Carmax and the Struggle of Small Dealers


Circuit City is no more but its legacy lives on. CarMax was formed as a subsidiary of Circuit City in 1993. How and why an electronics retailer expanded into used car reselling is anyone’s guess, but as I will try to discuss at length, CarMax’s future will not follow the demise of its predecessor. Today, CarMax operates 209 stores in 40 states and has a presence in the whole car-buying value chain: wholesaling, retailing and financing. The used-car value chain can be described as such: the wholesale purchasing of vehicles from auctions by dealerships; the reconditioning of vehicles to prepare them for sale; the sale of the vehicle from the dealership to the customer; the market for “add-on’s” and other service plan for sold vehicles; and the financing of the vehicle.


Going down the list, the wholesale market is dominated by Adesa (owned by KAR Auction Services, a great company in its own respect) and Manheim (owned by conglomerate Cox Enterprises) who together control 70% of this $74 billion market. CarMax didn’t get into the wholesale market until later in its history, but it is the third largest wholesale auctioneer of used vehicles (~$2.2 billion in wholesale sales). Auctions are conducted on its own lots and primarily consist of cars deemed “not suitable” for sale to the consumer market. The consumer-facing dealership market is large and fragmented with the top five publicly traded players- CarMax (KMX), AutoNation (AN), Lithia Motors (LAD), Penske Automotive Group (PAG) and Carvana (CVNA)- controlling only ~$30bn of the $764 billion market. CarMax is the largest player in this segment with ~$15bn in used car revenue in FY2018; in other words, a ~2% market share. The other >90% of the market is controlled by small mom and pop dealerships with significant local presence but little national representation. The used car financing industry is dominated by big banks (most prominently Ally Financial) but dedicated car financers like Credit Acceptance Corporation (CACC) also exist (however, as will see later, CarMax doesn’t really compete with companies like CACC).

No company competes in all these segments. The first half- wholesaling, reconditioning and retailing- is extremely capital intensive but naturally lends itself to a few large national companies with exceptional economics like Adesa. The second half- selling extended service plans, aftermarket sales, origination and financing of purchases- is high margin (>70%) but requires a balance sheet and significant technical skills. CarMax is by and large the most diversified player of the bunch; in a non-traditional sense of the word, it can be characterized as “vertically integrated”, but more on that later. Normally, a car that retails for $20,000 in a dealership was purchased for $5,000-10,000 (depending on whether is was sourced from an auction or a trade-in) by the dealership and undergoes another $5,000-7,000 in reconditioning and ~$1,500 in SG&A. This leads to an average “gross profit per vehicle” of ~$1,500-2,500 depending on which company you’re talking about. On top of that, at the time of purchase, normally a dealership will have a contract with (1) Captive financial firms of OEM’s like Ford Finance, which exist to incentivize dealerships to purchase inventory, (2) Independent financial institutions like CACC and (3) traditional banks. Dealerships may also have contracts with outside/local contractors to recondition vehicles or service extended service contracts (for example, Carvana has an agreement with DriveTime, a private network of dealerships, to service its extended service contracts).

Unlike what we’ve seen in other segments of retail, this industry has not changed much in the past 30 years, but a few general trends are worth mentioning. First, while smaller dealers still control the vast majority of the market, this is changing. Larger companies like AutoNation and Lithia are rolling up mom & pop stores; the vast majority of mom & pop dealers are family owned and younger generations are less inclined to take over the family business; and the market share of the top five companies, albeit small, has doubled over the past decade especially with the roll-out of national advertising by companies like CarMax and, more recently, Carvana. Second, the bang-for-buck of a used car relative to a new car has increased dramatically, and growth in used car sales are expected to surpass new car sales for the next decade. The average price of a used vehicle (~$20,000) is ~46% less than that of a new vehicle (~$37,000) with little compromise in quality. According to a conversation I had with an AutoNation representative, “A 5-year old car in 2019 has backup cameras, seat warmers, rearview mirrors, assistive control, telematics, navigation systems… a 5-year old car today is a luxury car in 2010 or 2005.” For the publicly traded dealers who deal both new and used cars, the fact that used car revenue made up an increasingly large portion of their total revenue over the past five years attests to this development. Finally, Carvana is a revolution in itself and this deserves its own paragraph.

