post, but the gist is that oftentimes a great business can be covered by a mediocre one. Financials are volatile, the stock does not screen well. It is for this reason that I find PDB to be undervalued — the high quality SaaS business is buried beneath inconsistent earnings masked by a mediocre service business.

—

ARPU Uplift Opportunity

Some analysts have expressed discontent with PDB’s top-line growth — pointing to the fact that PDB has grown the cloud business revenue at below double digit for the past several years and that customer acquisition has been slow.

Here’s why this does not matter to me:

1. 90% of intrinsic value is from the terminal value. The growth rate over the next 4 years matters far less than how much PDB can sustain growth into the far future. The fact is, while PDB’s software business has not grown fast, it’s grown consistent. There has been no year where they’ve saw a decline in sales.

2. There are plenty of successful compounders that despite not growing sales volume, have done excellently from growing price higher than inflation. It costs marketing spend to increase revenue from getting a new customer, but it costs nothing to increase revenue from raising prices when practical. (See’s Candies is a classic case study. In terms of volume, See’s failed to grow with California’s GDP, yet they’ve consistently priced up their products due to their competitive advantage of brand image).

While top-line has been decent, PDB has increased operating profits >20% for close to the past decade. Pricing power truly is enough to provide excellent returns on capital and long-term earnings growth.

Not convinced? Let’s look at an example.

Appfolio has been described, even by Mark Leonard, as a truly excellent vertical software business. The stock, since IPO, has compounded at an annual rate of 32%!

How has Appfolio done this?

Aside from the usual property management playbook, where Appfolio has done particularly well in the accounting-side of their product, Appfolio has grown significantly from cross-selling a multitude of value added features, consistently raising ARPU.

These services include tenant screening, risk mitigation/insurance services, and most importantly, e-cheque transaction processing fees, which is estimated to contribute the most to the Value added services.

The above is from Appfolio’s 2024 annual report. Value Added services are now around 3/4ths of Appfolio’s revenue, over 2x what they make from their core subscription sales.

While core solutions have grown at a near 20% CAGR since 2021, its Value Added Services has more than tripled! This primarily came in 2023 when Appfolio started charging $2.49 per transaction for their eCheck feature which allows tenants to send cheques electronically to collectors.

What baffled me when I came across Appfolio is that not only did they introduce a huge change in pricing structure, their competitors did not match:

Competitor Yardi has been charging $0.95 per transaction before 2023, yet has in fact since removed the fee in 2024. Buildium, targeting smaller clients, has also charged cheaper at $0.5 per transaction even prior to 2023. The layman would think that this significant pricing difference would cause a mass exodus of clients…but lo and behold, Appfolio saw a 13% increase in managed units!

While there was pushback, the switching cost was strong enough to gain more from the ARPU uplift than from the lost customers. This is TRUE evidence of the pricing power.

Currently, ‘@property’ allows tenants to connect their bank account and automates statement to invoice reconciliation, standard for most property systems. They’ve yet to levy any additional revenue-gaining fees (aside form payment gateway fees of course). While there is no guarantee they will ever do this, the point of the matter is, the property software is extremely sticky, ergo the module cross-sell and potential monetization of their clients stakeholders, e.g. the tenants in Appfolio’s case, create endless opportunities for ARPU uplift in the future.

—

In the next post, we will go over PDB’s competition, valuation, and management.

Stay tuned and thanks for reading!

Thanks and Merry Christmas!

—

This is part 2 of the deep dive on Plus Alpha Consulting. If you haven’t yet, please read our general write-up on the company — Plus Alpha Consulting - Part 1.

As we’ve discussed before, Plus Alpha is not just a SaaS company, they are at heart a text mining company. To reiterate, text mining is an application of machine learning that enables the analysis and interpretation of any text or language input through big data models. Plus Alpha is in the process of making the transition from having a machine-learning focus to an AI focus. We believe because Plus Alpha has been a text mining company since the beginning, the transition will be a lot easier and a lot faster than most average SaaS companies.

Additionally, while our analysis is mostly centered on Talent Palette, we don’t want to ignore the potential future free cash flow that could come from Plus Alpha’s other products.

Let’s start by turning our attention to Plus Alpha’s other existing offerings: the marketing solutions.

Mieruka-Engine allows firms to use text mining to analyze and visualize customer feedback from direct forms, call logs, or social media. The tool can scan text and voice data, pinpoint key words, interpret feedback results, and more.

Mieruka-Engine is more of a tool than it is an enterprise software. It lacks the feature richness, data integration, and mission criticality of Talent Palette. Customers mainly use Mieruka-Engine as an add-on to their existing CRM systems.

Mieruka-Engine is unique in which there are very few similar products that do text-mining for customer analysis. It seems like a good product overall from the feedback we hear, but customers are definitely a lot more price sensitive.

The second marketing SaaS product, Customer Rings, allows firms to analyze and visualize customer purchase histories, web access logs, etc. They also serve as a CRM system where users can manage email campaigns, SMS bots, feedback forms, and dynamic website graphics.

We like Customer Rings more than Mieruka-Engine, mainly because getting on the system requires integration with all customer and marketing data as opposed to just feedback results and log data. In fact, it takes about 1.5-2 months to set up for large enterprises in order to log in all the data. There is also a moderately large variety of functions, with most of the value proposition coming from the data analysis and visualization tools. Better switching costs, but the CRM space is by far the most crowded vertical.

Overall, the marketing solutions have ok economics. There are 870 customers, mostly mid-size and large enterprises. Customer base has grown just 5% YoY for the past several years. The product is generally on the expensive side, ARPU stands at ¥355k, increasing 4-5% YoY. MRR churn rate is pretty high, averaging 0.9-1%, over triple that of Talent Palette. Most churn comes from cancellation due to price increases, which tells me the marketing solutions are generally less mission-critical and face tougher competition.

Mieruka-Engine YoY growth has stayed at 4-5% for the past 4 years, Customer Rings on the other hand has grown >10% per annum over the same period. This is in part because Customer Rings has an established distributor network compared to Mieruka. However, Customer Rings has worse profitability with an operating margin of about ~30% while Mieruka has averaged >50% because Mieruka is a simpler product and doesn’t require as much infrastructure or customer support.

In summary, a highly profitable business, but with mediocre expected growth for a SaaS business.

—

While not important to the thesis, It’s also worth talking about something we believe could range from a slight negative to very large positive:

Mimuro has been launching multiple products related to text-mining since he started the company. Most of them were in the marketing space, and most did not stick. To this point, Mimuro has not stopped, and with valuable experience from his past failures and success of Talent Palette, Mimuro has been up to some very interesting new projects. While most will fail, even one success could lead to a homerun……

Mimuro launched Yorisoar in 2022. Yorisoar is a learning management solution (“LMS”) for schools and universities that helps with managing teachers, classes, and students, with a variety of features from staff attendance and grading tools to student performance analysis with Plus Alpha’s signature big data analysis tools that can help teachers understand what students need to improve, help them get jobs or get into their desired university, or prevent them from dropping out.

Like Talent Palette, Yorisoar is a mission-critical system that integrates with the entire organization of the user. Teachers, students, and administrators require the use of Yorisoar for everyday school activities.

There are some synergies Yorisoar can provide to Talent Palette which Mimuro has pointed to. Utilizing the student data from Yorisoar and the employee data from Talent Palette can help better match employees with employers. In late 2022, Plus Alpha acquired Grow Up, a ¥600M revenue HR SaaS company whose main product is Kimisuka, a tool that helps match companies with job applicants. This will provide the infrastructure to eventually bridge the gap between Yorisoar and Talent Palette (Mimuro acquired Grow up for about ~8x 2024 operating profits. Compare that to Kaonavi’s acquisition of Work Style Tech for a high single digit sales multiple. Both companies were unprofitable at acquisition with the same M&A motivation to incorporate the acquiree’s features).

Yorisoar was used by 10 schools in FY2023, expanding to 70 schools this year with an ARPU of ¥170k. Plus Alpha is targeting to expand their customer base by 40-50 new schools a year. Yorisoar currently generates ¥142M in annual revenue with a -¥166M operating profit.

Among universities, the LMS space is already very crowded. LMS software was widely adopted during the pandemic which forced schools to offer classes online. >70% of the roughly 800 universities in Japan are already using LMS software. Moodle and Manaba are the largest players which together likely command >50% market share.

The main opportunity is within high schools and vocational schools which likely have lower ARPU but are still mostly unpenetrated.

Don’t get your hopes too high though, a declining population means less and less kids in school. Nevertheless, the roughly 15,000 high schools and 2500 vocational schools in Japan makes for a pretty wide addressable market. And it seems that Yorisoar has generally performed better than management expected.

—

Also in 2022, Plus Alpha released Sales Square. Sales Square is essentially a combination of Talent Palette and Customer Rings — a HRM and sales management solution for sales and support teams. Sales Square allows companies to manage sales skill development, manage and track KPIs, manage leads and cases, build forecasts, and analyze and visualize marketing data. It’s a very small operation and there’s not much disclosure on Sales Square so far.

Aside from these two, Plus Alpha also just soft-launched a cloud hospital management system called HiCare Wellness which utilizes text-mining and GenAI to help improve patient experience and manage doctors/nurses/staff.

Only time will tell how these new ventures add contribute to the consolidated business. But it’s a positive to see this failure-is-learning mindset that we feel is kinda rare for a Japanese company.

Mimuro: the man behind the curtains

Ex-employees describe Mimuro as quiet and analytical, and the work culture has been described as “venture-like”. Rules are looser. Unlike most traditional Japanese companies (often referred to as black companies), the work hierarchy is quite decentralized. Employees at Plus Alpha are given significant amounts of autonomy to come up with their own ideas and suggestions. When employees come up with a new idea, they are given 2 months to build and implement — people are expected to fail fast. This kind of work culture is imperative in a business where innovation is key.

Though management are only paid with a salary, we think there’s sufficient skin in the game as insiders own 40% of Plus Alpha. Mimuro himself owns 21%. VP Kenji Suzumura, who was at Plus Alpha since the beginning and was best friends with Mimuro while they were at Nomura, owns 16%.

In some aspects, capital allocation is similar to most Japanese firms — fat cash balance, no debt, dish out some dividends. Japanese businesses have been notorious for being very risk-averse when it comes to capital allocation.

Plus Alpha hasn’t really done much with cash flow until the past couple years when it started making several acquisitions. Aside from Kimisuka, Plus Alpha made 3 acquisitions this year:

1. Attack Inc. is a recruitment service and consulting business. They were acquired for ¥80M at 0.47x sales (unprofitable). Plus Alpha wanted Attack to expand their consulting team, acquire their customers, and increase expertise in recruitment.

2. D4DR is a business consulting and IT transformation business. They were acquired for ¥144M at 11x after-tax earnings. Like Attack, Plus Alpha’s motivation is to use D4DR to provide customers, talent, and use their data to improve their products.

3. OMN was Plus Alpha’s largest acquisition for the year. OMN’s main product is R-Shift, a service that uses optimization to match the number of employees required for each store/branch with the available working days and hours, skills, and experience of the employees. They also provide shift management, payroll management, and budget management. Plus Alpha wants to incorporate OMN’s software into Talent Palette. OMN was acquired for ¥1.6 billion at a 6.6x operating profit multiple.

Acquisitions are mostly product and expertise led. We think management is pretty disciplined as all the acquisitions so far have been made at reasonable or cheap multiples, though they’ve used some stock issuance to fund the purchases resulting in a 6% dilution since IPO.

Figure: changes in 2025 projections

So far, Kimisuka has been a bit of a hiccup, generating only ¥177M vs their projection of ¥300M in operating profits. Further M&A-related costs, unprofitability in new product launches, and more aggressive discounting for Talent Palette have resulted in 2025 guidance revised down.

But the market is throwing the baby out with the bathwater. Sure there will be hiccups, just like any company, but 5 to 10 years from now, none of this will matter. All we know is Talent Palette will continue to dominate the TMS space. Management seems to agree as they’ve recently announced a share buyback of nearly 6% of shares outstanding in response to the sell-off — a positive sign that management recognizes the stock is undervalued and has consequently taken action to add value for shareholders.

Valuation

Plus Alpha is valued at ~11x 2025e ev/ebitda, likely ~10x when counting the share buybacks. (Price/2025e FCF of ~19x, ltm ev/ebitda of ~14x).

This is a historic low compared to the >20x ntm ev/ebitda Plus Alpha used to trade at over the past few years. Moreover, Kaonavi trades at close to ~25x forward ebitda (though they trade at a much lower revenue multiple due to lower profitability). Other profitable ERP software companies like OBC also trade above 25x.

Even when you mix in the lower quality marketing segment which rightly should have a lower multiple, Plus Alpha still seems quite cheap. Furthermore, as Talent Palette outgrows the marketing business and other solutions, overall profitability will likely go up.

If this company was in America, we wouldn’t doubt this could trade at above 20x ntm ebitda easy, as has been the case with Salesforce and Workday since IPO, and remember that these firms have much larger dilution from share-based compensation.

In summary, we think 11x forward or 16x ltm ebitda is a substantial discount for a wonderful software business with high switching costs + a customer base that absolutely adores the product.

