Greenhaven Road Capital commentary for the third quarter ended September, 2021, discussing their largest position, Digital Turbine Inc (NASDAQ:APPS).
Dear Fellow Investors,
On the first page of every recent letter, I have noted that we will have down months, quarters, and years. Well, we just had a down quarter. The funds returned approximately -6% net for the third quarter, bringing YTD returns to approximately +14%. Please check your statements for actual numbers as they may vary by entity, investment date, and class. This may be of little solace to those like my Italian cousin who just joined the partnership, but since the Q1 2020 depths of the pandemic, we have had five straight quarters of positive performance returning well in excess of 250% over the past 15 months. We remain focused on the long term. Declines along the way are inevitable and sometimes offer buying opportunities. We are not trying to time the market; we are trying to generate attractive risk-adjusted returns over a multi-year period.
Summer Of APPS
When a top five holding starts the quarter at $76 and drops to $48 in a virtual straight line, questions should be asked. This was my July and August with Digital Turbine Inc (NASDAQ:APPS). If I liked the stock enough at $76 for it to be a top five holding, I should love it at $48, but Mr. Market was clearly flashing warning signs. While some of you may have enjoyed a carefree summer, I dove back into Digital Turbine to try and figure out what I might be missing and what the market might be missing. I did the best I could to parse every word and action of CEO Bill Stone, connected with dozens of investors, looked at every sell side report and model I could find, studied competitors, and spoke with former employees. There was such a wide gap between my perception of the company and the daily drubbing of its share price, and I was eager to understand if I was missing something or if there was actually a compelling buying opportunity in front of me. I went deep into the weeds, looking at what was said as well as what was unsaid. Over the summer and into the fall, a more detailed mosaic emerged, and I reached three conclusions.
The first is that, after making four acquisitions in a little more than a year, the company has many more moving parts and is very difficult to model at a granular level. Nobody has a robust multi-year model that they have any certainty in. This includes yours truly, who (via the funds) has a large percentage of his family’s net worth invested in APPS, as well as sell-side analysts and even firms that specialize in building financial models for public companies. Digital Turbine’s model has gone from requiring three inputs – phone activations, revenue per phone, and percentage of revenue shared with carriers – to requiring dozens in order to achieve any real detail. The acquisitions each have different drivers, and the company’s new complexity is compounded by the fact that the most recent acquisition closed in Q2. As a result, Digital Turbine has not yet issued one full quarter of reporting with all of the acquired companies included. As best as I can tell, very few investors are letting themselves imagine what the economics will look like if there actually is an industrial logic to these acquisitions – if the companies truly are better combined than as stand-alone entities. To be fair, management was not spoon-feeding investors and was cautious with their public statements, spending months restricting their comments/guidance to saying that operating margins will improve over time. They finally opened up a crack in September, acknowledging that they were tracking 13 areas of potential revenue synergies. The combination of complexity and management conservatism created uncertainty, and many investors sell uncertainty and ask questions later. There was plenty of selling this summer.
The second conclusion I reached is that Digital Turbine’s underlying business is extremely healthy. The core app discovery business was profitable last quarter and had organic revenue growth over 90%. The companies they acquired are growing revenue quickly, e.g. AdColony at +46% and Fyber at +198%. Even excluding the opportunities presented by the acquisition discussed below, there is the possibility of substantial future growth from increased penetration. Digital Turbine’s software is still on only 700 million phones, so there is a long runway for international growth with both carriers and manufacturers. Further, the company’s Mobile Posse division (a Q1 2020 acquisition) will be rolling out a new offering to Verizon and AT&T Android customers, and Samsung will be rolling out SingleTap on all devices worldwide. Despite what the share price was implying during Q3, I don’t believe the growth story is impaired.
The third conclusion that I reached was that the acquisitions were done to control the advertising transaction from end to end, allowing Digital Turbine to capitalize on both a distribution advantage and a data advantage. They bought AdColony for its relationships with large advertisers and Fyber for its inventory of ads, effectively adding demand and supply, respectively, to their digital advertising platform.