As the company makes very clear in its sales-markety SEC filings, 97% of consumers today do online research before purchasing a car, 81% do not enjoy the car buying process and 75% would considering purchasing a car online. Enter Carvana: the online-only “Amazon of car buying” with $3.7bn in annual run-rate used car revenue (up from $150 million in 2015). And as the saying goes, monkey see, monkey do: CarMax rolled out an online-only platform in Atlanta with one-day delivery earlier this year with a national roll-out by 2021; AutoNation made a $50mm investment in Vroom and Lithia made a $54mm investment in Shift Technologies, both online-only car retailers. The economics of online car-buying are similar to that of brick and mortar regardless what Carvana’s investor relations team tries to propagate. Yes, you do not need to maintain facilities, but facilities were always cheap to maintain- the vast majority of a dealership is pavement and it requires relatively few employees to maintain. Given that the capabilities have existed for some time, one is hard-pressed to understand why companies like AutoNation and CarMax didn’t invest in online capabilities earlier if economics of online car retailing allowed for a gross profit per vehicle of ~$3,000 (the number touted by Carvana’s IR team) vs. the current industry average of ~$1,700.  Carvana’s current SG&A per vehicle sold is ~$4,000. Assuming this expense is perfectly scalable (which it isn’t) and revenue continues growing at triple digits for the next five years (which it won’t), at CarMax’s scale (~$15bn in used car sales), Carvana’s SG&A per vehicle will be ~$1,111 ($4,000*$4bn/$15bn) vs. ~$1,361 for CarMax, and this is an extremely unrealistic scenario.

Why CarMax Wins- The Hard Stuff

Why are we using gross profit per vehicle when comparing these companies? In the car-dealership industry, the name of the game is inventory turns. Cars are expensive to buy, recondition, maintain on a lot and sell; essentially, the longer a company holds on to the car, the drag on returns is exponential on a per car basis. Thus, the name of the game is twofold: (1) how quickly can we take the car through the value chain and (2) how much value can we extract per car. CarMax wins on both of these fronts. Let’s start with the wholesale side.

Companies can source cars in two ways: from an auction or from a trade-in by a customer. The latter method is much more profitable than sourcing through auctions because trade-in’s typically garner a lower price and there is no shipping required. In a sense, the consumer subsidized the shipping cost by driving to the dealership and has little bargaining power over the dealer relative to the bargaining power of, say, Adesa. So, being able to source from your customers is a huge advantage, and CarMax sources anywhere between 38-52% of cars from customers in a given year which is larger than every other publicly traded used car dealership. Why can’t the other players do this? They don’t have the scale to generate this percentage of sourcing: the next largest competitor, AutoNation, is a third of the size and also deals in the new cars market. AutoNation additionally does not have a wholesale segment and is thus not free to purchase any vehicle brought by potential customers. And for each car that CarMax sources, it sells more quickly than its industry peers (50 days in inventory vs. industry average of ~66):

Additionally, CarMax is the only “vertically integrated” national dealership in America (using my admittedly broad definition of the term).  Over the life of a given car, therefore, CarMax earns more revenue dollars while keeping costs relatively fixed. On the cost side, CarMax spends similar amounts to, say, AutoNation (for cars sourced through auctions). Both incur similar reconditioning and general SG&A expenses, and both employ a similar workforce in the dealerships; however, CarMax, using these same resources, has the infrastructure in place to finance a car and the optionality to auction it off given low customer demand/lack of salability. In fact, over the life a purchase of a car with a price of $20,000, CarMax will generate ~$3,000 incremental revenue and $500-700 incremental gross profit with minimal SG&A support. With smaller competitors like Lithia Motors and Carvana, the difference is even more pronounced (reaching ~$1,000 incremental gross profit per vehicle vs. Carvana).