If Talent Palette can grow customer base at 15-17% with flat ARPU growth and zero margin uplift, they would end up with ¥8 billion ($51 million USD) in just 4 years. As Talent Palette outgrows the Marketing solutions segment and assuming everything else grows very modestly, we estimate the whole business can easily grow ebitda by at least ~18% for the next several years with an operating profit of ~35-37% — about ¥8.8 billion in EBITDA; slapping on the same growth rate on the cash pile, we end up with ¥19 billion in cash.

Applying a 20x EBITDA multiple gives us an IRR to our estimated value of >25%. If Plus Alpha ends up with the same multiple it has now, you still get an IRR of ~18%!

Conclusion

Plus Alpha Consulting through Talent Palette provides an opportunity to invest in a fast growing talent management enterprise software with extremely high switching costs and robust unit economics all in a highly unpenetrated space facing ever increasing demand for software. And because of Mimuro’s continuous commitment to developing text mining and gen-AI vertical software products, new growth opportunities could be near endless!

We initiated research on Plus Alpha in early August but thought the stock was too expensive. Since then, the stock has come down to more reasonable levels, so thought now would be a good time to share.

—

Disclaimer: The following write-up is no investment advice. The author may own, buy, and sell securities mentioned in this post. Please do your own due diligence.

The main idea: Plus Alpha Consulting owns Talent Palette, a talent management software for large enterprises. Talent Palette is an excellent business for the following reasons:

1. Talent Palette holds the no 1. position for large enterprises, exhibiting extremely high switching costs.

2. Talent Palette has the best features and capabilities in the market by far. It is also the best-positioned to implement AI.

3. Talent Palette has the best customer acquisition strategy and cost structure.

Pre-ramble:

As Japan’s birth rate and population continue to decline, there has been a massive push to increase labor productivity.

Japan continues to have one of the worst labor productivity in the world for a developed country.

It’s very well known that Japan has a very tough work culture. Employees work extremely long hours and have extremely rigid hierarchies.

Employees are expected to stay in the office until their boss leaves, even if they don’t do anything. Turnover and workplace competition is extremely low; once someone is hired, they are typically employees for life.

Downsizing is often looked down upon. In fact, many companies are known to have “useless departments” that simply do not provide much value to the company yet are kept on.

Aside from that, employees are often overworked — loneliness is at an all-time high leading to tired and unhappy workers.

While younger Japanese are seeing changes, there has been urgent demand for software which improve workflow efficiency and labor productivity.

The software space in Japan has been rocky:

Japan’s industry was extremely successful during the 80s and at the time was projected to surpass the US. However, Japan did not embrace the rise of software and its implications on the rest of the economy was drastic.

The IT business was controlled by large oligopolistic system integrators. These firms handled outsourcing and development of IT systems and software as well as provided consulting and advisory services. While these firms did offer software, they didn’t have the speed of iteration or the culture of a startup to innovate against foreign peers.

As Tim Romero from the Disrupting Japan podcast put it:

“Software development was an exercise in box checking. You implemented a feature once the customer had asked for it and the contracts had been signed.”

Because Japan depended so much on system integrators, it stifled competition and investment in startups. Even today 75% of IT talent is still concentrated on System Integrators in Japan vs 35% in the US.

The fail-quick-and-learn-fast model adopted by the successful Silicon Valley startups never took off in Japan.

Software developers, even very talented software developers, were not encouraged to take risks and go against convention. Software developers were not respected like in North America, and were oftentimes considered a boring back-office low-skill career. They were not prized, and so there was not much of a career path they were encouraged to pursue.

By the time US software giants hit the shores of Japan, domestic software companies could not put up much of a resistance as these firms had vastly superior capabilities and technology that improved fast and continuously.

This was one factor that foreshadowed Japan’s lost decade in the 90s where domestic industry fell behind as IT and technology took over the world.

From the 2nd largest economy poised to take over the US, to one of the slowest growing as of recent, Japan has only now been able to recover from this…

The 2010s marked a revival of Japanese technology startups.

The process of software development became easier. The introduction of cloud computing, app development, more accessible programming languages, and better IT education and resources all helped to lower the barriers of entry.

Simultaneously, cloud software was quickly taking over on-premise software. Because cloud software can be accessed without heavy installation and maintenance requirements on your own systems, acquiring customers no longer required a lengthy sales cycle or tedious negotiation. As a result, the number of cloud applications exploded, many of them run by young visionaries.

As Japan’s workforce became more flexible and innovative, demand for talented software developers boomed.

The early 2010s marked the beginning of new software unicorns that could challenge legacy players.

Still, Japan has a lot of catching up to do. While on-premise solutions are on their way out, only 40-50% of software usage is on-cloud.

In the next several years, most on-premise software support will likely be shut down, which will force many firms large and small to adopt the cloud. Furthermore, companies will be forced to embrace AI, which is likely to be done by firms with more innovative culture rather than a system integrator run by older hierarchical teams.

The age of the startup has come.

This leads us to discuss a company that caught our eye.

—

Plus Alpha Consulting (TYO:4071) — market cap: ¥77B

Plus Alpha Consulting was founded in 2006 by Katsuya Mimuro as a text data mining and IT consulting business. Text data mining involves analyzing text and calls and using machine learning to interpret the data. Mimuro previously worked for Nomura Research where he was obsessed with developing text data analysis tools. Mimuro would build a variety of solutions with applications in demand forecasting, customer analysis, AI, and information systems. While Plus Alpha started as a consulting service, Mimuro would pivot to focus on enterprise software. For the next 18 years, Mimuro would release a variety of products involved with text mining in a variety of verticals.

Today, only three products contribute the vast majority of revenue: Mieruka-Engine, Customer Rings, and Talent Palette.

Talent Palette (and other HR solutions): 73% of revenue, 50% operating margins for Talent Palette, 40% operating margins for consolidated segment

Launched in 2016, Talent Palette is a talent management software targeted to mid-size and large enterprises. Talent Palette provides a software solution to a variety of HR activities including managing employee data, matching projects to employees, performance and KPI monitoring, recruitment, feedback and evaluation, professional development, rewards and benefits, data visualization, and more. Talent Palette has increased revenue Five-fold over the past 5 years (>50% CAGR), and we believe Talent Palette is the best poised to benefit from the Japan-wide drive to improve labor productivity.

Mieruka-Engine: 14% of revenue, 53% operating margins

Mieruka-Engine was one of Mimuro’s earliest products which Plus Alpha launched in 2008. Mieruka-Engine is a text mining software which analyzes and visualizes data from social media, call logs, surveys, and text. Mieruka-Engine is primarily used for improving marketing and services. Customers primarily consist of large enterprises which use the software for their contact centers and marketing departments. Mieruka-Engine has seen low-growth over the past 5 years (4% CAGR).

Customer Rings: 13% of revenue, 30% operating margins

Customer Rings was launched in 2011. Customer Rings is a CRM software which specializes in using text mining to analyze and interpret customer communication. Customer Rings is used mainly for digital marketing. Customers are mostly in retail, apparel, cosmetics, general goods, and more. Customer Rings has grown a bit faster, largely driven by price increases - a 10% 5-YR CAGR.

Note: consolidated Plus Alpha operating margins are 33% which includes company-wide corporate overhead

Why the opportunity exists?

This is a small cap high growth stock trading at ~11x 2025E ev/ebitda — a more than reasonable price considering the strength of their competitive position.

Plus Alpha is down over ~30% ytd. The last quarter results sent the stock down ~25% due to higher than expected costs coming from newly acquired firms and lower forecasted profit growth which has sent signals of management’s concern about the increasing competitiveness of the HR tech space.

But we are not thinking about where the company will be in 2 years, we think about where it will be in 10 years. And in 10 years, we believe Plus Alpha is most likely to come out on top.

Plus Alpha has no debt and a ¥10B cash balance.

We think Plus Alpha can grow sales at a 20% cagr for at least the next 4 years with potential multiple expansion.

Competitive landscape

The talent management space (“TMS”) can be divided into three segments: SME (<200 employees), mid-size (>200 employees), and large enterprise (>1000 employees).

The SME segment is dominated by the likes of SmartHR, Money Forward, and Freee. These companies offer very user-friendly features that’s simpler, less features, but with easier set-up. The space has been getting crowded with a variety of new entrants, typically single-product apps. Furthermore, talent management software isn’t as mission critical as managing a 50-person team doesn’t exactly require a whole system. Instead, the primary competition comes from accounting software giants like Money Forward which have been expanding their HR and payroll systems in addition to their current finance software suite.

The Mid-size segment has been the main battleground for TMS companies. Main players include HRBrain, Talent Palette, and the no. 1 in this space, Kaonavi, which is currently the largest enterprise TMS player in terms of customer size. Even though it’s “mid-size," 200 people is a lot — this is where they become an enterprise. As such, there is exceeding demand for managing employee data and evaluating performance.

Then we have the large enterprise segment. Talent Palette is the no 1 player, followed closely by Kaonavi, then foreign players like Workday. These customers comprise all of the keiretsu, trading companies, and well-known Japanese brands. With over a thousand employees, moving from Excel to a proper software system increases productivity many times over. Data visualization, performance flow, text analysis, and other tools become more useful.

The table below shows comparison between Talent Palette and Kaonavi’s customer breakdown.

Talent Palette (1798 customers, ¥712M MRR)

Kaonavi (3963 customers, ¥763M MRR)

Note: Kaonavi and Plus Alpha have different definitions of mid-size and SMB (Kaonavi defines mid-size as >200, Plus Alpha is >300), so there is going to be overlap.

While this helps us understand the customer base, both Kaonavi and Talent Palette reports a firm with multiple subsidiary firms as one customer. In reality, the no. of registered companies that use Talent Palette is >3400 and approx. 4000 for Kaonavi — the gap isn’t as large as it seems.

Also, we can categorize each customer segment like this, but there’s no hard boundaries. The higher up you go in size, the more the customer prefers features and capabilities, while the lower you go in size, the more customers prefer user friendliness and simplicity.

Talent Palette’s Competitive Advantage: Switching costs

The enterprise software business is all about switching costs. How difficult and/or costly is it for the customer to move over to your competitor?

Talent Palette’s switching costs vastly exceed its competitors due to 3 key factors:

1. Feature richness and pace of feature roll-out

2. UI/UX complexity which creates a high learning curve

3. Huge data onboarding which makes migration tedious

Talent Palette has thousands of features and counting, many of which customers will likely never fully use. In fact, Talent Palette has added a whopping 5962 features since it’s release! We estimate this is at least 4x that of peers.

Talent Palette’s functions are extremely vast, covering personnel evaluation, project matching, employee data management, job applications, professional development, benefits, and more. Feature complexity ranges from something as simple as digital paperwork to AI analysis of employee peer reviews.

Here are some examples:

Keyword maps - pinpoints common words/phrases from feedback form

Employee view - manage employee data

Data dashboard - shows multiple data sets including distribution of employee relocation preferences

The key differentiator of Talent Palette is the many different ways users can analyze their workforce (mostly based on feedback, survey, and other consentual employee data). Talent Palette offers dozens of options for visualizing data, far better than any other peer.

Because text mining involves complex data analysis, AI has been in Talent Palette’s DNA since launch, with features such as automatically generating profiles, recommending KPIs, and matching employees to different assignments. The more employees Talent Palette acquires, the more data it will be able to use to improve its AI modules.

Through their immense data bank, scientific rigor, and complexity of their text mining and AI models, customers can learn many things about their employees such as:

• if the employee is depressed, tired, or hates their job

• if the employee has performed better than peers

• which employees should be moved to a different department

• which employees are most likely to resign

• which employees should be promoted and which ones should be fired

Customers tell us Talent Palette helped them easily view identify depressed workers to give them mental help, understand the overall team morale and satisfaction, make better decisions on which employees to promote or fire, and more.

Talent Palette reduces the time needed to implement employee surveys, create presentations, do paperwork, and manage data.

As it takes time to get used to and understand Talent Palette, customers may start to require new HR hires to be trained with Talent Palette on the get-go. In fact, we’ve heard that knowing how to use Talent Palette is now becoming a requirement for getting HR jobs in the big firms.

Overall, these factors boil down to:

1. The sheer amount of functions, data input, and complexity creates a pretty steep learning curve. This undoubtedly reduces user-friendliness. However, this means customers are extremely entrenched in the product ecosystem.

2. The ROI of using Talent Palette is greater than peers due to the ever increasing capabilities of features and data applications. Customers find the product so useful that if they switch off they won’t be able to find anything comparable.

This makes Talent Palette have the highest switching costs in the space.

Let’s compare with Kaonavi:

Kaonavi was launched in 2012, being one of the earliest enterprise TMS software. Because of this, they quickly gained dominant market share early on.

Kaonavi has focused more on user friendliness than capabilities. This was confirmed when we spoke with Kaonavi’s CFO and ex-employees.

