With a rate hike inspired bear market and coming on the heels of a strong 2020 and 2021, 2022 had few winners for us. Only Datto (MSP) which was acquired earlier in the year, Tecnoglass, and Ensign were up for the year. Our largest negative contributor on the year was Goosehead primarily as a result of tremendously reduced profitability in the personal lines insurance industry. Goosehead was followed by Progyny, Armstrong World Industries, Xometry, Repay, Floor & Decor, Avalara, Allegiant Travel (which was exited early in the year) and Interactive Brokers. Although there are recurring themes, each holding has a unique recent history and outlook, and given all the volatility in the economy and market we think it would be a useful exercise to review each of the core holdings. This letter we’ll start with four of the most important holdings, and we’ll discuss the others in the next letter.

The environment in the skilled nursing industry continues to be very tough as a litany of hardships including increasing regulatory burdens, demanding payors including fast growing Medicare Advantage payers, a tight nursing labor market driving huge wage inflation, higher rents/higher financing costs, and the wind down of COVID support lifelines all compound miseries for subscale players. There are approximately 15,000 skilled nursing facilities in the country, and the median facility does not make money after rent. However, Ensign’s unique combination of hyper-focused and empowered local management at the facility level supported by the best practices, financial strength and scale of one of the industry’s largest players continues to pay rich dividends. For 20 years, Ensign has executed and refined a playbook designed to earn the trust of all the referral sources in their immediate community through great care, great outcomes, and leading systems for sharing data with those referral sources, which has in turn driven a client mix of higher acuity patients as opposed to filling beds with simple, low margin custodial work.

Impressively, Ensign managed to grow revenues and earnings in the mid-teens in 2022 despite being impacted to some extent by labor challenges which can require them to turn away business. In addition to the organic growth it generates in its existing facilities, Ensign grows by acquiring the businesses of small operators who cannot make the necessary investments to get the talent, systems, and scale needed to prosper. Ensign has been an opportunistic but disciplined acquirer for two decades with an unequaled track record. Acquisitions at distressed prices yield fantastic returns once Ensign installs its management and systems and builds that facility’s reputation in its local market. The cash-on-cash returns are so great, in fact, that since its IPO in 2007 Ensign has grown by 635% but has never needed to offer any additional equity and remains by far the strongest and most liquid player in the industry. Ensign’s opportunity set is expanding in unprecedented ways as a result of broader industry pressures, and 2023 and 2024 promise to be great years for Ensign as smaller players cry “uncle” and capitulate and sell out to a consolidator at very attractive prices. And with less than 2% of the industry, Ensign has ample room to grow. Longer term, there is also the interesting demographic development as the “silver tsunami” strikes and the baby boomer generation enters the prime age for needing Ensign’s services. The number of facilities in this country has been shrinking for years. A huge new and inexorable wave of demand without an easy way to accommodate it may lead to very interesting possibilities for Ensign and even turn a sleepy, consistent and profitable plodder into something of a glamour stock.

The status quo “good luck” strategy of going without a managed fertility benefit is clearly unsustainable and collapsing. And not just for reasons of equity. One by one, the leading companies in many industries have adopted Progyny, which causes all the dominoes to fall over time as other companies in each of those industries discover the challenges of recruiting scarce talent when a competitor enjoys the halo of a strong culture, evidenced by offering robust benefits including the one that is most important to a 30-something. After all, infertility affects one in five couples in America of child-bearing age. Progyny enables companies to do the right thing, attract and retain talent, produce vastly superior clinical outcomes with a program that largely pays for itself with the hard savings that come from reduced expense on the back end that would otherwise have to cover the costs of complicated, premature and multiple births. And with only single-digit penetration of its initial target market, there is no wonder why Progyny’s outlook is so strong.

