Columbia Business School's Graham & Doddsville Features Scott Miller of Greenhaven Road Capital

October 1, 2018

Scott Miller formally launched long-biased value hedge fund Greenhaven Road Capital in 2011. Prior to founding Greenhaven, Mr. Miller was the Co-Founder and CFO/Chief ofStrategy of Acelero Learning, a Head Start education services company that has grown to over 1,200 employees. He was previously an Analyst at Litmus Capital, an Associate at NewSchools Venture Fund, and has further experience as a business owner-operator. Mr. Miller earned a B.A. in Political Science from the University of Pennsylvania, an M.B.A. from the Stanford University Graduate School of Business, and an 
G&D: What was the biggest lesson you learned at Litmus?

Scott Miller: I learned how important it is for your investor base to be aligned with your investing strategy. At Litmus, our two primary LPs had very generous liquidity terms – they could effectively pull their money whenever they wanted to.  When the world was blowing up in 2008 and 2009, and people were taking liquidity wherever they could find it, we became very concerned that the LPs would pull their capital. This shortened our investment horizon. I couldn’t present an idea that I thought would work in three to five years because we didn’t feel like we had three to five years. I started looking for ideas that could work very quickly, such as those focused on earnings beats and misses. The terms of our money at Litmus ultimately encouraged us to play a very difficult short-term game, a style I am not suited for. 

The way I set up and have grown Greenhaven Road was informed by my reactions to the Litmus experience. We don’t have “hot” money. Greenhaven Road has about140 LPs, a majority of whom are high net worth individuals such as portfolio managers and former portfolio managers. We do no outbound marketing, active follow-up, capital introduction, networking, or cold calling. People read the letters and can go on our website (www.greenhavenroad.com) to fill out a form requesting more information. They choose to invest because the thought process and philosophy resonate. It’s a very stable capital base – nearly all of our investors have subscribed to a three-year lock-up. The arrangement is ideal because the stickiness of our capital allows me to buy low liquidity names with confidence and minimal distractions. Many companies we own are devoidof short-term catalysts but are attractive long-term opportunities. I make no promises to my investors about short-term performance, instead focusing on longer-term returns and the power of compounding. The strategy and the capital base are well aligned.

G&D: Can you touch on your investing philosophy and if anything has changed since you’ve started investing?
Scott Miller: At my core, I’m a value investor. When I first started, I loved 50-cent dollars – situations where the valuation gap may close by the dollar declining in value – but as I’ve evolved, I’ve become more enamored with higher-quality companies that I can hold for longer. I’m attracted to network effects, two-sided marketplaces, and platform companies – these are the modern monopolies – as well as businesses with high insider ownership, recurring revenue, and operating leverage. 

Today, there are increasingly better tools out there that make it much easier to access and process information. Communities like Manual of Ideas, SumZero, Value Investor Club, Corner of Berkshire and Fairfax, and even Seeking Alpha have exponentially increased idea flow relative to when I started. Along with my greatly expanded personal network ofincredibly talented investors, these resources give me a torrent of idea flow to sift through – far more than was available when I first started out. 

In terms of sizing, my sweet spot is between 12 and 18 companies. If my mandate were to outperform the market by the most dramatic amount possible, I’d own one stock. But I wouldn’t sleep at night. This is still very concentrated by most standards, but given that I generally hold names over longer time periods, I can be extremely selective about where I actually deploy capital. 

Overall, I believe that having the right temperament is critical for investors, and I don’t think that changes over time. The ability to be comfortable with a divergent opinion, the ability to not panic, the ability to buy more if there is an overreaction to prices – I think you either have those qualities or you don’t.

G&D: In addition to having your traditional fund, you also manage a fund of funds. What is the Partners Fund?
Scott Miller: The Partners Fund is a boutique fund of funds focused on emerging managers whom I believe are talented, underappreciated, and wellpositioned for long-termsuccess. The data suggests that smaller managers outperform larger managers, but the dynamics of capital allocation make it so that some really talented portfolio managersare running peanuts and, for a variety of reasons, investors don’t want to take the headline risk. But, ironically, that’s where the opportunity is. I’ll take hungry over fat and happyany day. The Partners Fund invests in managers that are similar to Greenhaven Road in that they meet the following criteria: an investment committee of one, concentrated holdings, reasonable AUM, significant personal investment, original thinking, and a mindset where getting rich is not the point

Now, a fund of funds focused on small managers is not a very good business – charging a few basis points on relatively modest amounts of capital is not a great set-up. Thus, most funds of funds want scale. They need hundreds of millions of dollars to make the business work, which means they need to invest larger checks in larger managers. I didn’t launch the Partners Fund to make money on fees; I launched it to formalize my relationship with select other managers and give my LPs access to them as well.

