EFTA00295552.pdf

ii0t , Iiii(117 jet". 4/10/2019 • • • Clifford Sosin CAS Investment Partners, LLC 8 Wright Street Westport, CT 06880 Performance Summary Sosin Partners, LP* Spy" 2012." 14.0% -1.5% 2013 66.6% 32.3% 2014 6.3% 13.5% 2015 14.5% 1.3% 2016 22.0% 12.0% 2017 31.2% 21.7% 2018 29.8% -4.6% 01 2019 20.3% 13.5% YTD 2019 20.3% 13.5% Cumulative return since inception 478.1% 121.1% Annualized return since inception 31.1% 13.0% See disdaimer regarding comparison to indicies at the end of this letter. * Performance net of 2% management fee and 20% performance allocation. ** Includes dividends reinvested. *** Sosin Partners LP launched 10/9/2012; e ormance or both the and and SPY shown rom that date. To My Partners: As shown in the table above, Sosin Partners, LP reported gains on a mark to market basis net of all fees, expenses and performance allocations of 20.3% during the three months ended March 31, 2019. The broad market as represented by the SPY ETF was up 13.5% including dividends during the period. Since its inception on October 9, 2012, Sosin Partners, LP has reported gains on a mark to market basis net of all fees, expenses and performance allocations of 478.1%; this represents a 31.1% compound annualized rate of return. The SPY ETF is up 121.1% including dividends during that period, representing a 13.0% compound annualized rate of return. Clifford Sosin • 1 EFTA00295552 The Balance Sheet We ended the quarter with six stock positions of any significance on the long side of the balance sheet. Our long holdings total 105% of equity capital. As previously discussed, our largest position at 37% of equity remains Carvana common stock. There are a few other small items on our balance sheet but none are material. Intellectual Conservatism and Risk From the uninitiated, I frequently get the feedback that our investment portfolio seems "risky." After all, we have only six holdings. For a long time our largest holding was a much maligned multi-level marketer (critics said pyramid scheme, but they were wrong). Then it was a subprime installment lender and now we have over a third of our capital invested in a money losing used car dealership! You won't be surprised to learn that I don't think our portfolio is nearly as "risky" as it appears. Instead, I believe that an unappreciated attribute of the portfolio — one that brings me a great deal of comfort as I have nearly all my net worth invested in the Partnership — is the intellectual conservatism of the analysis of our investments. Intellectual conservatism? Allow me to explain... Think about investing as a portfolio of probabilistic bets. We are betting on how companies will do over time, but one could equally imagine betting on sports teams or horses. There are two ways to lose on any given bet. First, one could suffer a loss due to a bad outcome from a known or properly assessed distribution of outcomes. Call this a Type 1 loss. Second, one could suffer a loss arising from improperly assessing the distribution of outcomes in the first place. Call this a Type 2 loss. Typically, investors try to manage the risk of loss by making "conservative investments" and investing in situations with "margin of safety." When investors refer to a "conservative investment," they are usually referring to making investments where the distribution of outcomes is such that the perceived probability and severity of loss is low. Clifford Sosin • 2 EFTA00295553 i f le •stet .u2r. When value investors (including me) refer to investing with a "margin of safety," they are usually referring to investments where the mean expected return of the distribution is high such that "things can come out a lot worse than expected" and the investment can still be successful. Clearly, "conservative" investments with a lot of "margin of safety" are very attractive, as the buyer stands to have a high return on average with a very low potential for loss. However, both "conservative" investing and "margin of safety" deal with the potential for Type 1 loss. No amount of "conservatism" or "margin of safety" can protect an investor if his/ her assessment of the distribution of potential outcomes is faulty, i.e. Type 2 loss. The spectacular losses incurred by AAA rated MBS investors during the Great Recession illustrate this point. The buyers of these securities were practicing "conservative investing" as they had consciously chosen to invest in securities which they believed had a very low potential of a Type 1 loss. However, this assessment of the distribution of outcomes proved to be highly faulty, leading to a substantial Type 2 loss. It's one thing to lose money when you are unlucky (a Type 1 loss). It is quite another to lose money when you didn't understand your bet (a Type 2 loss). Needless to say, I make avoiding Type 2 losses a priority. This is to say that I try to make investments where my understanding of the distribution of potential outcomes is unlikely to be meaningfully wrong. How though, can this lofty goal be accomplished? To answer this, we must step back and discuss how I assess the distribution of potential outcomes of an investment in the first place. I try to assess the distribution of outcomes of a social system (such as a company) in very much the same way scientists or engineers try to understand the distribution of outcomes of any system. I start with the facts. Facts represent the state of the system (world) as it is now. I then try to connect these facts with theories. Theories describe how the components of the system (parts of the world) interact with each other to determine how the system changes with time. My goal is to combine facts and theories into a simplified model of a company in its markets. To be clear, this is not an excel model or a financial model but an intellectual model which describes how the company might do under various circumstances. Finally, I use this intellectual model to run various simulations in my mind of how the world might work out, thereby generating some sense of the range of potential outcomes. Clifford Sosin • 3 EFTA00295554 O12aranulir .