Once the buyer approves the quote and an order is confirmed, Xometry queries its database of suppliers and offers that order to certain suppliers based on the quality of the match – basically how well this supplier has done before manufacturing similar parts and how likely this supplier will be able to make these parts at the lowest possible all-in price. If the first person doesn’t accept the order, they offer it to others and gradually increase what they are willing to pay for the piece, decreasing Xometry’s margin since the price at which they have to deliver it has already been set. This is brilliant, but it is unprecedented at scale, and impossible to refine without a few bumps along the way. Xometry has had great success getting better at the “match” as it has scaled – this is a big moat for any future competitors – and has seen enormous gross margin expansion over the last twelve quarters (from 18% to 27%) as a result. As a young company, Xometry has generally operated in an environment of stable or increasing prices and of stable to strong economic activity, which leads to stable or tight manufacturing capacity utilization. By the middle of Q3, the backdrop had suddenly and dramatically changed. The price of hot-rolled coil steel, for example, which had soared in 2021 fell over 40% from late May to mid-August. Costs to manufacture items went down, and Xometry’s matching algorithm got even better. Almost instantly, Xometry’s pricing algorithm adjusted, and pricing went down a lot. Customers were thrilled, and suppliers were happy to have good orders that kept their equipment humming. The sudden change of events, however, made for unhappy shareholders, including us, since a fixed margin on lower costs equals lower revenue and gross profit. We, in fact, have had “robust” conversations with the management over this algorithm not having old school, common business sense embedded in it. I know nothing about how to program an algorithm, but I know a fair amount about selling gasoline since I grew up in a family that owned convenience stores. As gasoline costs rise, you raise price quickly. As they fall, you take your time and max out a little margin opportunity. Xometry’s algorithm, instead, passed on all the benefits real time to the buyer, at its own expense. And based on our conversations with customers, this was completely unnecessary since customers buy from Xometry for many reasons other than price including breadth of offering, convenience, speed, and reliability. We should add that management wasn’t as obtuse as we make out above, and in fact we continue to hold them in very high regard. But this was new and happened quickly, and for better or worse, they had to gain a certain degree of experience before adjusting the pricing algorithm.

Q4 brought another new set of circumstances to train the algorithm, this time in the form of a sudden air pocket in order rates. In a newly disciplined financing environment, companies across the nation pivoted quickly to focus on profitability and cash generation for the year-end financials, and Xometry saw its quote-to-order conversion rate, which had risen meaningfully in Q3, slow suddenly and unprecedentedly. Rather than being completely passive, the company experimented to see if cutting price would stimulate conversion of quotes to orders. It didn’t. Pricing wasn’t the problem. People were simply pushing out future orders. In the end, overriding the algorithm by lowering price simply cost the company extra margin. All of this is frustrating to us, but the conversations with customers and suppliers have been more positive than ever. Counterintuitively, these issues should actually demonstrate and reinforce the strength of Xometry’s position. Q3’s issues speak to how eager suppliers are to get business from Xometry which in turn increases the value to customers and Xometry’s gross margin opportunity. In fact, the company has accelerated its target date for getting to optimal mature gross margins. Q4’s challenges, in turn, showed that the customers have bought into the full value offering and aren’t as price sensitive as might have been suspected. Taken together, these two quarters show how difficult it is to build and scale such an operation, which is why others have failed to gain any traction. We’ve spent a great deal of time with management since the snafus. We appreciate their candor and humility, but more importantly we are convinced that they are on the right path and building an extremely valuable business.

Which, at last, brings us to 2023. 2023 is also likely to be the tale of two halves. It will take some time for the company to get back to revenue per buyer levels seen in 1H 2022, but this metric is improving and there will be a nearer-term recovery of gross margin. The most important metric, however, is the consistent growth of active buyers. Active buyers are expected to grow north of 30% in 2023, which provides a larger base for Xometry to drive their land and expand strategy. Xometry offers the breadth of processes, materials, and finishes along with the depth of global suppliers to be a one-stop-shop for prototyping, small batch production, and production runs for the largest businesses globally.

We own Goosehead because we think they have the best mousetrap in a huge and stable industry and a long runway of growth with barely 0.5% share in its industry nationally. Nothing that happened in 2022 dissuades us from our belief that the next five years will see them make great progress towards their ultimate goal of being the largest agent in personal lines insurance, which would require them to pass State Farm which is more than 25x their size currently. Goosehead delivered a strong year with total written premium growing 42%, policies in force growing 27%, and margins expanding by 3.7%. And this was accomplished according to some measures in the least profitable year for the industry in decades as claims costs soared with inflation, increased driving, and weather events. For example, loss claims for GEICO expressed as a percent as premiums increased in an unprecedented way from 74% in 2020 to a staggering 93% in 2022.