Despite these advantages, CarMax doesn’t look that profitable (note Carvana is excluded because it is not yet profitable):

During this period, CarMax’s average return on assets (~7%) actually lags the industry average (~7.9%). Why is this the case? Remember how CarMax also finances its vehicle sales in house? It does this through a captive financial unit called CarMax Auto Finance (CAF). Normally, a dealer will sell a car, originate financing and then sell that loan to a third-party financial institution for a small commission (~3% of the loan face value). In CarMax’s case, that loan remains in-house, that is, it is originated and serviced by CAF. Because of this, there are two things clouding this picture: (1) CarMax has a huge loan book (~$13bn of $20bn in total assets) that is earning ~4% interest every year and dragging down CarMax’s dealership ROA and (2) Other dealerships are selling these loans for a commission, both realizing 100% margin income without having to book an asset on their balance sheets. On a fully apples-to-apples basis, if we wanted to compare the relative profitability of these companies’ dealership operations, we must adjust CarMax’s financials assuming it sells its loans upon origin for a similar commission as its competitors:

This gives us a very different picture. The way I view CarMax is a holding company with two distinct operating subsidiaries: a bank that finances the sales of the dealership operator. As such, they ought to be separated when comparing CarMax to its competitors that operate in only one of these industry verticals. But this treatment also has cash flow implications which is what makes CarMax such an interesting case study. How are these loans funded? Through the issuance of non-recourse debt. That is, CarMax will securitize a pool of financed car purchases and sell this security to institutions who don’t have recourse to assets outside of the securitized pool; the spread on these securitizations is amazing (assets earning ~4% vs. interest payments of ~1%) but more on that later. Two things are peculiar about this treatment: (1) the change in finance receivables is treated as a working capital item and (2) the funding for these receivables is treated as a financing activity. As a result, CarMax doesn’t screen well and traditional profitability and valuation metrics are skewed (we’ve seen how this occurs with ROA, but this leads to an inflated EV/EBITDA ratio and an obscenely low FCF yield).

Finally, and this is a more nuanced point, comparing growth rates between KMX and competitors as stated is another futile exercise. The counterintuitive point is that the nature of the accounting is such that as long as the industry is growing (as it has been for the past decade), its growth rates will lag that of competitors. When a company like AutoNation sells a car, it automatically gets to book the commission it earns on the sale of the loan as income. CarMax, on the other hand, must recognize finance receivables over time. The classic revenue recognition issue. As a result, in a period with growing car sales, companies that sell their loans (every publicly traded car dealership besides CarMax) will enjoy higher growth rates. On the other hand, in periods of declining car sales, a company like CarMax will get “hit” less severely because of the more “conservative” nature of accounting. To see this, lets refer back to a previous graph:

In other words, a period with high growth in Auto Loan Receivables (that is, a period with many car sales) will correspond with lower growth in operating income in the short run due to the deferral of finance revenue (vs. other dealerships who will recognize this income at the point-of-sale). Over the long run, however, growth in operating income should correspond to growth in auto loan receivables (unless we assume interest rates on these loans tend towards zero, a remote possibility):

Moral of the story: CarMax Auto Finance distorts the profitability of KMX’s dealership operations, it distorts growth rates and profitability/valuation metrics, it results in a GAAP misrepresentation of KMX’s actual free cash flow and it prevents CarMax from being able to be compared to competitors on an apples-to-apples basis. But CAF itself is a wonderful business. Firstly, it has captive customers (people are already in the process of buying a car and CAF financing is the first thing they see). Secondly, it has high margin revenue (CAF’s operating margin is >75% and CAF income averages ~4% on receivables while the financing for it has a cost of ~1%) and predictable/recurring revenue (much easier to predict cash flows of a run-off credit book than car sales for the next decade). Thirdly, it is, by far, the safest loan portfolio in the industry (average credit score of 705 and CarMax divides its customers into three risk groups, and CAF selectively only finances the cream of the crop- more on this later). CAF incurs little incremental spending and capital investment per customer: since 2006, CAF operating expenses have averaged 0.46% of revenue and 4.5% of total SG&A while comprising ~25% of operating income. Finally, it cannot be replicated by any of CarMax’s competitors. Other dealerships simply do not have the scale to justify investment in risk management infrastructure and an in-house financing program. Moreover, CarMax’s unique singular focus on used vehicles implies more conformity across its national customer base, allowing its risk management program to be easily replicated to new stores. To give an example, AutoNation owns both Ford and Mercedez dealerships; naturally, the credit profiles of customers in each of these dealerships is radically different and difficult to aggregate into one risk management system, a precursor to an effective origination and servicing financing program. According to the AutoNation representative I spoke to, “CarMax is unique because of the volume they have and a very similar credit profile per person… being able to underwrite on [our] scale is very difficult.”