-' '165 II 'Sill a2ers With accurate facts and connected by sound theories, this method of analysis can make reliable predictions about the distribution of outcomes in the future. In the physical sciences, for example, if one were trying to predict the motion of a weight suspended by a spring, this method of reasoning works with precision. Unfortunately, we live in an extraordinarily vast and complex society. So when applied to social systems, this method of analysis is fraught with potential for error. Many "facts" are unknowable, uncertain, or worse yet, wrong! The theories are even worse. We must rely on a hodgepodge of inconsistent, incomplete and sometimes inaccurate theories that represent the state of the art in the social sciences. It is in this void of awareness and understanding where the potential for Type 2 losses can creep in. Fortunately, not all our facts are faulty, and some of our theories make reliable predictions under the right circumstances. And this is where I turn to intellectual conservatism. I compensate for our general epistemological poverty by attempting to invest only in situations where the critical facts are exceedingly well verified and the essential theories upon which the analysis depends are among the most reliable and permanent. Essentially, I try to limit our investments to situations where the irreducible ambiguity (if not the risk) of the situation is kept to a minimum. Call it swimming in the intellectual shallow end, or intellectual conservatism. Consider our investment in Carvana, which I described in detail in the October letter. The range of potential outcomes is in no small part predicated on a) consumers' willingness to buy a car from Carvana, b) Carvana's unit economics serving those consumers, and c) the potential for competition to reduce Carvana's ultimate share and margins. As to consumers' willingness to buy a car from Carvana, my expectation is based on 1) a theory of how consumers choose retailers and 2) a set of facts about Carvana's offer to consumers. The theory is that consumers choose where to shop based on some combination of price, selection, Clifford Sosin • 4 EFTA00295555 service, and convenience. A prediction of this theory is that a retailer who can offer a much better proposition along all four dimensions will eventually win a lot of customers. That this notion seems obvious is exactly the point. I am hanging my hat on a theory that is about as close to ground truth as any in the social sciences. I am practicing intellectual conservatism. As to the facts, i.e. how does Carvana's offer compare on the basis of price, selection, service, and convenience, there is also little doubt. It is possible to directly observe price and selection via web scraping. Service and convenience can be inferred by speaking with former employees, reading customer reviews, and discussing the matters with management. While nothing is certain, the quality of Carvana's offering is about as certain as things get — intellectual conservatism. On the second consideration, Carvana's unit economics serving customers, the matter is largely factual as it pertains to understanding the underlying economics of the business today, albeit assuming greater scale. This area received a great deal of attention in my diligence, as the unit economics are not immediately obvious based on GAAP accounting. However, here too it was possible to develop a great deal of confidence. I prepared estimates of the cost to serve customers from first principles (the cost to move a truck a mile, how far cars travel etc.). I compared my estimates with the recollections of former employees, with the statements of management, and ultimately, with the help of some disclosures made by the company last summer, to Carvana's economics in its earlier markets. I further compared my understanding of the economics to the economics of other car retailers and shared notes with other smart investors who had done similar work. In sum, while no one analysis is conclusive in an uncertain world, by triangulating across a variety of sources it was possible to develop an understanding of the unit economics, which (I believe) has a very small likelihood of being meaningfully wrong — intellectual conservatism. On the third consideration, the potential for competition to reduce Carvana's ultimate share and margins, I am drawing from a set of theories (each applicable based on a set of facts) about what makes a company difficult to compete with. From a theory perspective we are using the theories that: a) companies with large economies of scale can fend of competition, as their scale allows them to sustainably compete at levels which are unsustainable for entrants, Clifford Sosin • 5 EFTA00295556 Is •sist . asa. b) companies