As an agent, as opposed to a carrier who bears risk, Goosehead doesn’t get directly impacted by underwriting losses. Hence, while Allstate actually lost money in 2022, Goosehead had a remarkably positive year by most accounts. But serving as a leading distribution arm of an industry makes it difficult to be impervious to the downturn in that industry. Goosehead was affected in two primary ways. First, appreciate that most carriers operate in highly regulated markets which makes it an arduous and long process to raise price sufficiently to cover the new level of losses. And what carrier wants to grow when rates are inadequate? Accordingly, many carriers withdrew from writing new business, especially in certain important geographies like Florida and California, limiting the ability of Goosehead agents to flex the full advantage of choice which accrues to their robust independent model representing approximately 150 carriers nationally. The bigger impact on the year’s P&L came from dramatically lower contingency commissions which are paid to distributors like Goosehead based on the profitability of the business they bring as well as the growth that agent brings to the carrier. Contingent commissions are inherently volatile in any given year for reasons that are out of Goosehead’s control, but they are an important long-term source of profitability especially since they have no direct costs associated with them and fall straight to the bottom line. Given the backdrop of underwriting losses at the carriers, contingent commissions collapsed in 2022. Compounding the macro challenges was the slow down in real estate transactions. People tend to passively renew their insurance until they buy a new home, at which point they actively shop for a new policy. That passive renewal habit contributes to the great value of what Goosehead is building because it is essentially an ever-increasing annuity. However, more real estate transactions and more shopping for insurance typically favors Goosehead since they have the superior offering. In Texas, for example, they were the insurer in more than 15% of the housing transactions in the most recent quarter, which portends staggering growth for a company with only 0.6% share nationally. But not all of the challenges Goosehead had to overcome in 2022 were macro in nature.

For the second straight letter, the early days of CarMax can be illustrative. CarMax was created in Richmond, VA as a side project for electronics retailer Circuit City (RIP). The model was refined, a degree of infrastructure was in place, and management decided it was time to expand. It was an era of good feelings, cheap money and dodgy initial public offerings (dot com bubble of late 1990’s), and CarMax was able to float an offering of “tracking shares”. The success invariably attracted competition, most prominently Wayne Huizenga’s AutoNation USA. Huizenga’s appetite for and creativity around doing deals was truly incredible, and at times he was thought to have a Midas touch. He would be a formidable competitor, or at least one with access to tons of capital. Feeling the need to lock up big markets before AutoNation, CarMax engaged in a land rush buying up entry into a number of big, expensive markets, notably South Florida and Texas. CarMax had to make certain assumptions as they tweaked their midsize market model to large, urban markets. How much land did they need? How many cars did they need in inventory? How far would people drive to save $500 on a car? The battle for used car supremacy began, and soon it was clear that both were losing as result of rolling out untested business models. CarMax had guessed wrong. They overbuilt and lost money, overburdened with overhead. AutoNation’s situation was worse because they had been financial opportunists without any real and proven model. They capitulated in December 1999, closing all 23 locations and laying off 1,800 people. CarMax, on the other hand, quickly adapted, selling off excess real estate and shrinking inventory and changing the configuration of stores in major metros. CarMax’s stock sank as confidence was lost. Even its IPO underwriter dropped research coverage of them, but the changes were working. The numbers were obscured by the runoff of the costs associated with prior sins, but proforma for that adjustment the stores began making good money. And wonderfully, the competitive environment was infinitely better than it had been. AutoNation was gone, and it had become impossible for a new entrant to raise money for a used car only concept since it was now general knowledge that you could only make money as an auto retailer if you had both used and new cars, and service. The rest is history. CarMax expanded and was a 150 bagger off those lows as it grew unchallenged until the next era of good feelings, cheap money and dodgy IPOs spawned the next challenger and created a new level of investment required to stay ahead.

We tell the story of CarMax above because of its relevance to Goosehead Insurance. At the onset of the pandemic, Goosehead had proven the uniqueness and durability of their model in almost every state of the union. Carriers valued their distribution, their corporate and franchised agents valued their platform which enabled great careers, and their customers valued their savings, service, and professional advice. In 2019, they grew policies in force by 44% and were extremely profitable with better than 20% EBITDA margins. They justifiably harbored grand ambitions, even though they remained but a footnote in the industry rankings. And then the COVID shutdown happened. One of the beauties of this company is that people must insure their homes and cars at all times, even in a pandemic. And then the real estate market picked up as lifestyle changes and low rates drove transactions. And Goosehead was even better positioned as the ossified State Farm and Allstate models that require physical locations in strip malls became even more preposterous to agents. (When was the last time you or anyone you know stepped foot in one of those places? Yet, they remain a requisite financial millstone around the necks of those agents.) Amidst all the turmoil and new-found work force flexibility, torrents of new people from both inside and outside of the insurance industry sought to join Goosehead. An opportunity for a generational landgrab emerged, and Goosehead seized the day rapidly accelerating the pace of hiring both corporate and franchised agents. It was true that they could only interview and train people over Zoom, but it was a risk worth taking given the opportunity to quickly become a true national force. And for a while, it worked. In 2020, policies in force grew by an astounding 48%, and margins expanded. Policies in force grew by another 42% in 2021. Cracks, however, began to develop, although at first any issues were easily covered by surging contingent commissions as industry profitability soared amidst lower driving during COVID.