Kaonavi’s primary feature is being able to pull up employee faces. The product is called “Kao” (face), “navi” (navigation). Kaonavi has the core features one would expect of a TMS including employee data management, task management, evaluation forms, recruitment management, rewards, and benefits. Where they are lacking is the analysis functions such as textual analysis, GenAI recommendations, data visualization, personnel simulations, and performance analysis. They’ve recently introduced their own text mining module, but we hear it’s far too basic to compare to Talent Palette’s capabilities. Overall, Kaonavi has far simpler features, easily navigatable dashboard, and smaller no. of functions.

Kaonavi, like Talent Palette, charges different plans based on the number of employees and additional features requested. Kaonavi, when comparing on similar company size tier is cheaper than Talent Palette. Also, because Kaonavi has a larger mix of smaller customers, their average price consolidated is much lower as well. As such, Kaonavi’s monthly ARPU is lower than Talent Palette’s, ¥188k compared to Talent Palette’s ¥402k.

This explains why Kaonavi has become more popular among mid-size enterprises who may not have need for all the complex functions Talent Palette. But the drawback is smaller customers will always lead to higher churn than larger customers.

It’s much harder to change systems with 1000 people than it is to change with 200. TMS systems tend to store a ton of data. One employee alone could have 10+ years of data including projects they were involved in, key developments in their career, conversation and communication information, and more. The time it would take to migrate the data from 1000+ employees is astronomical! Because TMS systems involve each employee downloading their app and creating a separate account, migrating would also require each and every one to set up the software all over again. The larger the workforce, the more hassle it is to have them do that in an orderly fashion.

While Talent Palette has averaged an MRR churn of around 0.3%, Kaonavi has averaged a churn of >0.4%.

Talent Palette’s higher switching costs also gives them superior pricing power. On an annual basis for the past 5 years, ARPU has risen >12% while Kaonavi has risen 9%. This is a combination of upselling new features and price increases. (though we expect ARPU growth for Talent Palette to stay flat as competitive pressure has forced Plus Alpha and other firms to offer larger discounts, though price hikes on existing customers continue)

—

Kaonavi is in a structurally worse competitive position compared to Talent Palette:

The HR software space is getting very crowded, especially on the small customer end.

We believe the SMB TMS space is a no go — accounting giants like Money Forward and Freee have already replicated payroll and benefits, it won’t be long until they include check-in tools and data management, which are quite simple to build. Smart HR, the largest player in the SMB HCM space, is seeing competition on it’s core payroll tools and are trying to migrate to the TMS space by adding data management, feedback forms, and job application tools.

The mid-size segment is also facing increased competition. Kaonavi’s fiercest rival, HRBrain, has similar features with similar pricing.

They were acquired by Swedish PE firm EQT in 2023, a rare case of foreign private equity activity in Japan. EQT said HRBrain was the most user-friendly system with the best reviews. While HRBrain reportedly has about half the revenue of Kaonavi, we hear they are growing much faster in the lower end of the mid-size space.

We are also seeing a lot of SMB players trying to attack upscale segments. For example, Team Spirit, an attendance and time tracking system, historically focused on the SMB segment , has recently started targeting the mid-size enterprise space in an effort to boost margins.

The mid-size segment is huge — nearly 60,000 potential customers, >60% of which are still using Excel for HR functions or legacy systems. But all this competitive pressure forces price competition, which will hinder TMS players’s ability acquire customers.

The Plus Alpha Strategy: Customer acquisition structure

It’s tricky to compare customer acquisition costs because Plus Alpha doesn’t disclose sales personnel expenses.

Nevertheless, we can confirm that while Kaonavi has slightly lower revenue vs Talent Palette (¥8.6B vs ¥10.1B ltm), Kaonavi’s SG&A was 14% higher than Plus Alpha’s on a consolidated basis LTM — EBITDA margins was ~10% vs Plus Alpha’s ~35%!

This is not due to advertising: Plus Alpha has historically been more aggressive on advertising and trade shows and spent ~55% more LTM than Kaonavi. An ex-employee even told us during the late 2010s, you could not enter a taxi without seeing a Talent Palette ad.

The reason Kaonavi’s expenses were higher was due to personnel. Kaonavi has 2x more employees than Talent Palette and spends over ¥2 billion in personnel a year, mostly sales and product people. This is more than double what Plus Alpha spends on a consolidated basis.

Our guess is, again, Kaonavi is facing extreme competitiveness in the mid-market segment. More and more companies with similar functions and UI are offering better terms and extending the free trial period, putting pressure on firms to ramp up feature roll-out and sales.

Talent Palette is no doubt facing this too, but we think the sheer differentiation of the product and focus on the enterprise segment simply gives them a better cost structure which has allowed them to decouple ARR from SG&A whereas other HR SaaS cannot, all while growing faster than peers.

—

Talent Palette has another ace up it’s sleeve over Kaonavi: distribution network.

Kaonavi acquires almost 0 customers from sales partners or system integrators. Talent Palette on the other hand acquires about ~25-35% of it’s customers from distributors, which structurally leads to a better CAC. The lion’s share of distributor customer acquisitions come from consulting and IT service companies. The key drivers are NTT and Accenture. These system integrators will promote Talent Palette to their clients, which also tend to be mid-to-large enterprises. They take a commission, which is overall much cheaper than direct marketing. Talent Palette has been dedicated in adding more HR and IT service companies and training them to use Talent Palette.

Other key partners include large enterprises who themselves use Talent Palette,. Keiretsu like Mitsubishi have multiple relationships with other medium-to-large enterprises (suppliers, customers, etc.). Having a customer base of larger companies will generally lead to a larger aggregate number of companies in your network.

Plus Alpha’s marketing strategy involves a ton of consultants. About ~20% of the workforce are consultants. These consultants would go to a customer’s HR or executive team and provide advice for using the software and help customers learn about the various functions. They go the extra mile by initiating HR consulting projects with clients to help them better manage their employees and solve specific problems. This does the following:

1. Deepens Talent Palette’s relationships with existing clients which further drive down churn.

2. Provides an easy source of customer learning and finding ideas for new potential features.

3. Because customer service is so good, this raises the reputation of Talent Palette for both existing and prospective customers.

While it may seem overkill to put that much resources on customer service, dealing with large companies is a huge deal. In 2023, Mitsubishi Heavy Industries subscribed to Talent Palette, and not only implemented the software for it’s corporate workforce, but also implemented for all 8 of its domestic subsidiaries. The result was an acquisition of over 40,000 employees on Talent Palette! With organizations this large, having a strong consulting group to provide support is absolutely crucial for maintaining relationships.

In comparison, Kaonavi’s only major consulting partner is Recruit Holdings. Recruit owns about ~20% of Kaonavi. Because Recruit is one of Japan’s largest HR service companies, we initially thought Recruit was a major driver of customer acquisition, but our sources tell us Recruit does not contribute at all. Instead, the only reason Recruit invested in Kaonavi is to utilize their data to help improve their own services. Moreover, due to Kaonavi’s reliance on mid-size customers, they cannot get very good sales partners as mid-size enterprises will naturally have less connections or reach.

Nevertheless, management told us they want to have a stronger distributor network when TMS system popularity reaches an inflection point.

Overall, Talent Palette is evidently the better business. They already caught up to Kaonavi in revenue despite launching 4 years later. Indeed, ex-Kaonavi employees all tell us the scariest competitor is Talent Palette.

What about competition with Workday and SuccessFactor?

The core competencies are completely different where systems like Workday are used mostly for payroll and finance while Talent Palette is used for talent management.

Many of the keiretsu are using foreign HCM software like SuccessFactor, Taleo, or Workday, which also have talent management and visualization features:

In some ways, these systems are better than Talent Palette: they have better customizability and offer far more languages. But Talent Palette is obviously the clear winner in terms of employee analysis capabilities.

Actually, what usually happens is that most firms will use Talent Palette or Kaonavi for talent management on top of their payroll and benefits system. They might also have an international system for their head office but use Talent Palette or Kaonavi for the rest. For example, Panasonic uses Successfactor for the international headquarters in Japan but uses Talent Palette for all their local subsidiaries.

These established foreign software compete more with SmartHR Enterprise than they do with Talent Palette.

Growth runway

In Japan, there are about 58,400 mid-size enterprises, and 4,500 large enterprises.

Among large enterprises, we estimate about <40% penetration. Whereas among mid-size enterprises, we estimate about <10-15% penetration. Of course, the segmentation is more complicated than that: firms with >500 employees seems more likely to opt for Talent Palette than firms with less, so it’s not like Talent Palette will hit a wall once the 4,500 are filled. Overall, it seems like the penetration is low enough that TMS systems can grow for quite a long time.

What about the declining Japanese population?

It’s a valid point. Japan may lose a 1/3rd of their population in 50 years. Since Talent Palette charges per employee, this will no doubt bring some client's to lower pricing tiers overtime. But this also means there will be more labor shortages and thus more demand for labor productivity tools, which in our opinion should lead to a stable ARPU without accounting for further module cross-selling.

As seen from the graph above, the labor shortage continues to worse, motivating more companies to find better ways to attract and retain top-tier talent. If anything, the declining population will actually be a tailwind for software, possibly for more than the next decade.

What about concerns regarding Labor DX Lite?

One of the main criticism of bears is that Talent Palette is going to run out of growth when they reach saturation in the large enterprise segment. Talent Palette has been trying to enter the SMB space with the near-term release of Labor DX Lite, a simpler version of Talent Palette targeted to firms with <300 employees with a focus on payroll and benefits. Because this will naturally bring ARPU down, investors have been worried that this “land grab” is seen as a signal that Talent Palette is struggling to grow.

But…history tells us the story usually plays out differently.

Cloud software companies like Workday and Salesforce have seen their profits surge amidst decelerating ARR growth. Why? Because with less customers to acquire, the focus shifts towards cost optimization. Direct marketing is less important while distribution gets all the focus. New modules surrounding the current product continued to drive up ARPU.

Some examples:

Salesforce, which was an enterprise-focused business, also released an SMB version in 2017, right at the time their sales growth started to decelerate. They did not displace HubSpot, but it did not affect the fundamentals of the enterprise business. Although Labor DX Lite may not be the best move, we do not think this is a risk.

But Talent Palette is already very profitable, how much more profitable can it go?

OBC, a great case study for Japanese enterprise software, reached consolidated EBITDA margins of >45% despite relatively slow growth (8% 10-YR CAGR) through a combination of its dominant market position and strong network of 3000+ distributors. And remember, their software targets SMBs with a bit over half their revenue still in on-premise software. While it would be difficult for Talent Palette to achieve that distributor network, we think a stronger shift to a distribution model in the future could still be very accretive to margins.

It might also be worth noting that most new customers will often use Talent Palette for a small portion of their workforce before doing a company-wide implementation. We think most customers using Talent Palette for <1 year are doing this, which implies there is going to be TAM in customers eventually expanding Talent Palette’s coverage to the entire workforce, thus increasing ARPU and margins.

—

We expect CAC for all firms to go up over time as the space gets more competitive, though we think there is enough space for multiple firms to cement themselves — this won’t be winner takes all.

For these reasons, though Kaonavi is in a worsening position, they still have a very bright future ahead. Kaonavi’s efforts to replicate some of Talent Palette’s features like text mining and its larger focus on enterprise customers is a good signal despite decelerating YoY sales growth. In fact, if Kaonavi is able to build up a distribution network and be more disciplined with costs, they might be able to drive up margins closer to industry average, which would likely cause a multiple re-rate from its 2x price-to-sales.

Both Talent Palette and Kaonavi have a long growth runway while being quite cheap for an enterprise SaaS company.

—

In the next post, we will be covering other software products as well as our thoughts on management.

Thanks for reading!

Started just in 2023, Michele & the team have been acquiring small HVAC install and maintenance service providers all across Italy. 

Timeline:

Acquisition of SMIT Srl in Feb 2023 → 23k EBITDA plumbing, boiler, and heater maintenance firm operating in Piacenza and Parma. Lindbergh has used SMIT as a platform company to use in further acquisitions. The initial 52% stake has now been raised to 100% after the original owners decided to cash out.

Acquisition of Gatti Ermanno Srl in Jun 2023 → About 70k EBITDA plumbing and heating maintenance firm operating throughout Italy. 100% stake.

Acquisition of Climan Srl and Idro Calor Srl in Sept 2023 → About 135k combined EBITDA air conditioning maintenance businesses based in Parma. 55% stake.

Acquisition of Alberti Srl and RCR Impianti Tecnologici Srl in Mar 2024 → Two heating, air conditioning, electrical, and plumbing businesses with over 30 years of experience based in Verona. 100% stake. About 120k combined EBITDA.

Acquisition of Vergottini Srl in April 2024 → thermo, gas, heating, and photovoltaic maintenance firm based in Milan, and active in Lombary since the 1920s. About 160k EBITDA. 100% stake.

Acquisition of Energy Pro System Srl in June 2024 → 600k EBITDA firm with over 30 years of experience based in Lecco. 100% stake

In total, in a year and 5 months, management has already acquired about €1M in EBITDA excluding minority interest.

Lindbergh has purchased these businesses at a disciplined price, paying about 3-4x EBIDTA, often from owners who want to retire (although when possible, Michele prefers to keep the current owners in charge to maintain the team morale). Once the business is acquired, Lindbergh centralize the ERP, warehouse, and back office, lifting EBITDA margins from typically 8-10% to 15%. This gives a post synergy multiple of 2-2.5x EBITDA. Since Lindbergh trades at about 10x ltm EBITDA, this alone already creates a multiple arbitrage as it immediately pushes the valuation within the company to a higher multiple. 