Digital Turbine’s distribution advantage is derived from a new offering called SingleTap, which allows an Android phone user to click on an app’s ad and download it in the background without being taken into the Google Play store. This patented offering leads to 2X-5X improved advertising efficiency. In the app world, this is massive, lowering the cost per install by 50% to 80% just by running ads with SingleTap embedded. What started as a $1M/quarter business is quickly approaching a $100M/year business. On an October investor call for Oppenheimer clients, management revealed that this $100M run-rate had been achieved with only 12 clients. A number of “last mile” technical issues have slowed growth and are being addressed, but the logic of taking the SingleTap technology and introducing it to AdColony clients to purchase ads served on Fyber’s networks is very interesting. Controlling the transaction and technology from end to end can yield superior outcomes for all parties. To date, none of the reported numbers reflect the opportunities presented by owning AdColony or Fyber, and there are reasons to believe that a technology that lowers acquisition costs by 50% will be used by more than 12 clients.
The second reason that Digital Turbine wants to control the digital advertising from end to end is to take fuller advantage of their data advantage in the marketplace. They do not publicly emphasize their data advantages, but because Digital Turbine software is on the phone and operates under the carrier or OEM’s user agreement, they have access to a lot of data that other digital advertisers lack. How much data and how they can use it is determined, in part, by each carrier or hardware manufacturer, but at a high level, they know the model of phone and when it was activated, the demographics of the owner, what apps are installed, and which and when they have been opened. In a marketplace where advertisers are paying for app installs and activations, these data advantages can yield significantly lower cost per activation and drive advertising dollars towards the Digital Turbine ecosystem. It will take time to integrate the data advantages into campaigns and sell those campaigns to AdColony clients to be run on Fyber ad networks, but the combination of SingleTap and a measurable data advantage for a growing client base could yield lollapalooza results.
It is easy to look and feel smart by being pessimistic and snarky, but optimism is usually more profitable. Buying more shares of a company that is up over 10X from the start of 2020 is not easy, but we did. I think Digital Turbine is in an enviable position with sound strategic rationale for controlling the digital advertising from end to end, pairing significant supply with significant demand, and layering in distribution and data advantages in a massive mobile advertising market where their overall market share is still small. On the left tail of negative outcomes, there are execution risk as well as risks associated with Google’s control of Android. At the same time, there are also massively positive potential outcomes such as the possibility that Digital Turbine eventually runs Verizon- or AT&T-branded app stores for the carriers. SingleTap could also be licensed to Snap or other large platforms. Again, SingleTap currently has only 12 clients. What will their revenue be with 100? When you layer in a below-market multiple with potentially dramatic sustained growth, the set-up is attractive. APPS’ price has rebounded from the lows of the summer, but I don’t think my summer was wasted. I now have a different (and potentially more informed) view of Digital Turbine’s distribution and data advantages than many in the market and believe that the market will gain a greater appreciation for the magnitude of the opportunity after the company’s analyst day in November. I think that the summer of 2021 will wind up being a lot more financially rewarding than the summers I cleaned pools.
Here is a link to a brief slide deck on APPS that I presented at VALUEx Vail. Given the limited time window to present, it does not get into the details of data, but I think it will still be informative.
Top 5 Holdings
Our top five holdings should be familiar to limited partners, as we have owned them all prior to this quarter. The largest position is Digital Turbine Inc (NASDAQ:APPS) – discussed at length above. Below are brief updates on the remaining four:
PAR Technology (PAR)
PAR Technology Corporation (NYSE:PAR) continues to make progress towards moving quick-service restaurant chains to a cloud-based point of sale (POS) system, positioning itself to take advantage of all the opportunities that creates, such as integrating payments, inventory management, and loyalty program apps. The company raised capital during the quarter, which some may view as a negative, but I thought was incrementally quite positive. CEO Savneet Singh’s excellent capital allocation decisions are core to our PAR thesis. So far, every time that he has raised capital, he has made an acquisition. I look forward to seeing what he buys. The other positive development in Q3 was the announcement that PAR’s defense business won a large contract that they had been waiting on. This non-core, non-strategic asset greatly muddles company reporting, and I believe the announcement increases the likelihood that the defense business will be sold, simplifying the story and giving Savneet more capital to invest.
KKR & Co Inc (NYSE:KKR) remains an extremely resilient business with an A+ team enjoying the secular tailwinds of the migration of investable dollars toward alternative assets, where large allocators like the returns and love the muted volatility.