Why CarMax Wins- The Soft Stuff

People hate buying cars, but CarMax has made the experience better than most other publicly traded used car dealerships. They introduced the concept of a “megastore”, a concept very similar to the Walmart model applied to used cars. Compared to other competitors, CarMax locations offer a much larger selection with the ability to ship any car on the company’s website to any dealership within two days. Moreover, by 2021, the company expects to roll-out nationally its online-only model to complement its physical location. Next, while other companies have expanded horizontally into new cars, CarMax has stayed true to its core competency, that is, used cars. This has both allowed it to become more recession-proof than its competitors but also improve the customer experience within its stores. Salespeople at CarMax are the only ones in the industry that are paid a fixed commission regardless of what cars they sell (contrary to salesmen in other dealerships that are incentivized to sell customers more expensive cars with many perks). Overall, customers report higher satisfaction on every measure when shopping at CarMax relative to its other competitors (specifically, Carvana and AutoNation).

This is extremely unique. CarMax wasn’t founded by fancy/seasoned businessmen who knew how to maximize customer value. In fact, while they had a 3-year head start, their largest competitor in the early years, AutoNation, was founded by none other than Wayne Huizenga. In 1962, Huizenga borrowed $5,000 and founded Southern Sanitation Services with one garbage truck. That small company later became the renowned Waste Management (WM). In 1987, he bought a handful of stores that later became the largest movie-rental retailer in the United States: Blockbuster Video. Rest assured, CarMax was going against one of the greatest businessmen of modern day. However, while both were rapidly growing, AutoNation purchased other dealerships, very similar to what he did when consolidating the garbage and movie-rental industries. This is also similar to the growth strategy employed by many present-day national car dealerships like Lithia. Alas, each dealership had its own systems, culture and people that were difficult to integrate into a national company; AutoNation found itself in the new car and used car business, owning luxury and economy dealerships alike and dealing with significantly different customers depending on which region you were talking about. CarMax employed a very different strategy that ran counter to the general trend of consolidating mom & pop shops within the industry: all its dealerships are CarMax branded and they were all built from scratch. While this required larger capital investment than AutoNation (and AutoNation grew much faster in the early years through the roll-up strategy, at one point being the largest used car retailer in the United States before CarMax took the lead), CarMax was able to retain its culture, offer a similar buying experience for customers coast-to-coast and ultimately develop the capabilities to retain an in-house financing unit, CAF. After witnessing CarMax’s success with the megastore concept (so many companies early on decided against the megastore because it required much more inventory on hand, and as I’ve discussed, cars are the worst kind of inventory- they didn’t think about the customer experience, it seems) many competitors attempted to replicate it. AutoNation tried the concept- it was shut down after reaching ~$1bn in revenues. Ausbury Automotive Group (ABG) tried it with their “Q Auto” stores and Sonic Automotive (SAH) had an “EchoStore”; both of these concepts were shut down soon after their inception.

It is only with the megastore and customer homogeneity when an auto dealership can create a successful financing business like CAF- and what a business! On the soft side, CarMax has an embedded “first right of refusal” in its financing business, only offering in-house financing to its most creditworthy customers. As previously mentioned, CAF divides its customers into three segments based on perceived risk: the top tier gets CAF financing, the second is sold to traditional banks and the third to specialized subprime lenders. Of those who receive a financing offer from CAF, 95% accept despite having three days to find a competing offer. Moreover, institutions must yield precedence to CAF to get access to the large and growing “left-over” market of second and third tier customers who didn’t meet CAF’s strict underwriting criteria. In this sense, CarMax isn’t dependent on credit institutions; they are dependent on CarMax. An indirect result of this relationship is that, relative to competition, CarMax less prone to vacillations in the availability of credit, allowing it to continue to sell vehicles even when banks pull away from the market in recessionary periods. To give an example, in the Great Recession, CarMax experienced a 5% drop in used vehicle sales vs. ~15% for AutoNation, in large part due to its ability to continue to access internal CAF financing. Finally, on the securitization side, the market is very confident in “CarMax branded securitizations”. According to an S&P Credit Analyst covering the CarMax securitizations, the name CarMax itself garners a lower interest rates because of its historical reliability and recognition in the market (this is not a direct quote and was said in an unofficial capacity).