with long and steep learning curves (economies of skill) can fend off competition since the time and cost required for a competitor to get down the learning curve makes building a viable competitor uneconomic, c) companies that earn consumers' trust in markets where trust is important and uncommon can fend off competition as competitors need to provide even more value to lure customers away from a trusted source to a new and as of yet untrusted source, d) being far ahead of the competition (along with fostering an energized culture) facilitates attracting the best talent who can help to further the advantage, while being behind/ losing causes talent to erode, making it harder to catch up, and e) a-d above combine in a non-linear way to produce an even more formidable competitor than one might surmise from each of these effects on their own. While these theories and their relative importance may not be as self-evident as my theory about how consumers choose where to shop, they are, based on my general studies of microeconomics, retailers, businesses, and business history, nearly as dependable individually and quite dependable and powerful collectively. Also, while I am not providing the factual basis that justifies applying each of these theories here (for brevity), they are also well substantiated — intellectual conservatism. Using the facts and theories laid out above, we can now come up with a probability distribution of outcomes for Carvana's key long term drivers, and from those make estimates of Carvana's long term profitability and value. I can't say exactly whether Carvana's long term market share will be 5%, 20%, or 50%, but I can say that it is likely to be a lot higher than its current 0.5%. I can't say exactly what Carvana's margins will be at scale, but I can say that they will likely be quite healthy. I can't be sure that Carvana will ultimately be the only meaningful online retailer of used cars, but I can venture that a) it is pretty likely they will be and b) if they aren't, then the number of meaningful competitors will likely be very small - probably one, possibly two. Thus, we've arrived at a distribution of outcomes for Carvana as a business. While the potential outcomes within the distribution vary widely, the distribution itself is unlikely to be meaningfully wrong due to the intellectual conservatism in its underpinnings. We've arrived at this distribution by practicing intellectual conservatism, hopefully avoiding Type 2 errors. By comparing this distribution of outcomes to the purchase price, we can assess an investment in Carvana's stock. The investment has a high expected value (i.e. "margin of safety") and low potential for loss (i.e. "conservative"). From a portfolio perspective, the key determinants of the Clifford Sosin • 6 EFTA00295557 ii IP •stit wen investment (share, unit economics, competition) are not meaningfully correlated with our other bets (i.e. the other investments in the portfolio). So this is a good bet! The point, though, of this exercise is not to rehash Carvana as an investment, but instead to point out that while Carvana seems risky by many conventional measures (new company, long history of losses, etc.), the groundwork of the thesis is actually quite intellectually conservative. I am relying on simple, reliable theories supported by well demonstrated facts in producing my distribution of expected outcomes. Let me contrast this investment with another company I studied recently. This company provides technology to manage online grocery businesses, including extremely advanced robots and related software that allow for the operation of super-efficient, large grocery distribution centers. Proponents of this investment point to the company's blue chip customers, vast technological capabilities developed over many years, competent management team, and large opportunity. On these points, they are right. However, great as the company is, greatness today is not enough. Betting on this business over the next 5, 10, 15 years requires believing that the company's lead, particularly its lead in technology, will be sustained over many years. Will it be? I simply don't know. I have no reliable theory that predicts how relative technology gaps between businesses evolve over time in this sort of case. From what I've seen, sometimes these gaps endure, but other times they collapse, often quite suddenly. Many of my homework assignments as an undergraduate electrical engineering student required me to build what would have been absolutely cutting edge technology just 10-20 years earlier. Will this be the case for sophisticated robotics and warehouse management software? I can't gather facts that would alleviate this ambiguity. As far as I know, there is simply no way to compute reliable odds that the company's competitive advantage will endure. Formulating a view — any meaningful view — would seem to be very intellectually aggressive as it would assume away this fundamental and critical ambiguity. Unable to assess this bet, I moved on. I should be clear that the intellectually conservative analysis underpinning each of our positions does not mean that our investments will work. First, even if my assessment of the distribution of outcomes is correct, this does not preclude the potential for bad outcomes from within that distribution (Type 1 losses). Second, even what I believe to be intellectually conservative Clifford