For me, the Partners Fund allows me to collaborate frequently with whom I consider to be some of the most promising investors of my generation. If I get one good idea a year that finds its way into our main fund, that is time very well spent.

G&D: Are there synergies for idea generation from the Partners Fund?
Scott Miller: It was a combination of a glaring opportunity that I didn’t think others were seizing – investing with small managers in a fund of funds structure – an opportunity fordiversification for my family, and a potential source of ideas. It didn’t hurt that I thought it would be fun and interesting.

G&D: Why do you think the market for allocating to funds is skewed such that larger funds get the vast majority of the assets? 

Scott Miller: I often compare performance and quality of ideas against AUM and generally find a disconnect. I don’t believe allocation of assets to fund managers is as efficient as it should be. Part of the challenge is if you are a gatekeeper, the accepted strategy is to wait until the statistical evidence is incontrovertible – when the performance of a manager is so good for so long, and the fund has enough AUM. Nobody gets fired for investing with a billion dollar hedge fund. The incentives are to keep your job and not stick your neck out for an emerging manager. Interestingly, when people are investing their own personal capital, the calculation is different. For example, one of my LPs runs a multi-billion dollar family office. He is comfortable putting his money in the fund, but not that of the family he works for. Part of what we’re doing in the Partners Fund is acceptingsome of the risks that come with smaller managers in exchange for hopefully outsized returns. My diligence process for the Partners Fund is also much different than thatof many allocators. It is less quantitative and more focused on individual ideas and the entire thought process. I would much rather have three years of investor letters thanreturn statistics.

G&D: In general, what’s your research process like from sourcing ideas to making an investment?

Scott Miller: The research process depends on both the source of the idea and how close it is to something we’ve done in the past – effectively how much domain knowledge I have. In general, I’m trying to get comfortable with product, market, team, and execution risk. I look for certain attributes to filter ideas quickly. For example, I prefer high insiderownership, asset-light business models (even though Fiat Chrysler – which we own – is not asset-light) recurring revenue, expanding margins, and the potential for operatingleverage. 

The other piece is what Murray Stahl calls invisible companies – companies that don’t screen well, aren't necessarily telling their story well, and aren’t covered by analysts. In those cases, the research process is initiated by other people explaining the idea to me. Then it becomes, “What are the pieces I have to fill in?” 

G&D: What are some themes you’re seeing in the market today, and where are you finding opportunities?

Scott Miller: We’re nine years into a bull market – it’s expensive. However, I’ve found opportunities in companies that are under-monetizing either assets or transactions insome way. Under-monetized companies can be attractive because you can have earnings growth without a significant increase in capital spending or SG&A. Success is dependent on making tweaks to existing products or pricing. Additionally, fixing monetization generally has lower execution risk.

G&D: Your Etsy investment aligns with this under-monetization theme. Can you discuss your thesis there?
Scott Miller: Prior to this year, Etsy’s commission fee has been set at 3% of sales since the company’s launch. A big part of the stock’s appreciation this year was driven by thecompany’s decision to raise its commission fee from 3% to 5% with no expected decline in the number of sellers on the platform, as the 5% rate is still very competitive compared to alternatives. A potentially greater than 50% increase in revenue with no associated expense increase is a great setup. However, the increase in revenue won’t immediately drop to the bottom line, as Etsy will reinvest a significant portion of the revenue into building out the demand side of the platform.

At the end of the day, Etsy has a very attractive, niche business that can grow many multiples of where it is today. There is room in the world for an Amazon alternative. Last year, I needed an outfit for an ‘80s costume party. Naturally, I searched “’80s costume” on Amazon Prime and received my shipment in two days. It couldn’t have been more convenient. The only problem was that every other guy at the party did the same thing. One guy had the exact same costume as me, and I recognized the costumes of a dozen other people. I don’t shop on Etsy for every purchase, but if I want something special, I go there. They operate in huge verticals and have an asset-light model with real barriers to entry.

G&D: What are some other examples of under monetization?
Scott Miller: TripAdvisor only monetizes somewhere around 1% of their traffic. Now, some of that traffic they can’t monetize; for example, if a traveler searches for best places to take a hike in a specific destination. But TripAdvisor is working hard to get bookings – restaurants, museums, attractions – done directly on the site, so I think it’s quite likely they’ll succeed in increasing monetization rates with attractions. 