The HVAC market in Italy, like in most countries, are some of the most fragmented, with 19000 tiny local providers, making up a nearly €19 billion industry. In the US, there are plenty of private equity funds rolling up these HVAC businesses while in Canada you have large public companies such as GDI doing the same thing. 

What we've learned is that the key benefits of being a consolidated player in this industry are the following:

• Having a turnkey one stop shop solution: We infiltrated Italian HVAC FB groups to learn more about local players, and one thing that stood out to us is how narrow the capabilities are of each provider. Typically, each business has about 5-50 engineers, and they only specialize in a few activities, such as refrigeration, photovoltaic systems, ventilation, air conditioning, etc. They might also have expertise in certain brands. Because Lindbergh acquires players specialized in multiple HVAC needs, they can offer everything instead of you having to find one provider for each service. Moreover, during the summer, there is larger demand for air conditioning while in winter, there is larger demand for heating. This means customers can rely on one provider all throughout the year. We found that this wasn't as significant as we thought. While many property managers prefer a turnkey solution, a lot also don't want to rely on just one provider in case of mishaps. We also have smaller customers preferring to work with their local provider who they know rather than a branded well-known player. This is why Lindbergh doesn’t slap on their brand when they acquire and tries to keep the execs and team in order to keep their local mindshare.

• Barriers to scale: While studying GDI, a comparable Canadian HVAC and janitorial rollup, we noticed while there are low barriers to entry, there are high barriers to scale. Each project and contract requires putting up working capital upfront to provide the materials and equipment for the job. This means the larger projects you have, the larger the upfront capital required. This is why GDI was able to win out on larger contracts as they faced less players with the financial capability to pay upfront. Furthermore, having a recognized brand makes you appear more reliable for big property managers. This worked wonders for GDI because Canadian real estate is more consolidated, meaning property developers and REITs hold large amounts of real estate. In Italy, it's extremely fragmented, most property owners own either one or a few compounds. Still, for Lindbergh, being a national player might get them better bids on large contracts such as airports, data centers, malls, offices, etc. 

• Customer retention and acquisition: While not applicable to simply being a national player, we do think switching costs keep Lindbergh's rollup defensible. The reason why HVAC service space has stayed so fragmented is that it's hard to steal customers. Local HVAC businesses operate through word of mouth. The typically generational-old family business has friends of friends of friends to act as consistent sources of customer revenue. Moreover, services are often done on a project-by-project-basis or on contracts spanning 1-5 years. Property managers don't want to deal with switching, and since it's a commoditized service, there are little to no pricing differences between providers, this is the case with Lindbergh too for now. There is also a benefit when it comes to having more slots on supplier registries. Many HVAC service providers are under supplier registers where customers can find and contract your team. Having more businesses simply means there’s a higher likelihood of customers choosing you.

• Economies of scale: Being able to buy equipment and materials in bulk might produce cost savings. This alongside centralizing overhead, pushes up margins and allows for price undercutting. However, we've seen that there is often little to no price undercutting as firms simply prefer to enjoy higher profits.

Nevertheless, there are a couple key risks:

→ Forward integration by appliance brands. This applies for branded appliance fields like refrigeration and air conditioning. There is a trend, right now mostly in Germany, of OEMs building out their own installer and maintenance team to service their own brand. Because they have better scale and knowledge of the product, they should be able to have better expertise in doing the job. What I think is more likely however, is that these OEMs will outsource install services to these local players who already have the training, certification, and knowledge. This could actually benefit Lindbergh as now they are dealing with bigger accounts who might prefer working with a national player with a better reputation.

→ Property managers bringing the install services in-house. This is a trend we noticed in Canada where GDI faced contract losses when big property developers decided to just build their own janitor and install team. But because the property development is more fragmented in Italy, we think most customers will still prefer to work with their outsourced services.

Overall, because there doesn’t seem to be strong competitive advantages with being a national player, the HVAC service space doesn’t make for the best rollup.

Still, the timing is definitely impeccable as we are seeing more and more demand compared to North America. This is primarily due to the tightening environmental regulation within the EU.

For example, in recently released environmental rules…

• the largest emitting properties will have to be renovated by the owners to achieve zero carbon emissions by 2023.

• Vehicles and structures are now required to be regularly checked for leakages.

• From 2027, only certified technicians will be allowed to work on HVAC systems at any time.

• Some refrigerators that use too much power will be banned by 2025 and must be modified to reduce emissions.

What this means is 1) higher barrier to entry as to be an expert certified technician requires more and more training, which is already a longer process than in the States due to the vocational training, apprenticeship, and testing that could last up over 5 years for a full suite of certification, 2) much higher demand than other regions due to the added requirements to reduce emissions, and 3) more types of services required (i.e. installing and maintaining hydraulics, leakage detection systems, modified refrigerators, as well as more documentation and renovation services). We estimate the rollup could achieve 3-5% organic growth.

Lastly, we arguably could eventually see Lindbergh using their HVAC arm to compliment their in-night delivery business, mainly by cross selling their services for technician's who need their vans to comply with regulation. 

Lindbergh plans to acquire €1M in EBITDA per year. Michele himself said he expects the HVAC rollup to reach €20M in revenue by 2026. In 5 years, ~€4M in EBITDA should be reasonable.

There are some constraints:

Lindbergh generates just about €1M in free cash flow and has a €2M cash balance. Despite maintaining minimal debt and despite the fact that most of the HVAC businesses come with a good net cash position, if Lindbergh is to spend €2-2.5M a year on this (matching Michele’s EBITDA target), they would likely have to eat into their cash balance, which we estimate could burn €200-250k a year. We think this is overall sustainable when you count the cash flow growth in the following years, but we maintain caution in ensuring Lindbergh stays disciplined in M&A.

That said, following Michele’s guidance, applying today's EBITDA multiple would imply a future valuation of ~€40M just for this segment not counting for organic growth. If GDI could reach $2.4B in revenue and $60M in operating profit with a <3% market share in Canada and the US, we bet Lindbergh could reach at a 1% market share eventually, meaning a total TAM of €190M in revenue, which would amount to a potential €200M valuation at maturity, but this is very far off.

Share buybacks

Aside from their M&A, management has also been doing share buybacks. Last year, they used a quarter of FCF to repurchase shares, buying back approx. 1-2% of S/O. This year, they’ve already bought back over 0.3% of S/O. This is minimal right now, but considering they are a small company which depends on their cash flow, it’s still a pretty significant indicator of their care for investors. And they should care since management owns approx. 65% of shares.

Valuation

To get an estimated valuation of Lindbergh, let's add up the HVAC and in night delivery segment.

Since the technician base has grown about 10% annually since 2021, if we model out this 10% growth, we end up with about 12,000 technicians using Lindbergh in 5 years, or abt a 3% total penetration. Let's stay conservative and assume revenue per technician, after the value-added services are cross-sold to France, remains at €370/technician (for details on that, read Lindbergh - part 2). We should get about ~€53M in revenue.

If we apply a 13% EBITDA margin of 2023 fiscal in which we don’t account for a higher profitability in France after a few years — at a 10x multiple, we should get a valuation of about 70M in 5 years. Add that to our 40M valuation for the rollup leaves us a 110M business, an IRR of ~30% from its current price. As I see it, 10x ltm EBITDA is undoubtedly an absolute steal! 

—

Above all however, there is one aspect which I think is Lindbergh's biggest asset. 

That is management culture. 

After speaking with Michele and members of the team. This is, by far, the most humble, open, and persevering group of business people I have ever had the Honor of meeting. 

I encourage you to read Michele's letters. Management openly admits their mistakes, and even when they talk about their accomplishments, they follow it by telling investors everything they could be doing better. They are real leaders who are in the thick of the business. In fact, Marco Pome himself led the first waste disposal. It's this willingness to get your hands dirty that sets them apart. 

See an excerpt from Michele’s 2022 shareholders letter:

This is the reason why I gained so much confidence when I heard William Thorndike was an investor. He knows they are outsider leaders who have what it takes to build a true compounder.

Typically, micro caps don't possess the moats required to succeed in the long-term. Lindbergh is an exception. Lindbergh recognized the need for more customized technician services, and understood the playbook to sustain their position. With a healthy balance sheet and disciplined capital allocation skills, Lindbergh is both building an excellent business while maximizing shareholder value.

I personally am excited to see how Lindbergh's story plays out. 

—

Always happy to listen to any comments, thoughts, or ideas from my audience. If you have any you'd like to share, comment them and I'll respond. 

Before we begin, if you enjoy this content and wish to be updated on future research, subscribe below.

Subscribe now

Thank you!

—

Cintas - a brief case study

Cintas is an American uniform rental company that is a must-read case study for any route-based business.

Cintas was started in Ohio during Great Depression by a circus couple who noticed factory workers used rags which would be thrown away after a while. They would collect them, clean them, and resell them as brand new.

By the 60s, Cintas was doing uniform rentals and was being managed by Richard Farmer, the grandson of the founding couple. Cintas worked mostly with hotels, restaurants, and mechanics. Instead of having to purchase uniforms, clean it themselves, maintain them, and replace it themselves, Cintas allowed you to rent uniforms for about a couple bucks per day per employee and Cintas takes care of everything for them. Customers have less to think about, allowing them to just focus on their business.

This was a mission critical service for many customers in sectors requiring complete cleanliness of work-wear. Factories and hospitals cannot have a single contaminant enter their workplace. And Cintas, due to economies of scale, could do this a lot cheaper than what customers would be able to do themselves.

Overtime, Cintas expanded into other services that the customer could use such as uniform and outfit laundry, supplying kitchen and bathroom supplies, supplying rags and towels, delivering first aid kits, conducting on-site safety inspection, on-site fire inspection, and deep cleaning. These facility services make up nearly half of the company’s total revenue.

Via a series of disciplined tuck-in acquisition, Cintas took out its competitors. Now, Cintas is the number 1 uniform rental player with a roughly 40% market share. The number 2 and 3 players, Aramark and UniFirst, have a vastly inferior 14 and 12% respectively. Nearly half the country still insource their uniform services, leaving a long growth runway for even the largest players in the space. Cintas boasts an impressive 20% return on operating assets.

Cintas shares are now worth $77 billion and has delivered 25% compounded annual return for the past 15 years, a return higher than Berkshire Hathaway. But how could a boring simple uniform rental business deliver better returns than Microsoft, Facebook, and Google?

• Order density → for a route-based business — transportation, warehouse management, and labor is fixed to the extent that your trucks and hubs reach full capacity. This means the more parcels you can fit in your truck and warehouse, the cheaper the per unit cost to deliver it becomes, thus achieves a scale advantage which generates cost savings that can then be used to undercut competitors on price. More deliveries in one route also achieves greater turnover on a similar amount of fuel and labor expense. Cintas achieved their dominant scale position by offering various related services so that each truck could deliver a lot more parcels to a single customer and simply acquiring more customers so that each truck has more to deliver each ride. This makes it harder for competitors to undercut Cintas on price.

• One-stop shop effect → a common strategy in many great distributor businesses. Businesses typically prefer to get most of what they need from one supplier versus having to manage accounts from multiple suppliers. By offering various related services aside from just uniform rentals, Cintas becomes a one-stop shop for customers where they their kitchen, bathroom, cleaning, and inspection services can all be managed in one account. This achieves 2 things: 1) customers prefer Cintas over smaller players who only have the scale and expertise in one line business, and 2) with this much services attached to one company, Cintas becomes more integrated into their customer’s operations, raising the switching costs. In fact, Cintas has a 96% retention rate, typically locking them in contracts spanning over 3 years.

• Roll-up acquisitions → As analysts said about Cintas: “In good economic times, uniform companies grow by increasing the size of existing accounts and convincing companies that have never rented workers’ uniforms that it’s cost-efficient and enhances their corporate image, says analyst Jim Stoeffel of Smith Barney. In bad times, the big companies grow more through acquisition, because more of the 700 or so small uniform-rental companies become willing to sell their businesses for lower prices, says analyst Craig.” Cintas has operated with minimal debt, allowing it to make investments during both up and down cycles, acquiring companies at multiple lower than what they were trading at, slowly and quietly kicking out competition. We will see that because there’s not a lot of small in-night delivery players, there’s not much utility to Lindbergh doing the same roll-up in logistics.

• Corporate culture → Cintas came from a humble environment and management remains humble to date. Management has consistently been transparent about their mistakes, and ensures their employees are too. Every employee is required to read a book written by Richard Farmer detailing the values of honesty and integrity required to work at Cintas. Executives are long-term focused as well, which in 1979 they laid out their principal objective of “exceeding our customers’ expectations to maximizing the long-term value of Cintas for its shareholders and working partners.”

As you have and will continue to see, Lindbergh is trying to follow the Cintas playbook.

—

Value-added technician services — creating a one-stop shop effect

The key differentiator between Lindbergh and all of traditional and in-night delivery services is that Lindbergh thinks in terms of revenue per technician, not number of deliveries. It’s this mindset that we believe gives Lindbergh their competitive advantages.