Elastic Software (ESTC)
Share prices are up more than threefold since our first purchases of Elastic Software. Elastic NV (NYSE:ESTC) continues to report best-in-class net revenue retention (amount generated from existing customers) of 130%. With recent acquisitions, they are continuing their expansion into security. This is the company with the highest product velocity and largest addressable markets in our portfolio. With a massive base of customers using freemium/opensource products, there are fertile hunting grounds for growth.
During Q3, MarketWise Inc (NASDAQ:MKTW) reported their first quarter of earnings as a public company. I think it is fair to say that the market was underwhelmed. As of the writing of this letter, shares are down approximately 30% from our purchase price, which I thought was a fair one. As a reminder, MarketWise sells subscriptions to financial newsletters and related products. It is one of a very small handful of businesses that I know of that has grown to $500M+ in annual revenue with only $50,000 invested in the business to date. Gross margins are higher than most software companies at 86%, the company has been profitable all 20 years of operation, and revenue has grown for 18 of the 20 years. In the second quarter, they grew revenues 71%, generated over $50M in cash flow from operations, grew paid subscribers 45%, and grew free subscribers 75%.
MarketWise has many attributes we seek from the businesses that we own, including:
- High Insider Ownership: 92% of outstanding shares are owned by insiders
- Recurring Revenue: 90%+ customer retention
- Operating Leverage: Incremental subscriber additions are highly profitable
- Long Runway for Growth with No Additional Capital: Customers are added at less than 1/5 of their lifetime value and marketing spend is paid back in less than 9 months.
By most measures, this is an extremely healthy business, so why the decline? One can never be certain on this question, but my supposition is that the main driver has been the fact that MarketWise does not have a long history in the public markets or any publicly traded peers. Their forward guidance indicated a softening of demand as consumers focused elsewhere with the economy, recreational opportunities, and travel opening back up. These facts – combined with the market’s lack of familiarity with the business, its lack of publicly traded peers, the general apathy for SPACs, and overall uncertainty – led MarketWise to join a very long list of SPACs trading below their $10 IPO price.
My working theory is that the company will get better at communicating and the market will get more comfortable with the inputs of the business and the strength of its operating model. There is a large list of free subscribers to convert and a history of growth and operating profits. If growth and profits continue, there is little room for multiples to compress further, and thus we believe the price should rise over time.
The partnership remained short major indices and no individual companies. We have identified a couple of SPAC-related shorts that were not actionable; in one case there was no borrow available, and in another it was at the rate of 200% per year. While we continue to look for diamonds in the rough, rest assured, SPACs are still a fertile hunting ground for shorts.
I generally believe that if we own good businesses run by great management teams, we will do well over time. I tend to discount many of the macro themes of the day. For instance, in 2015 when Greece’s issues were roiling global financial markets, I took some solace in the fact that the GDP of Greece was one-quarter that of the state of Ohio and represented far less than 1% of revenue for any of the companies that we owned. As I look at the wall of worry that is facing investors today, most items, such as supply chain issues, seem temporary and isolated to specific industries. The one worry that is more insidious than supply chains and has a far broader reach than a country-specific issue is inflation. Could the hangover of printing trillions and trillions of dollars include inflation? Yes, there are certainly indications that it could.
When I look at our portfolio, I take some solace in the fact that most of the businesses that we own should be able to navigate an inflationary period well. KKR and Digital Turbine are not apparent beneficiaries of inflation, but they should be able to bear the environment as they operate with high margins, have no debt, don’t suffer from major labor costs, and lack long-term contracts where increases cannot be passed on. I also believe that my being right or wrong about my variant perceptions related to core holdings will have a larger impact than the CPI index movements. Thus, I am not building a bunker and we are not going to pivot immediately to gold, but of the thousands of variables out there in the cacophony of worry, inflation is the one I am focused on the most and it may influence our portfolio over time.
Just as I have ended many of our letters... as volatility arises, I will attempt to take advantage of the opportunities it creates. We will continue to invest with a long-time horizon, and we will continue to invest like it is our own money – because it is. Thank you for the opportunity to grow your family capital alongside mine.