Like Etsy, they sit between consumers and businesses. It’s a very valuable resource, and TripAdvisor has an excellent ecosystem with the most downloaded travel app and the deepest reservoir of content related to travel. The question is: can they get more booking activity on their site? If they can improve the monetizationof their existing traffic, I think the investment will work out quite nicely. 
G&D: Is under-monetization the main theme in your portfolio?
Scott Miller: Since I have the vast majority of my life savings in the fund, my interests are highly aligned with my investors, so I don't want to only own companies that are under-monetizing. I see value in diversification across investment theses, market caps, and even geographies, but I do think under-monetization is one of the current main themes. Theopportunity for improved earnings without massive spending – taking what you have and just monetizing it – is attractive and carries less execution risk.
We own Scheid Vineyards which is a growing sum-of-the-parts story. It is a family-controlled wine company where the land value is worth 2x the share price. So, this is an example of a 50-cent dollar, but what’s interesting is that they are transitioning from selling grapes to selling their own branded products. They have gone from a standing start to selling 600,000 cases of finished goods per year. They have the capacity to produce approximately two million cases with minimal incremental capex. If they are successful in their continued path towards selling more branded products, the economics should work out well for shareholders. 

G&D: Fiat Chrysler is one of your top five positions. Can you walk us through your thesis?

Scott Miller: Let me start by saying this idea lacks some of the criteria I discussed before in terms of recurring revenue, but it does have high insider ownershipand operating leverage. I have owned Fiat Chrysler since it was just Fiat and traded only in Italy. Despite returning multiples of our original purchase price already, I stillbelieve it is very attractively valued. This is a company that has a portfolio of valuable brands yet is also a turnaround story with legs. They are making a very accretive shift in manufacturing capacity away from low margin Fiats and Lancias to high margin Jeeps and Alfa Romeos. The margins on a Fiat Panda are sub-5% while the margins on a Jeep Wrangler – although the company doesn’t disclose them – are probably around 35%. If you focus solely on car volume or top-line and don’t want to focus on the mix of what those cars are going to be, you miss a major part of the opportunity. Fiat Chrysler is reducing the low margin fleet business by getting out of sedans and focusing on SUVs,aligning themselves with customer preferences and higher margins. They are also going to either spin off or sell their parts division. If you back out the parts business, you're getting the core business for less than 3x earnings excluding net industrial cash and the parts business. That’s an attractive multiple for a growing earnings stream and a business that should remain profitable even if U.S. new car sales decline by 30%.

G&D: You mentioned in your last letter the importance of understanding the motivations of key actors. Can you expand on that in the context of Fiat Chrysler?
 
Scott Miller: A large part of this job is trying to put the puzzle together – trying to understand what the key players’ incentives and preferences are. Fiat Chrysler is controlled by the Agnelli family. Their holding company, Exor, holds 30% of the stock. John Elkann inherited the Fiat holding from his grandfather. However, the investments that John has initiated during his time as chairman of Exor are decidedly not industrial. He clearly prefers asset-light businesses as he has invested in media, reinsurance, and now start-ups. I don’t think Elkann ultimately wants to own Fiat Chrysler or its parts unit because it’s not a great business—it’s cyclical, and over the cycle it should have relatively low returns oncapital. I think spinning off the parts business makes it easier to sell the entire company.
Charlie Munger discussed that over a 40-year period, returns start to mirror return on invested capital, regardless of what you pay for the company.I think Elkann wants to reset what they’re invested in. He won’t give Fiat Chrysler away, but I suspect that over the next couple of years, Elkann will be out of the parts business and will sell the core auto business to another OEM so that he can redeploy the capital at higher rates. Keep in mind – combining large OEMs will yield enormous savings.

G&D: You mentioned that you really admired late Fiat Chrysler CEO Sergio Marchionne. What specifically did you like about him?

Scott Miller: To me, Sergio was like a five-tool player in baseball. Those are the guys that can hit for power, hit for average, run, throw, and field. At Fiat, he executed at the highest level. He led the acquisition of Chrysler without having largemerger experience. He spun off Ferrari and articulated a vision for increasing volumes and expanding margins. He created so much value over his time. Shareholders got over a 30x return. I also found his speaking on conference calls to be operatic. He was enthusiastic, dismissive, and honest in a way that felt authentic. You don't see that very often. He would be on my Mount Rushmore of CEOs.