By thinking about how better they can serve the technician, Lindbergh has expanded into multiple value-added services other spare parts delivery providers have not. These include:

• waste management, as we’ve already discussed in Lindbergh - part 1.

• tool testing - Lindbergh picks up the technician’s tools which require periodic maintenance checks to meet quality regulation. They test the tool in to ensure there are no quality changes and that the power range complies with regulation. Adding a tool testing to your spare parts delivery adds up to +€18/tool.

• protective equipment - Lindbergh supplies technicians with gloves, overalls, shoes, equipment, and accessories directly into the van. Lindbergh has an online catalogue with around 40 items that can be accessed on T-Linq. With scale, Lindbergh can negotiate better prices from its PPE suppliers. This segment alone generated >€500k last year.

• equipment delivery - Lindbergh delivers can deliver equipment from the OEM warehouse to the technician’s van.

• spare parts return - Lindbergh picks up an unusable spare part from the van and return it to the OEM’s warehouse or repair center.

• clothing repair and laundry - Lindbergh picks up the technician’s outfit, cleans it and fixes any damages, and returns it to the technician’s van.

• shared equipment delivery - Lindbergh collects equipment from the OEM’s warehouse and sends to the van of the technician.

• warehouse management - started with Jungheinrich which requested Lindbergh to manage their forklift warehouses. Lindbergh delivers the product, manages the warehouse, and maintains the inventory.

The addition of all these services helps both the customer and Lindbergh.

For the customer, when using multiple services in addition to the spare parts delivery, could save hours. Previously, technicians had to drive to the OEM’s headquarters to dispose their waste, collect new equipment, and return spare parts. Technicians had go to a separate shop for repairing or laundering their clothes. Technicians had to spend time doing maintenance checks on their tools. OEMs had to spend time finding and purchasing protective equipment.

Why use another in-night spare parts provider and do all these tasks separately when you can have 1 account on Lindbergh where you can request for all these activities in 1 platform. This removes hassle, and more importantly, saves precious time that would have been expensed as labor costs for the technician.

For Lindbergh, these services increase order density. In fact, all these value added services aside from warehouse management contribute >50% of in-night delivery turnover (abt 48% for spare parts delivery).

The order density can be broken into two scale economies:

1. More stops to deliver per ride - with a larger variety of services, Lindbergh is guaranteed to have more stops compared to just spare parts delivery on the same fuel and labor cost.

2. More to deliver per stop - with a larger variety of services, Lindbergh has a higher chance of delivering more parcels per stop. This increases the utilization of the vehicle spread across the same fuel and labor cost. This is less significant than the first point because at some point, you will need to get more trucks when your current capacity gets fully utilized.

Since 2020 when these ancillary services were introduced, average monthly revenue per technician has gone up from about €180/technician to over €370/technician in Italy. The order density should have increased about 2x! This has increased return on operating assets, which now stand at a strong 15%.

While most customers use at least 2 Lindbergh services, we expect further cross-selling and integration to increase this, further improving order density. Moreover, the amount of possible services that Lindbergh could offer is nearly endless. Lindbergh could provide van inspection, first aid, sanitation materials, and more.

France

Lindbergh expanded to France in 2021, a was the primary reason they went public. France now contributes 40% of sales to Lindbergh.

The French logistics space is a different story compared to Italy.

The French space barely provides the value-added services available to Italy, so Lindbergh France primarily deals in spare parts delivery and waste management. This has led to a much smaller revenue per technician of about €170/technician. Yet the French segment already serves over 5000 technicians compared to the >2000 served in Italy. This is because 1) France is generally a larger market for large OEMs with typically over 2x the amount of engineers in France vs Italy, and 2) because Lindbergh already had strong relationships with multinational OEMs in Italy, it found no problem getting their French technicians onboard.

You also have a much larger proportion of waste management customers. Lindbergh’s largest customer in France is LVMH. Again, the LVMH story of choosing Lindbergh is similar to what we discussed:

Before using Lindbergh, LVMH used various local partners around France to dispose their waste. These local waste disposal services did not have any material recovery capabilities.

With Lindbergh, LVMH could track all their waste disposal on a centralized platform, allowing for more efficient data management.

With Lindbergh, LVMH doesn’t have to deal with so many accounts and suppliers. They can handle everything on a national level.

With Lindbergh, waste can be recycled and some can be recovered, saving LVMH on costs.

But while the French segment looks very promising, there are some issues:

→ The French segment is not very profitable yet. In fact, as of 2023, it’s just breakeven compared to the 20% ebitda margin enjoyed by the Italian segment (remember, industry average ebitda margin for logistics in 10-12%). This led to total ebitda margins declining to 13% as of 2023.

→ Cultural and language barriers caused road bumps for the Lindbergh team to set up a sales network.

→ Because the French segment was losing money initially, Lindbergh resorted to outsourcing drivers, while this increased profitability, it does lead to control and quality issues.

→ The competition in France is small but more developed than in Italy. There are a couple of large European logistics firms that offer in-night spare parts delivery.

• One of these is Sterne. Sterne offers both traditional and in-night delivery for a variety of niches. Still, in-night spare parts delivery is reportedly <1% of total sales. Sterne was formerly the top in-night player in France, but according to management, Sterne now controls just 25% of in-night spare parts delivery space in France.

• Another is Ciblex. Like Sterne, Ciblex is a large logistics firm involved in both traditional and in-night delivery in a variety of niches. Ciblex is a much smaller competitor which we estimate holds <5% of the market.

Once again, we see the case where large logistics firms cast a wide net, and as a result, are not able to focus on competition within a small niche.

Chats with large national logistics firms from other European countries confirm this. The Swiss Post, for example, has an in-night delivery segment, but it only comprises <0.5% of revenue. Again, they do all kinds of niche delivery from healthcare to auto parts. They are not able to think in revenue per technician because they are inherently a volume-centric business.

Lindbergh now likely controls over half the in-night spare parts delivery space simply due to the adoption of their major OEM clients they had while they were just in Italy. As Lindbergh eventually rolls out the value-added services, we expect this differentiation to enable Lindbergh to control majority of the market.

Overall, the French market is Lindbergh’s key area of growth. Despite already making 30% more volume in France compared to Italy, the market is arguably 2-300% larger. The only issues is profitability and scale, which we think Lindbergh could improve on over the next few years.

Growth Runway

How much market is there left for Lindbergh to penetrate?

Lindbergh has roughly 7500 technicians using their service. 5300 in France and 2200 in Italy.

In Italy, there are about 90-100,000 maintenance technicians that could potentially use Lindbergh. This means total penetration is just 2-3%.

In France, there is probably about 250-300,000 maintenance technicians, with about 600-700,000 engineering technicians in total across the country. This means total penetration is also about 2%.

As you can tell, the total addressable market is huge!

In Italy, the main source of growth will come from more OEMs and technicians adopting in-night delivery, switching from traditional delivery like DHL which dominate the industry.

In France, the main source of growth will come from mostly from adoption but possibly from customers acquired from Sterne.

Key risks

• While this means Lindbergh has plenty of room to grow, it also provide a gap for new entrants to establish themselves. Lindbergh is still a small company and has not established enough order density and scale economies to be completely defendable against new entrants.

• There is also the risk of larger logistics companies focusing more on the technician in-night delivery niche. While this is possible, as we’ve discussed, this is unlikely as the market for this is incredibly small compared to the traditional delivery market. Even if DHL enters in-night delivery, they will need separate training, better IT integration, a key management system, waste certification, and capabilities for other ancillary services outside the typical logistics activities.

• Finally, there is a key person risk — In the next and last Lindbergh post, we will discuss our thoughts on management, their culture, and capital allocation skills, and we will see how important Michele, Marco, and the rest of the team is in making Lindbergh succeed.

—

Disclaimer: This is NOT investment advice. All content on this website is for informational and educational purposes only and should not be considered to be advice of any nature. Due your own due diligence.

Lindbergh is a small Italian and French logistics company involved in 3 primary types of business:

• in-night logistics (80% of revenue) servicing mostly technicians of OEMs, such as those who produce forklifts, elevators, trucks, etc. Lindbergh services primarily include spare parts delivery but may also include spare parts return management, PPE delivery, shared equipment delivery, laundry, and tool testing.

• waste disposal services (10% of revenue) that are complimentary to in-night logistics for technicians or in other cases a separate contract. This segment uses the same logistics network as the in-night business.

• HVAC installation and maintenance roll-up (5% of revenue) which they started this year and which we think could, alone, be potentially worth over the entire market cap today.

Lindbergh was founded by CEO Michele Corradi and Chairman Marco Pome. The two met as consultants at Steer Davies Gleave in 2003, where they worked on macro-logistics projects, typically for state infrastructure, as well as industrial projects for major companies like Vodaphone. Michele and Marco wanted to continue doing work in industrials, but due to differences with their superiors, the two eventually founded their own consulting firm in 2006.

Slowly, they shifted more and more to waste management, obtaining their waste collection permits in 2008. Lindbergh started to help offer the first in-night logistics services in Italy in 2013 as a consultant of TNT (now owned by FedEx), one of the largest delivery firm in Europe. TNT eventually exited the in-night business due to downsizing and the in-night business being a very very small portion of sales, passing the business to Lindbergh which kept their customer contracts.

Lindbergh expanded to France in 2019 through a JV which Lindbergh eventually acquired majority control of a couple years later. It was also at this time that the T-Linq platform finished development, allowing technicians and customers to control all their logistics needs digitally.

Lindbergh delivered 28M EUR in revenue in 2023, making for a whopping 38% compounded annual growth in the past 4 years.

Financials

Lindbergh has grown sales at an annualized rate of 35% since IPO. Lindbergh has exhibited strong track record with a return on equity >20% every year since IPO on very relatively low debt levels (earnings can pay interest 24x over and equity covers debt). Because the France margins has lower margins (due to lower scale), the expansion caused ebitda margins to decline from 20% to 13%, which is still close to the average for logistics firms. It’s a moderately capital light business. Scale works with order density as you can fill up your trucks and hubs as you get new customers so Lindbergh’s return on assets, currently hovering at 5%, and their margins, should go up as order density improves unit economics.

Why does this opportunity exist?

→ The company trades at around 10x ev/ltm ebitda of about 3M EUR, which we think is a fair price for a wonderful business.

Lindbergh is a micro-cap in Italy. There is almost no sell-side coverage. It’s a boring, dirty, complicated business, and we know how much compounders have come out of the “boring” industrial service/distributor playbook.

Our view is that Lindbergh provides a unique opportunity to own a potential hundred million dollar compounder while it is still a microcap, for these reasons:

1. Lindbergh provides a mission critical service that massively improves technician productivity characterized by high switching costs and significant cost savings for the customer

2. Lindbergh operates in niche and complex space often ignored by logistics giants such as DHL and FedEx, leading to virtually no direct competition in Italy and minimal competition in France.

3. Lindbergh’s cross-selling of other value-added services for technicians in addition to the regular spare parts delivery contracts creates a one-stop shop effect and improves order density which further drives up potential scale advantages.

4. Lindbergh’s HVAC service roll-up is very promising due to the fact that the Italian HVAC service environment is extremely fragmented. While Lindbergh just started doing this this year, we expect this segment to contribute a sizable chunk of revenue within the next 3-5 years.

5. Lindbergh has a management team that openly admits mistakes, has immense skin in the game, and are disciplined capital allocators. 

Let’s dive in and break down Lindbergh’s value proposition.

Mission critical service

First we need to know what value Lindbergh actually provides to their customers.

Lindbergh charges €20-24 per deliver, which is >33% more expensive than a traditional delivery which would charge €15 per delivery.

Why is Lindbergh more expensive?

OEMs/manufacturers, think elevator, forklift, auto brands keep large amounts of maintenance technicians who are there to serve their end-customers to conduct maintenance, fix, and repair their products.

For example, Jungheinrich, Europe’s largest forklift manufacturer and one of Lindbergh’s longest running customer, would have a large team of technicians, each with their own vans, to help support customers when they need it. Now, technicians require delivery of spare parts when they run out.

This is where Lindbergh comes in.

In a traditional delivery service like DHL, the spare parts are delivered to DHL’s warehouse where the technician must pick it up.

Jungheinrich tells us their process was like this:

“5am - Spare part picked up by DHL from their German warehouse

6pm - Spare part is dropped off at a DHL warehouse and then sent to a Florence connection site

9pm - Spare part is dropped off at the connection site in Florence where the technician has to pick up the item.”

See how time consuming this is…

Lindbergh on the other hand delivers the spare parts inside the van of the technician (yes, they manage a replica key), overnight. This is their flow:

“All spare parts are stored in Lindbergh’s facility rather than Jungheinrich’s own warehouse.

9pm - Spare part is collected by Lindbergh and sent to the technician’s van

7am - By this time, the item is already in the van”

So when the technician enters their van, the part is already inside. No effort, no waste.