Sosin • 7 EFTA00295558 14,01 .d., analysis could prove to be faulty. I may not be as intellectually conservative as I think. So the risk of Type 2 losses remains, if hopefully attenuated. In spite of these risks, I think that, at the portfolio level, the combination of good bets assessed using intellectually conservative methods is actually a fairly low risk/ high expected return investment into which I am happy to invest nearly all of my wealth. Of course time will tell. Other I want to thank my friend Peter Blaustein in particular for his input/ assistance with this letter. We spent a fair bit of time discussing this letter and his input definitely improved it. Also, if you are huge fans of my writing, last year Peter and I authored a paper on bank equity capital reform. Immodestly, I would suggest that we figured out a way to equitize the banking system which would eliminate banking crises, reduce the frequency and severity of recessions, simplify the role of regulators, reduce the costs and risks to the public, and lower bank private capital costs. If this sounds like a panacea, we think it is! We've been trying to get someone (anyone) serious to help us to get our idea considered more widely, but to no avail. Introductions or feedback on this idea is most welcome. You can find it here: https://papers.s.sm.com/sol3/papers.cfm?abstract id=3216230 Also, I've been negligent in thanking people for their help with my letters over the years. I routinely circulate my letters to my employees Ginny Moore and Lance Ettus who provide comments, as well as to Jennifer Bloom from BTIG, who has provided detailed and thoughtful comments on every letter since I started the fund. They deserve thanks too. Administrative The Partnership's Amended and Restated Confidential Offering Memorandum (the "Offering Memorandum") requires that I disclose whether my investment in Sosin Partners, LP represents over 50% of my liquid net worth. I am pleased to say it does. In fact, it represents over 90% of my entire net worth. My family and I are invested right alongside you. In addition, the Offering Memorandum requires that I disclose whether I have any other significant income generating activities. I do not. The management of the Partnership and its related or parallel funds is my sole occupation and source of income. Clifford Sosin • 8 EFTA00295559 fln Conclusion I am excited for the prospects of our Partnership. While I expect our short term results will be volatile, I believe that profits over time will be worth the volatility. You'll recall that we endeavor to benefit from a virtuous cycle wherein: 1) our investors trust us and allow us the space necessary to focus on long term investment performance, 2) this long term focus allows us to make better investment decisions unclouded by short term considerations, 3) better long term investment decisions, in turn, (hopefully) allow us to produce better long term returns thus earning our investors' trust and restarting the cycle. With this in mind, we continue to look for additional partners who understand and support our approach to investing. The wrong partners, partners who focus on short term performance, will not be welcome. Prospective investors should review the Amended and Restated Confidential Offering Memorandum for more information. I appreciate your continued trust in me. As always, feel free to call or drop by the office if you'd like to that. Sincerely, 441c Clifford Sosin CAS Investment Partners, LLC Clifford Sosin • 9 EFTA00295560 4-- c s 111“, III .IIIIWI• The information contained in this report is intended for informational purposes only and is qualified in its entirety by the more detailed information contained in the Sosin Partners, LP offering memorandum (the "Offering Memorandum"). This report is not an offer to sell or a solicitation of an offer to purchase any investment product, which can only be made by the Offering Memorandum. An investment in the Partnership involves significant investment considerations and risks which are described in the Offering Memorandum. The material presented herein, which is provided for the exclusive use of the person who has been authorized to receive it, is for your private information. CAS Investment Partners, LLC is soliciting no action based upon it. It is based upon information which we consider reliable, but neither CAS Investment Partners LLC nor any of its managers or employees represents that it is accurate or complete, and it should not be relied upon as such. Performance information presented herein is historic and should not be taken as any indication of future performance. Among other things, growth of assets under management of CAS Investment Partners LLC may adversely affect its investment performance. Also, future investments will be made under different economic conditions and may be made in different securities using different investment strategies. The comparison of the Partnership's performance to a single market index is imperfect because the Partnership's portfolio may include the use of margin trading and other leverage and is not as diversified as the Standard and Poor's 500 Index or other indices. Due to the differences between the partnership's investment strategy and the methodology used to compute most indices, we caution potential investors that no indices are directly comparable to the results of the Partnership. Clifford Sosin • 10 EFTA00295561