G&D: Given your position in the auto industry, do you have any views on autonomous vehicles?
Scott Miller: I’m very skeptical about autonomous vehicles. People have this vision that eventually we are not going to own cars because there are going to be on-demand fleets. As it stands, autonomous vehicles don’t work except for in the most mundane conditions, such as in Arizona where there are no pedestrians, snow, or rain. In additional to the technical challenges, there are also economic challenges because the LIDAR (Light Detection and Ranging) components are not currently cost-effective. Assuming you solve thetechnical and financial issues, there are still regulatory hurdles, consumer preferences, and execution risk surrounding production. Last year, the former head of Waymo (Alphabet’s autonomous driving subsidiary) forecasted no legal autonomous level-five vehicles before 2030. 

I’m currently teaching my 16-year-old daughter how to drive. Believe me, I wish we had fully autonomous today. Fiat Chrysler will be sold long before the autonomous fleets are swarming our streets. 

G&D: Can you run through your Yelp thesis?

Scott Miller: The thesis on Yelp is straightforward. When Yelp started, their primary competitor was the Yellow Pages, which was sold on twelve-month contracts. Initially, Yelp matched this time frame and went to small businesses saying, “Don't advertise in the Yellow Pages. Give us $3,000 or $4,000 and advertise on Yelp for the year.” That's a fairly big ask depending on the size of the business. Eventually, Yelp started testing month-to-month and even day-to-day advertising and found that shorter, more flexible time periods returned greater customer lifetime value than annual contracts. This effectively reduced the cost and risk of a trial period and, not surprisingly, more businesses turned to Yelp. They knew that churn would go up, but bet that would ultimately be outweighed by the increased number of advertisers. Yelp has also started rolling out a non-term model this year. As penetration increases, I think there is an ongoing opportunity to radically increase both the number of advertisers as well as the lifetime value of those advertisers. They have a long runway.

G&D: That’s the type of thing a quant model would not pick up on, as you discussed in your letter, because the historical data doesn't give any indication that it’s occurring. How do you evaluate a business that’s growing intrinsic value per share that doesn’t show up in GAAP financials?

Scott Miller: In general, most of the companies we invest in have progress that doesn’t necessarily show up on the earnings line. We own a marketing automation softwarecompany SharpSpring that is acquiring customers at one-sixth of their lifetime value. I think the most rational way to operate that business is to acquire customers. The ROI on marketing is fantastic. As long as marketing efficiency does not deteriorate, they should acquire, acquire, acquire. In fact, I would borrow money to acquire customers, which is what they eventually did. However, this company will screen poorly on traditional value metrics. Earnings? They have none. Book value? Their main asset is their customer base, which is not valued on the books at all. But, there is enormous value in the customer base if their lifetime values are anywhere close to right. With these technology companies, I end up looking more at customer retention rates, net dollar retention, customer acquisition costs, and growth rates.

G&D: I notice you’ve been a lot more active on the long side in recent years. How do you allocate your time, in terms of idea sourcing, between long ideas and short ideas? Is there a systemic reason why you’re not as involved on the short side recently?

Scott Miller: We’ve historically had a long bias. Our longs can be 15%+ (as a percentage of the portfolio) positions. For an individual company short, we tend to take a 1% or 2%position. The upside on a long can hopefully be 5x, while the upside on a short is a double – at best – if it goes to zero. We end up spending a lot more time on the longs. I’m nottrying to be market neutral. 

G&D: Do you short indices? If yes, can you talk about why and how?

Scott Miller: There are a couple reasons why we may short indices. Sometimes we want to take risk off without triggering a tax event. We are heavily long-biased and own many positions with large embedded gains. It’s not a perfect hedge, but shorting indices versus selling and going to cash allows me to sleep better. We don’t take on a lot of leverage – we might be 110% long and 12% short. It’s pretty slim; it’s not our core business.

G&D: Any advice for students who are trying to get into the investment industry? How would you suggest they develop their investment philosophy?
Scott Miller: In my experience, the hiring process is very idiosyncratic, so I would not read too much into inevitable rejection. If you think you really like investing, the most important thing to do is to invest your personal account. You learn far more from owning stocks than anything else.

I would also say, quite frankly, that there are higher callings in the world. If investing doesn’t really compel you and keep you up at night and excite you, go do something else. There are a lot of ways to make money. If you’re just investing to make money, the guy who loves investing and is thinking about his portfolio while he’s in the shower and while he’s walking the dog is probably going to kick your ass. Only do it if it absorbs and compels you
G&D: Thank you so much for your time.
 
https://www8.gsb.columbia.edu/valueinvesting/sites/valueinvesting/files/Graham%20&%20Doddsville_Issue%2034_v22.pdf