This saves the technician about an hour, which is huge considering that the average maintenance technician will cost you €50-60/hour. If you count the €15 per delivery + €10-15/hr (estimate) additional fuel costs for having the technician drive up to the collection site, the OEM saves about €50 per delivery, a nearly 70% cost saving — and remember, technicians need a consistent flow of spare parts, ordering an average of 2-3 parcels each week, meaning customers can’t afford for even a week to stop the delivery of spare parts.

—

Lindbergh thinks of it’s customers as partners. They do this by structuring it’s contracts where customers agree to pay for a set amount of deliveries over a set number of years, typically 2 or more. This creates a lock-in period, establishing a potentially long-term relationship with Lindbergh.

One of the best signs that told us Lindbergh’s in-night solution had high switching costs was the fact that Lindbergh had only lost 1 customer since inception, a small French customer contributing $200k rev, and it was due to the firm’s own cost-cutting measures rather than dissatisfaction with Lindbergh. This represents a retention rate since inception of >95%.

In fact, around half of Lindbergh’s customers have been with them for over 10 years!

So why are switching costs so high?

1. Implementation process

Setting everything up requires a ton of integration with your technicians, logistics network, and software. From start to finish, the implementation time may take up to one year.

• Procuring replica keys: From collecting the keys, to making the replicas, to returning the keys to every technician’s van, this is a long process. These are typically hundreds of employees — and for some, such as a KONE director we spoke with, maintenance technicians are nearly 80% of the workforce in his region. This is the key difference in integration compared with traditional delivery that makes in-night logistics much more sticky.

• Onboarding technicians on T-Linq: Technicians tell us T-Linq is quite easy to use; not much of a learning curve. There is still short IT training and data integration that has to be done. It’s a typical on-prem software implementation.

• Master data integration: This takes about 20-25 days and includes setting up T-Linq and Lindbergh accounts with the ERP system. For example, Toyota Material Handling called up their IT office in Sweden to set up all the data and ERP.

• Pilot program: Typically lasts a few months. This gives a chance for the customer to try out Lindbergh.

Overall, joining Lindbergh is a significant time investment for the company, but it makes a deep integration with their systems, making it very time consuming to switch to another delivery service, as well as a deep relationship with Lindbergh.

We’ve seen this relationship take place when Jungheinrich asked Lindbergh to manage their warehouses. Lindbergh’s response was to create a whole new business unit, which is now 3% of their revenue.

2. T-Linq platform

The T-Linq platform is an on-prem order flow software created to connect technicians with Lindbergh services all in one place. While the learning curve is small, there are still some switching costs in getting used to how the platform works.

Remember that the software is used by a hundreds of people within an organization. And this is not your typical order and tracking software used by DHL or UPS since it includes a variety of other functions built to serve a team of technicians.

With T-Linq, technicians can order spare parts with the press of a few buttons. T-Linq became extra important with the roll-out of other value added services such as laundry and tool-testing. Now, technicians have access to all these services in one screen where they can book and track services.

T-Linq also provides a shared calendar with the whole team, an alert system for issues and anomalies, documentation and report drafting, and a messaging system with video and picture compatibility. T-Linq is fully customizable: the layout of the apps can be rearranged, and the branding/design can be edited to fit the customer’s preferences.

T-Linq is not just used by the technicians but also the service directors. Service directors can control the number of PPE being requested, which equipment can be used, and can easier track the schedule and flow of technicians and deliveries.

Overall, the platform’s customizability and usage in many parts of the job across organization hierarchy makes it hard for technicians and directors to switch to another delivery provider and having to get used to another kind of software.

Complex space with little direct competition

In Italy, Lindbergh is the only significant in-night logistics player.

Considering the huge cost savings, customer-seller relationships, and stickiness of the business, that begs the question of why haven’t more entered the space? Why haven’t large firms like DHL entered the space?

Customer standpoint:

• In-night in-boot delivery took time to be adopted by OEMs. In Italy, many customers were hesitant to give logistics providers key replicas to their vehicles. Furthermore, delivering directly to technician’s vans put the control out of the OEM’s own warehouse management into the logistic provider’s hands, causing fears of inefficiency and untimeliness.

• Many firms initially did not understand the cost and benefits of improved technician productivity. Lindbergh’s main point of contact are service managers who man the technician team. The most common case of a failed sale is when the service manager’s proposal is rejected by the finance department or higher ups who don’t focus on the technicians and do not know the costs to that labor as well — so it took time to educate customers and to build the operational capability required to ensure 100% efficiency.

• Really the only firms that truly started to adopt overnight delivery was large OEMs. Small local Italian technician teams typically have only 3-4 engineers each who could potentially benefit from in-night delivery. Moreover, these small firms tend to keep to tradition. The hassle of integrating Lindbergh was oftentimes a major drawback. This is why >90% of Lindbergh’s customers are not just large OEMs, not just international OEMs, but German OEMs, who often are more keen on investing in productivity.

Competitor standpoint:

The in-night space was simply too small to really care about. Spare part delivery is just one of many niche in-night delivery verticals which also include medical equipment, uniforms, documents, and auto parts. Furthermore, compared to traditional delivery, in-night delivery typically involves a small amount of deliveries per customer.

The in-night space as a whole paled in comparison to the classic 3PL retail, supply chain, and e-commerce delivery services dominated by the likes of DHL and FedEx. The potential gain from instating in-night delivery is a rounding error to their tens of billions in sales.

Imagine you are DHL and you you had to choose between putting effort to secure a 3PL logistics customer which means making thousands of deliveries every week vs a spare parts technician customer which means making just hundreds of deliveries every week, which would you choose? The large OEMs can provide DHL immense order density on the 3PL side, but not on the in-night delivery side.

That does not mean these bigger players do not delve into in-night delivery. TNT was one of the largest European logistics providers. Before being acquired by FedEx, TNT was one of the early providers of in-night delivery. In Italy, TNT was the only provider of overnight spare parts logistics. They enjoyed some early success with their first major customer, Jungheinrich Italia.

However, following the FedEx acquisition in 2016, TNT was soon forced to sell-off it’s entire overnight delivery arm. It’s Central European segment was sold off to private equity, while it’s Italian operations were discontinued, giving Lindbergh the opportunity to grab the market for themselves.

Why did TNT let go of the segment? It’s because the overnight delivery arm was so small that it comprised <1% of TNT’s turnover. In Italy, overnight delivery was just €2M in sales compared to the €400M for the entire business. As TNT made an effort to centralize it’s CapEx on core businesses, overnight was naturally turned a blind eye and eventually dropped from the company.

→ It’s also important to know that developing overnight logistics is much harder than traditional logistics

• As we’ve mentioned multiple times, making key replicas is a time-consuming and relatively costly task. The act of even getting the keys from all the technicians is a bloody operation.

• Because in-night in-boot delivery requires spare parts to be delivered directly to the van, you need software to optimize the delivery route. Lindbergh had to develop this technology to do this whereas traditional delivery services only have to send the parcels to a small fixed number of locations

Waste Management

Lindbergh collects waste from technician’s van overnight, properly disposing it in their waste collection sites in accordance with regulation.

Previously, technicians would have to bring their waste back to the company’s hubs where the OEM would have to dispose it themselves via the large waste management providers.

• Again, similar to spare parts delivery, this is a mission critical activity as technicians need to dispose their waste to abide by the stringent EU waste regulation.

• Again, Lindbergh’s service saves a lot of time by not having to drive all the way to the OEM’s hub to drop off the waste. This comes at a very efficient cost too: when customers use Lindbergh’s spare parts, they can add on the waste management for cheaper, leading to a €30 price per delivery for both the part drop-off and the waste pickup, meaning for each waste pick-up, you just pay an extra €5.

• Again, technician micro waste collection is paid no attention by the large traditional waste providers. Overnight waste collection and disposal is a mere fraction of these what these firms take care of in a day.

• Again, most of these traditional waste management firms use a intuitive platform and routing software like Lindbergh. In fact, a lot of Italian waste documentation is still paper-based.

• The main barrier to entry here is that getting waste disposal certification in the EU is a pain in the ass. It took Lindbergh nearly 3 years to get a waste disposal permit just for Milan.

Like spare parts delivery, waste management is a complex, dirty, niche business that not only has the customer advantage to create long-term growth but also further improves the order density for Lindbergh trucks deepening the cost-savings.

Sun Mountain Partners, investors in Lindbergh (Chris Solberg is a seasoned investor who won a major Buffet case competition while William Thorndike wrote The Outsiders, an essential read on great management) put it like this:

“Think of the difference between Lindbergh’s model and the standard couriers’ and waste managers’ as low volume, high complexity, high cost per unit versus high volume, low complexity, low cost per unit.”

—

In the next posts, we will discuss how Lindbergh is winning in France, how their other higher margin delivery services are improving their order density, and how Lindbergh is building a new and promising HVAC rollup that could be worth the entire market cap of the company today!

Disclaimer: This is NOT investment advice. All content on this website is for informational and educational purposes only and should not be considered to be advice of any nature. Due your own due diligence.

Software platforms face cutthroat competition, constant pressure to innovate better than peers, and the fine line between burning through all your cash to grow and one day becoming profitable. Within the available pool of ideas, we have not found any software businesses that meet our standards at a reasonable price, until today.

Eventbrite is a quickly growing high switching cost event management tool and live experience marketplace benefitting from an extremely fragmented online ticketing industry. Eventbrite trades at a reasonable valuation of about 10x EV / NTM AEBITDA due to, as we believe, investor ignorance regarding the churn of Eventbrite’s core and non-core customers. 

Over several months, we have been studying the event management software space, an extremely fragmented industry, each player with different features, niches, customers, and capabilities. A ticket tool is just a payment processor where you make money by charging a fee on ticket sales. But our research shows it’s a surprisingly complex business.

Event ticketing demand is growing worldwide at breakneck speed. We realized if we found the best player in the industry, it would no doubt completely dominate the whole space. And we think this title belongs to Eventbrite as a result of their brand awareness, improving unit economics, and network effect through a shift in its business from being just a ticketing software to a marketplace for live experiences.

Overview

Eventbrite customers are mostly small to medium-sized event creators, below 100 attendees per event. Eventbrite targets these creators by having a simple easy-to-use platform customers can use by themselves. But this wasn’t always the case. In 2020, management made a huge decision: getting rid of its on-premises solutions that required a ton of customer support and set-up time. This effectively took out most of its high price-point operations, which at the time was half their revenue, leaving larger clients to either leave, or continue with their self-serve platform. 

It was harder to compete since the big creator event space is less fragmented. Ticketmaster, owned by Live Nation, holds a monopoly in the space, hosting celebrities like Taylor Swift, Ed Sheeran, etc. Ticketmaster had the customers, who would not switch, as well as better scale advantages. They also had an undeniable supply advantage since Live Nation owns so many concerts and stadiums. While this may have hurt Eventbrite’s top-line, the switch also made for a much leaner operation with a larger emphasis on improving the core self-serve platform that appeals mainly to frequent small creators. Churn has consistently stayed under 3% before and after the switch — until Q4 2023…

Investors were spooked following management raising the take rate from 8 to 10% in Q4. Paid tickets and creators still went up 7 and 8% respectively compared to double digits in previous years. Because management did not give a concrete plan to mitigate these issues, Eventbrite lost >40% of it’s market cap — not pretty.

We monitored a sample of almost 70 paid frequent creators over the past quarters and found that following the price increase, the churn among frequent creators was less than 10% QoQ for both Q4 and Q1! What this means is that the churn was not really felt among Eventbrite’s core customers that generate most of its revenue. Instead, we think the real churn was faced by non-frequent or free creators whose events tend to be less commoditized and thus don’t benefit much from Eventbrite’s marketplace, we will explain what this means. 

Business Model

Eventbrite makes money by taking a portion of ticket sales and a fixed amount for each ticket sold (no fees for users who sell free tickets) as well as a subscription fee (or per event fee) which gives access to the dashboard, tools, and marketing suite. 

Initially, Eventbrite offered two plans: the Essential plan charged 2% and $0.79 per ticket while the Professional plan charged 3.5% and $1.59 per ticket, a take rate of 7.5%. But this year, Eventbrite cut out the Essential plan and raised the price of their Professional plan, charging everyone 3.7% and $1.79 for all the features, raising take rate to 8%. Despite this, the price change resulted in just 1% of creators leaving. The take rate was further raised to 10% in Q4 due to the subscription fee addition and added the marketing suite. 

After comparing price points from a bunch of competitors, it appears that Eventbrite has a large price premium over its peers despite possessing an almost similar set of core features. 

Eventbrite and these three are the largest self-serve North American-based players. Although the space has thousands of competitors, we could barely find any remotely viable self-serve competitors through Google search or website trackers, especially internationally. Most ticketers barely had 100k monthly traffic. 

Eventbrite has different take rates in other countries with different payment structures, but take rates are still consistently below local players. We expected scale advantages to drive cost savings that would return in the form of lower prices relative to peers. But our research tells us this price point is justified as a result of their competitive advantages. 

Breaking down Eventbrite’s value proposition

Marketplace and SEO Ranking

One of Eventbrite’s competitive advantages is that they can promote creator events way better than peers. Approximately 29% of ticket sales are driven by Eventbrite.