CarMax vs. Carvana

Given the disruptive nature of Carvana’s business, its valuation and the lofty estimates of its IR team, I think it is valuable to spend some time on how this competitive dynamic may change the future of this industry. Carvana’s promotional materials, in my opinion, are directed towards those not knowledgeable about the industry, and a thorough understanding of the economics of these companies on a per-car basis does reveal a “cap” on how much gross profit per used car can be generated (not including high margin add-on or financing revenues). There is no better place to study the dynamic between CarMax and Carvana than Atlanta, Georgia. Atlanta is one of the largest used car dealership markets in the United States and one of CarMax’s largest markets. Atlanta has been transformed into the first market featuring omnichannel and online-only models. In Atlanta, Carvana has grown market share from 0.13% in 2013 to over 3% today. To begin, Carvana’s 2.94% market share, while registering impressive growth, pales in comparison to CarMax’s >15% market share. Furthermore, according to CarMax representatives, Carvana has not stolen any market share from CarMax in this market. Carvana’s IR representatives admit that the primary drivers of market share growth were at the expense of independently owned dealerships, with “there being enough space for [CarMax] and [Carvana]” to grow and how they’re “really competing against… the [independent dealerships] that are left”.To counter Carvana, CarMax introduced its own hybrid-online model in December 2018. CarMax’s model is superior to that of Carvana’s in many ways:

  1. CarMax offers same-day free shipping from a dealership to a person’s house
  2. CarMax registers an inventory of 55,000 vehicles to view vs. Carvana’s 15,000 vehicles
  3. CarMax’s on-site or in-house sales associate is on-call for assistance and in-home test driving

The second and third points are especially important. Everyone knows the psychology of test-driving a car before paying for it. Moreover, a car is usually a person’s second largest purchase after a home; most people would want to physically see a car before committing to a purchase. To be fair to Carvana, the company does offer a 2-day return policy but delivering a car to a customer and picking it up as a return is prohibitively expensive and I suspect Carvana easily incurs a >$1,000 gross profit loss for each return.

To understand Carvana, it helps to compare the industry to the airline industry. CarMax is more of a Southwest, with many stores similarly situated and close-by while Carvana is a traditional United, that is, it has centralized warehouses (hubs) and delivers cars to customers from these hubs (in case you were wondering, the vending machines serve more of a marketing purpose than a business purpose). You can imagine how the economics of these two systems would compare: Carvana does save on having a smaller store footprint but it trades these lower operating costs with higher delivery costs. In other words, Carvana incurs an incremental ~$700-800 cost per car in the form of higher delivery fees relative to CarMax. Moreover, Carvana doesn’t get the benefit of higher margin customer trade-in services. In fact, one can argue that it is more expensive to pick up cars from a residence from trade-ins. Although, Carvana would never tell you this- they claim that the 30% of inventory that they source from customers has a higher margin (the actual numbers are difficult to obtain from their financials).