Marketplace-driven purchases are made either through people actively browsing on Eventbrite or through events recommended by Eventbrite on the website or app, sources tell us the latter comprises the bulk of marketplace-driven orders. The more creators you have, the better the event selection and recommendation options, which leads to more customers on the marketplace, which leads to more sales creators will receive, motivating even more creators to sell on Eventbrite rather than competitors. 

Eventbrite’s traffic is multiples higher than competitors, which provide, by far, the largest audience for creators. Eventbrite also dominates the supply side with 850k creators, 10-16x more than all the direct competitors we’ve looked at. Eventbrite has the largest customer base, selling 93M paid tickets in 2023. The relative scale of creators make it by far the best marketplace for creators and consumers, a self-reinforcing value proposition that makes it more and more difficult for peers to compete against. 

Since the pandemic, around a quarter of their creators have been “frequent” creators, those who hold an event every few weeks to a month, with typically below 50 attendees, yet 60-70% of ticket volume comes from this group. These creators use ticketing platforms as a regular source of income. The most common profiles were small business owners. Restaurants realized they can host cooking classes in the afternoon. Bar owners saw how much they could earn through monthly stand-up comedy nights. Artists learned they could sell drawing workshops during off-hours. Eventbrite provides the best platform for these frequent creators to promote their business. 

Marketplace sales vary a lot. Some customers we've talked to had over 50% of sales from the marketplace while others had 10% and below. We find more commoditized events like speed dating, amateur comedy, parties, pub crawls, weekly yoga, and classes benefit the most from the marketplace. These events have similar formats, and less room for differentiation or brand building. Event listings like museums, attractions, conferences, music are different because they are typically tied to a name or brand beforehand. Their customers typically find them on Google, searching up the creator’s name. Bigger events also tend to be less commoditized. 

Since the shift to frequent creators, Eventbrite’s creator-base has shifted from being dominated by music creators to a more diverse group. Music creators now make up only 22% of tickets. No category makes up over 25% of ticket sales, which as defined by EB include business, music, health/sports, visual/performing arts, community, food/drink, education, and charity/causes. We think this breakdown can be misleading, as there are multiple niches that dominate Eventbrite that can’t be categorized strictly into these groups. From our analysis, the largest groups are nightlife, followed by amateur comedy, then speed-dating/social, then tastings/food. We find that these more commoditized events get more value from the marketplace. If you are selling tickets to a famous music festival, chances are the customers know you and would search you up directly. But for experiences like pub crawls, there are thousands of crawls that follow a similar format, leading consumers to consider multiple listings, which leads them to the marketplace. 

Customer acquisition is important in order to drive both creators and consumers. Most creators we talked to found Eventbrite through Google search or word of mouth. Majority of end-buyers also find events through Google (search pasta class, wine tasting, speed dating). This means SEO is important, and Eventbrite’s SEO advantage comes with scale: larger traffic size and creator landing pages organically raises your ranking. Eventbrite can also vastly outspend most competitors on SEO ranking. If you Google “events near me” right now, you can almost guarantee Eventbrite will come up at the very top, followed by a whole palooza of blogs, forums, and articles, except competing event platforms. 

For FY2022, the % of Eventbrite-driven tickets was 29%, 1 percentage point higher than FY2021. It’s slow, but it shows that the flywheel is spinning. The company is at the point where it is slowly losing its reputation as a ticketing tool and getting a larger reputation as a platform for live experiences. Note: FY2023 states platform driven tickets of 47% of volume, but since they now include affiliate links, we do not trust this figure.

Switching costs and Brand

With the focus now on frequent creators, for a large part of its creator base, Eventbrite is quite mission critical. Frequent creators typically host events every few weeks or monthly. Event ticketing is a SaaS tool. When used in scale, creators can’t afford to spend time trying to get used to a new event platform with a different user interface and system. Eventbrite has historically kept churn at 3% and below, even before its shift to self-serve.

This is the exact reason why nearly all creators we interviewed refused to switch. You need to familiarize yourself with tracking analytics, building out your landing page, creating your ticketing terms, setting up your emails, integrating your website and social media, etc. There are quite a lot of steps. Similar to many front-end business softwares, the learning curve for starting is low but the learning curve for scaling is surprisingly high. Creators have to learn how to schedule multiple events, automate multiple event content creation, and connect their social media accounts. As Eventbrite keeps adding more and more features, now with SMS notifications and microservices, switching costs should go up. 

There are plenty of Eventbrite creators who sell tickets on a website or bootstrap set of forms in addition to Eventbrite. It’s definitely more economical to use a bootstrap method, registration tool, or own website to do ticketing, but it doesn’t have the needed features and doesn’t work as robustly as a ticketing tool in scale, and it takes time and effort to build those systems whereas Eventbrite is really easy to get on and use it self-serve. This means larger creators could be more inclined to spend on building out their own ticketer, but for smaller frequent creators who just want something easy to get started with, Eventbrite provides more value.

Brand equity and social proof is high. A lot of creators we talked to used Eventbrite simply because it was popular and everybody was using it. Furthermore, Eventbrite, without a doubt, has the best user-interface across the board. One of the key things creators and consumers look for is trustworthiness of the site. Ticketers have been notorious for withholding creator-payments (EB is not entirely innocent of that during the pandemic) during tough times to keep the business afloat. Ticketers are also prone to overload for on-sales/holidays or be a house for scammers. Creators want to ensure consumers trust their ticketer and that their money is kept safe. Smaller or new ticketers would have a harder time building that social proof similar to what Eventbrite already has.

The most surprising thing we learned is that end-buyer switching costs also exist for recurring end-buyers going to the same creator events since they are used to using Eventbrite. Creators switching to another user could throw consumers off, but will likely not significantly affect sales. 

The only susceptible switchers we see are customers using Eventbrite solely as a ticketing tool for private events or those that make <10% of sales from Eventbrite-driven tickets. They are very price-sensitive since all the event platforms offer nearly the same product. But for frequent creators, Eventbrite’s bread and butter, the switching costs coupled with the audience is too good a value proposition, even if you are paying more than similar platforms. 

What’s the TAM?

We believe the growth runway for Eventbrite will come in three packages: 1) grabbing customers and creators from direct competitors, 2) geographical expansion, and 3) growth of the small events industry as a whole. 

Direct competitors

The event ticketing industry is extremely fragmented. We estimate Eventbrite currently holds at least 50% of the market in North America based on traffic volume. The next largest competitor is Feverup, Ticket Tailor, Ticket Leap, and Universe, followed by hundreds of tiny players. Competitors tend to fill a specific niche or price point. Dice, for example, targets music creators. Cvent targets larger creators at a higher price point, but with more specialized and complex features. Universe, the small event platform subsidiary of Ticketmaster, is very low cost at its lowest price bracket but has simpler features. We think Universe poses the greatest threat to Eventbrite due to its connection with the Ticketmaster platform, but since Ticketmaster and Universe have different audiences, their promotion capabilities are a lot worse. We see very sparsely populated event listings in the Universe search platform.

We expect Eventbrite’s superior marketing capabilities to eat away at the market share of direct competitors that aren’t niche or specialized enough to provide better value for creators, and consequently, for consumers. Eventbrite’s product is also way more user-friendly and feature-rich than most competitors as evidenced by better reviews and our own experience using the product. 

International growth

Breaking down the geographic sales, 60% of Eventbrite paid tickets come from North America, 14% from the UK, 6% from Australia, and 19% from the rest of the world. This has been quickly changing, with international sales gaining a larger and larger share of the pool.

Due to the nature of the business, the network effect only works when the consumers are near the event creators. If you’re in an untapped area, no matter how many creators Eventbrite as a whole has, your local platform will most likely have more near you. This makes it more difficult for the company to quickly expand to new regions compared to the typical software company that can scale globally at minimal cost.

Nevertheless, for English-speaking countries like the UK, Ireland, and Australia, Eventbrite has captured a majority market share purely organically with similar churn to North America. In fact, to date Eventbrite has done zero ticket subsidies and zero partner creator in any of the regions they’re in. Sources confirm that Eventbrite has consistently outspent competitors on SEO to grab market share. For non-English speaking regions, they’ve typically entered by buying local players, from Brazil all the way to the Netherlands. Since the pandemic, Eventbrite has more than doubled international revenue, but North American revenues continue to grow much faster, compounding >37% annually since 2020.

This is mostly due to the new regions having thousands of small competitors with poor user-interfaces, scam-filled markets, and/or overload-prone systems. The only exception has been Brazil where Sympla takes 1st place and generates larger traffic and charges a higher take rate. Although Sympla employees refused to speak with us, they mentioned that while everyone knows about Eventbrite, their threat has never been discussed in the workplace. The higher switching costs will make it hard for Eventbrite to dominate.

From what we hear, Eventbrite expects saturation in the North American market within the next several years. They also expect the UK and Australia to reach maturity in the near future. The longest chunk of the growth runway will have to come from non-English speaking regions. Right now, the 19% ROW revenues are mostly made of Europe and LatAm, and no doubt there are plenty of opportunities for geographical expansion. But we also believe their biggest potential market is India, where Eventbrite has an office and where a big chunk of their engineers reside. India has multiples the potential customers Eventbrite currently has combined and like other regions, there are thousands of small clunky competitors. 

Industry dynamics

When we asked former employees who the scariest competitor is, one of them stood out when they said that it was not the direct but the indirect competitors who pose the greatest threat. Eventbrite is competing against concert shows, sports games, and Netflix; all these guys are trying to be the answer to the question “What do you want to do with your buddies this Saturday night?” And on these terms, Eventbrite has a tiny share of the market, but it’s mainly because people aren’t used to finding small events on a regular basis. 

We think this generally underserved market has huge long-term potential, especially internationally where the landscape is barren, and even the small players present have nowhere near the robustness and user-friendliness of Eventbrite. As the industry grows, it may be plausible that happy customers will start getting recommended and attending small events more frequently, bringing their friends. And these friends also bring their friends, and so on, giving event platforms a stronger and stronger advantage in getting people to choose them as a regular choice for recreation. 

On the creator-side, there are still plenty of small businesses and people who may be drawn to Eventbrite as a possible source of income. While we don’t have numbers of this penetration, we know that every restaurant with downtime hours can host classes in the afternoon and parties late at night. We know every bar and club would reach a larger audience hosting parties or crawls or socials at Eventbrite. We know every hotel could host stand-up comedy, tastings, and socials in their lobbies on Eventbrite. We are confident that the potential penetration is significant, and it will take time for word of mouth and education through user base growth for this to happen. There’s just so many ways event platforms can be used to generate income for creators.

Improving unit economics

Just late last year, Eventbrite released Eventbrite Ads, which allows creators, by paying a fixed daily budget (similar to Google’s SEO system), to have their event on the top of the search list or on the recommended page. This could provide a huge source of revenue to the company, and should become a more and more appealing product as the marketplace gets bigger.  Eventbrite Ads is already at a $4.4M run rate, which we believe can easily double for the next few years.

The second “big” release was Eventbrite Boost, released in 2021, which is now incorporated into the general subscription plan. Boost is a marketing SaaS that allows creators to manage email and social media marketing all in one place. Creators are able to automate more emails in much larger volumes, create and automate Facebook, Instagram, and Instagram Stories ad campaigns, get marketing and targeting recommendations leveraging Eventbrite’s own data, and build multi-event marketing schedules. Instead of having to manage separate accounts on social media apps and email SaaS tools, they can do everything in Eventbrite plus get targeted data-based recommendations. Creators who use Boost marketing tools increased their ticket volume per event by 63%. As of 2022, Boost was at a run rate of $1.6M, up 4x from the year it launched. Boost reached 10k subscribers in Q2 2023, making up 1.25% of all creators. The switching costs are quite significant too, because the marketing activities of frequent multi-event creators are mission-critical. Once your email base is on Boost, creators need to spend time setting up their email contacts on another platform as well as learning how to use the new user interface. 

These services were eventually combined into the regular Eventbrite pricing in October, which meant all creators were forced to have them available. The TAM for marketing is immense. According to creator surveys, around 8% of event costs are associated with ticketing, while 20-40% of costs are associated with marketing.

Something interesting to note is that each ticket sold generates a mini float. Eventbrite holds cash generated from ticket sales for the majority of its creators until after the event is completed (qualified creators are exempted). This float period is short though, usually only a month before returning it to creators. Eventbrite has for the past couple years been investing most of these in short-term T-Bills, generating extra return on the held cash and can serve as a natural hedge for a high-interest rate environment. 

Barriers to Entry?

The barriers of entry to the event platform space is low. Let’s start with large entrants. What is the chance that big network firms like Facebook or Google, make their own integrated live experience marketplace. They have way bigger traffic, users, and network compared to Eventbrite. Facebook has way more users than all the event ticketing platforms combined. But our research tells us otherwise.

Eventbrite already has a brand that is difficult for other big firms to replicate. Google is known as a search engine. Facebook is known as a social networking site. Square is a POS system. Existing customers who are already browsing on Eventbrite may try out their event platforms but Eventbrite will still hold top of mind. They know what the experience of using Eventbrite was like; there’s little search cost. Creators know that when they launch an event on Eventbrite, it’s going to circulate amongst people who are specifically looking for events. If this was on a broader network site, it would be more difficult to target the people you want. Eventbrite’s focus on its niche is what will help it defend against large entrants. 

Moreover, investing in the ticketing software is more complex than it looks. Different countries have different ways of ticket payment processing. In Germany, wire payments are preferred over credit cards. Latin America has a very unique payment system where a coupon has to be printed out and scanned at a nearby bodega to confirm a purchase. You also need a ton of features: discount options, ad integration, email integration, reporting, on-site check in, etc. Different kinds of events behave differently. Music creators care a lot more about promotion and reporting. If their seats are not filled enough, audiences get the feeling people don’t want to watch, no social proof. Sports and conferences need very detailed on-site check-in features, printable name tag integrations, and updates before, during, and after the event. Parties and speed-dating events need detailed personal information forms, age verifications, and options to choose drinks or other items. It takes time and significant investment to find the hearts of each category of event creators. 

Speaking for ad-revenue giants, the margins on ads are way higher than the margins for tickets. It makes little sense for firms this big to light their money on fire all for it to end up as a rounding error on the top-line. Instead, integrating with Eventbrite as a payment processor partner is a far more cost-effective way to incorporate events in your site, which is what Facebook has opted for. A former business development employee at EB had this to say about established firms entering the space:

“Constant Contact was a company who through their email marketing product, they had done exactly what you said, like 5 or maybe even 10 years ago. They had created an event RSVP product, connect your PayPal or Off.Net account, take payment, and it was bundled into the Constant Contact subscription fee. They got, whatever, five plus years into that, and said, "This is a pain in the ass, we don't want to maintain this forever, and our product is not nearly as good as Eventbrite's," and so they came to Eventbrite and said, "Hey, can you guys just power an events product that we can bolt on?" And so, I think that's actually the opportunity for Eventbrite..”

While new entrants can easily set up ticketing software, it is notoriously difficult to scale. Unlike most software businesses, ticketing has volumes that behave in a very bizarre spikey manner. When popular events go on-sale, you usually have immense demand at the very beginning, simmering down overtime. What this means is that 90% of the time, running your software is easy, but 10% of the time, you need a crazy amount of server space and infrastructure. And every crash and error is an unhappy customer. Your site’s reputation gets ruined: customers are unhappy, now creators are unhappy. Even scaled first like Ticketmaster often face massive reputational damage due to systems not being able to handle the sheer amount of on-sale volume. Since they’re a monopoly, consumers have no choice. But for small frequent creators with thousands of options, it takes you out of the game. You need to ensure you get creators that are large enough to bring in profitable customers but not too large where your system breaks and everyone leaves. 

Development future

Research and development

Eventbrite is unprofitable, but only because of R&D spend., which makes up 30% of revenue, higher than 20% pre-pandemic. Immense investment had to be made to make the self-serve model as reliable and versatile as possible since Eventbrite made the bold move of cutting off their on-premise solution. Now cash is being thrown at AI to automate event creation. Frequent creators will soon be able to let Eventbrite publish events for them, generating all the required content and options, streamlining the creator’s business. They recently had a partnership with OpenAI which allows you to auto-generate captions and event descriptions. We think Eventbrite will be a net benefactor of AI and at its relative scale should be able to invest the most in it. Our guess is the price increases are meant to give more leeway for management to spend on R&D in order to accelerate feature rollout, so we don’t expect Eventbrite to be profitable on a net income basis anytime soon.

Most recently, Eventbrite made a huge infrastructural change: moving from their older rigid one-program software to a dynamic micro-service architecture. The new architecture allows different products and segments to be operated on separately. In other words, this change signals the decoupling of the core ticketing software from its ads product, platform, and other components alongside leaving their legacy software and tech debt. Many engineers at Eventbrite have complained about these issues, so it’s a good sign that their voices are being heard by decision-makers. 

Product Strategy

Many basic self-serve products such as a simple dashboard or adding Google logins were rolled out within the last few years. Eventbrite has very recently and is still at this moment becoming the self-serve platform it wants to be; the extent of its product capabilities' steady state has incredible potential. 

The past 3 years has seen a shift in focus from acquiring creators to acquiring consumers, and this meant putting more resources both into marketing and product development aimed at growing the consumer-base rather than the creator-base.  Eventbrite’s former Chief Product Officer and now Advisor, Casey Winters, has been credited with planning and executing most of the product-strategy since 2019. He was a major contributor to the decision and execution in cutting off the on-premise segment, revamping the legacy tech infrastructure, and rolling out Ads and Boost. Winters has had extensive experience in platform businesses, working with Grubhub in its early years, building out Pinterest’s network effect, and being an advisor for multiple firms including AirBnB and Reddit. We think this shift was somewhat his idea. 

Well, Ads and Boost, while used by creators, definitely boosted the demand for events. Ads improved the discovery experience by enhancing event targeting while Boost simply helped creators do marketing better, expanding their reach to more end-buyers. Something as obvious as releasing a landing page video upload feature saw creators who added videos to their listings get a 74% increase in page views. Or simply adding a rating system for consumers added social proof in event selection. A lot of resources were put into the personalization of the search experience. Customers who attended or searched up an event will get recommended similar listings. SEO spend is very large, but more effort is being placed into social media and influencer partnerships. Creator sales growth overall has been going up 40-50% YoY. It shows how much Eventbrite still could add to make their marketplace experience on par with mature platform businesses. 

Acquisition History

Eventbrite’s acquisition strategy has been quite opportunistic. Acquisitions have been made either to secure talent, engineering, or to enter a new market. 

Eventioz was acquired in 2013 at an undisclosed amount. Eventioz was founded in 2008 in Argentina. By 2013, they had operations in Brazil, Chile, Colombia, Mexico, and Peru, hosting 15k creators, 20k events, and selling 1M tickets annually. Eventbrite was neck-on-neck with Sympla in volume and revenue. We heard they had issues getting creator and consumer trust with the bodega coupon payment system, which caused reputational damage for a time, now we hear Eventbrite’s LatAm segment is growing faster than all English-speaking regions. The people are culturally outgoing, jovial, and social. Sympla still remains the biggest threat.

TicketFly was acquired by Eventbrite in 2017. It was Eventbrite’s biggest competitor in the US with a seventh of Eventbrite’s ticket volume and had a large market share among music creators. Ticketfly was purchased by Pandora, a music-streaming service, for $450M 18 months prior to Eventbrite’s acquisition. Pandora was forced to fire-sell Ticketfly, allowsing Eventbrite to buy it for $200M, a nearly 60% drop. At the time of acquisition, TicketFly generated $54M in revenue for the year, which means Eventbrite bought TicketFly at 2.4x revenue when most software companies were getting acquired at 4.5x revenue. Nevertheless, management faced difficulties integrating the creators. TicketFly also got hacked, leaking personal info on 27M accounts. Eventbrite only managed to absorb 70% of creators from Ticketfly, significantly hurting return. Fortunately, management has acknowledged the acquisition was an operational failure, and said they would be more careful on big acquisitions like this going forward. They have yet to make any investment nearly as large as this. 

TicketScript was acquired by Eventbrite in 2017, several months before the Ticketfly acquisition, for $33.4M. TicketScript was a pure market entry play. TicketScript was the market dominant player in the Netherlands and Germany, and has had long experience in the market, being founded at the same time as Eventbrite. Eventbrite bought the firm for approx. 3x revenue. 

In 2018, Eventbrite made two smaller acquisitions: TicketTea for $11.4M and Picatic for CAD 1.8M. TicketTea was a leading Spanish ticketer. It helped Eventbrite enter the Spanish market but Eventbrite primarily bought the company to gain access to its engineers which were just as good and educated as those in San Fran but could be paid at much lower cost. Eventioz and TicketTea would eventually house most of Eventbrite’s engineers following their workforce cuts during the pandemic. The Picatic acquisition was purely for the features they offered. Picatic, based in Vancouver, offered a variety of features including online guest surveys, more sophisticated inventory management, crowdfunding, and comprehensive reporting.

The last acquisition Eventbrite has made so far was ToneDen for $7.5M in November 2020. ToneDen, based in Los Angeles, was a marketing dashboard tool that could integrate with social media and email. The acquisition provided the framework and the engineering capability for Eventbrite’s marketing suite.

Management

Eventbrite was founded by Julia and Kevin Hartz in 2006, who started the firm in their home soon after getting engaged and moving to the Bay area. Early adopters included small tech meetups and speed dating. The Hartzs spent time studying customers first-hand, adding more and more features ad-hoc. It was a scrappy operation with less than 10 employees. None of them knew how to build the tech, but Kevin had a strong network, bringing in Renaud Visage who became CTO. 

Eventbrite got their big break when Sequoia Capital invested $8M in 2009. Eventbrite then secured $50 million in Series E Financing in 2011, with Tiger Global as the primary investor. In 2013, the company raised an additional $60 million led by Tiger Global and T. Rowe Price. Subsequently, in 2014, Eventbrite secured $60 million in a private equity round. Finally, in September 2017, the company successfully raised $134 million in a Series G funding round, bringing their total funding to $334 million. The company then went public in 2018 at $23 a share, a market cap of $1.76B. Julia Hartz has served as CEO since 2016, replacing Kevin Hartz who now serves as Executive chairman.

Kevin (Executive chairman) has had a long history in the start-up scene and in platform businesses. An ex-Paypal employee, Kevin founded Xoom in 2001 and served as CEO until 2005, remaining as a Director up until its IPO and eventual acquisition by PayPal. Xoom was a remittance platform, whose framework echoed that of Eventbrite.  It had a payment processor which earned for Xoom by taking 1-3% of the exchange rate spread, and had some network effect: families and friends who were sending money to each other would see the Xoom logo when they got the payment, prompting the receiver to eventually become a sender themselves. Like Eventbrite, Xoom relied on reputation in sending money as efficiently and securely as possible. Kevin and the team grew Xoom at a compounded rate of nearly 40% since the end of the 2008 crash, turning a profit in 2015 on a EBITDA margin and cash flow basis, the same year Paypal bought it for $1.09B, almost 6x revenue. 

Julia Hartz (CEO) worked in show business before starting Eventbrite. In the early 2000s, Julia was an executive producing shows such as MTV’s hit series Jackass, FX’s The Shield, Nip/Tuck, Rescue Me, and 30 Days. Julia has had extensive experience dealing with actors, production managers, executives, and directors. Event creators share similar challenges with people in show business: both need to appease an audience and have to get enough of them to earn a profit. We think Julia’s time with entertainers surprisingly makes her an excellent fit for the event space. 

Kevin enjoys growing young companies and then sending them to college while Julia wanted to be a parent to the company. We saw this in Xoom where Kevin took a backseat after running the firm for 4 years, and now after 10 years of Kevin as CEO. Both have been described as savvy, caring, capable operators, and incredibly open-minded. One former employee told us Britelings could send an email to Kevin and Julia about a new idea and they would always have a conversation about it. They trusted their people, and were only hands-on when necessary. We got the feeling that Eventbrite really prioritized culture. When Eventbrite wanted to acquire Sympla, Kevin chose Eventioz because Eventioz people were close-knit and allowed pets in the office. 

Management is bold. The pivot to frequent creators is no easy decision, considering that it wiped out half their sales. Sources say Casey Winters orchestrated this. We don’t know if he was the one to suggest the shift, but he likely led its execution. His time at Eventbrite was a 3-year “gig” for him, but he evidently had a profound impact on Eventbrite’s current strategy. Employees also thought really highly of him. With Winters’s mark and Kevin’s decision to take a backseat, Eventbrite on a managerial level is quite different now compared to pre-pandemic. 

Insider ownership is good. Julia and Kevin hold 10% of the Eventbrite, worth around $54M. Through Xoom’s acquisition, which Kevin held 2% of, Kevin received $24M, which puts the couple’s Eventbrite stake at around 70% of their net worth. Kevin increased his stake 4 fold since the pandemic.

Capital allocation

Kevin and Julia are generally hands-off when it comes to capital allocation outside M&A. Since CFO Lanny Baker joined 2019, he has significantly improved return on capital. 1) In March, he approved a share repurchase plan that would take out a fifth of shares outstanding — very well-timed! 2) Since about half of their is short-term reimbursable to creators, they invest this mini-float in short-term T-Bills, adding extra return.

Valuation

Following the Q4 release, Eventbrite lost nearly 40% of its market cap. Eventbrite now trades at 10x EV / NTM AEBITDA and 9x EV / steady state EBITDA. Eventbrite suffered from an overpriced IPO, leading to an over 80% drop in share price since. We think the market has thrown the baby out with the bath water. With the large share buyback plan and churn not as strong as expected, it seems the doom and gloom is centered on a short-term outlook, and valuation should pick up in the long-term.

Looking Forward

The event ticketing space is quickly growing with a massive long tail, especially internationally where penetration is still low. Competitors are unable to compete with Eventbrite without subsidizing their tickets. Eventbrite has the best product, the best reputation, and the best audience. Existing customers don’t leave when they raise prices, and new customers always find Eventbrite first due to their superior SEO spend. We think buying Eventbrite is a chance to be part of that growth runway at a price where even if management misses targets, we still don’t lose money in the long-run.