But Carvana’s growth doesn’t lie: the company has grown at a CAGR of over 100% for the past 4 years. Moreover, its market share in key markets has increased substantially (albeit not at CarMax’s expense) while being supported whole-heartedly by equity markets. However, much of this growth mirrors that of AutoNation back in the late 1990’s and early 2000’s. Let me explain: Carvana requires almost no capex to grow into additional markets because of a “Master Service Agreement” with DriveTime. DriveTime is, not coincidentally, owned by Carvana’s CEO’s father. Currently, Carvana conducts the majority of its reconditioning and maintenance in DriveTime facilities. However, once Carvana outgrows DriveTime (based on current growth rates, this should take about a year), it will need to build its own reconditioning facilities (IRC’s) in new markets, an initial investment much larger than the current $500,000 touted by CVNA’s IR. Finally, Carvana needs to expand by a factor of ~4 in order to reach the sales penetration of CarMax; even using the loftiest growth estimates, this should take at least 2 years. During this time, CarMax has the ability to replicate Carvana’s model nationwide. No matter how you look at it, qualitatively it is difficult to see the path Carvana will take to usurp CarMax from the throne. Relative to CarMax, Carvana is more vulnerable to economic changes (it cannot yet fund its own operations and is increasingly reliant on equity markets, and bank commitments for its financed purchases), less vertically integrated and doesn’t possess anything special. CarMax will also roll out its online-only Atlanta model, which is just as good as Carvana’s, nationwide by February 2020. As put succinctly by an AutoNation representative, “AutoNation and CarMax can replicate Carvana’s model by simply buying 2 or 3 pick-up trucks and offering free one-day delivery. It’s not that revolutionary.”


I first began looking at this company when it was trading in the mid-to-high $60’s. I didn’t know the price at the time, but here’s what was going through my head.

CarMax is a company that can invest capital at a much higher rate of return than its competitors. Its dealership operations are generating a return of ~14%. On top of that, the company has a loan book of ~$13bn that is generating 3-4% in net interest. The latter business requires almost no equity investment resulting in a consolidated pre-tax ROE of ~30-40% depending on what year you’re looking at. This is better than every other used car dealership out there. So how does CarMax look at saturation, that is, when it can’t reinvest capital at such high rates of return?

This is a hard question. 20 years ago, people thought dealerships could never be consolidated into national companies. In the early 2000’s, CarMax itself estimated the United States TAM as 200-300 stores, yet the company has 209 stores growing at 10-12% a year with no sign of slowing down. Management itself recently admitted that  they “significantly underestimated” the United States market opportunity. Today, CarMax controls ~2% of the market, a small percentage but a substantial increase from the <0.5% 15 years ago. In addition, this 2% doesn’t do the company justice because it is still growing. It has an established presence in only 16 states, while growing in another 24 states and no presence in 10 states. Moreover, not all markets are created equally. Downtown LA isn’t the same as a small town in Idaho. NY Metro isn’t the same as Austin, Texas. For a long time, in fact, CarMax couldn’t penetrate smaller markets with its megastore concept- there was simply too little demand. As a result, it introduced smaller-format stores in 2015 that carry less inventory (they’ve been very successful to date and contribute to about 30-50% of store openings in a given year).

How does the market share look like in more local areas? Well, the economic reality is that CarMax enjoys a 4.5% (and growing) market share in 98 of 210 television markets in which it has a presence and 15+% market share in its most established zip codes. The “saturation point” should be somewhere in the middle, perhaps around 7-10% (I’ll be the first to admit this is a wide range for the largest retail vertical in the world but such is the nature of valuation). CarMax still does not have a presence in the NY Metropolitan, Pittsburg and Detroit, all of which can support >4 stores. Medium markets, which make up about 32% of untapped markets, can support 2-3 stores and small markets (64% of untapped markets) can support one smaller-format store. If we use these as our estimates, the actual addressable market in the United States is ~360-400 stores.

If each store sells 4,000-5,000 used vehicles for an average price of $20,000 and 2,000-3,000 wholesale vehicles at an average price of $6,000 total vehicle revenue should be on the order of $33.1-$42.5bn. At the current gross margin of 18%, gross profit should be on the order of $6-7.7bn and at SG&A as 9% of revenue (10-year average), operating profit of dealership operations should be on the order of $3-3.9bn. What is CAF income? About 45% of customers fall in the tier one category, and of this 45%, 95% will accept financing. Therefore on 1.6m sold vehicles (360 stores*4,500 average per store), ~700,000 will be financed by CAF. Average LTV is 95%, so total originations are $13.3bn (700,000 financed*$20,000 average selling price*95% LTV) and total loan book will be ~$26.6bn (loan book has always been about 2x yearly originations). From this, we can derive net CAF income (after interest expense), which runs about 3% of the total loan book, or ~$800m ($26.6*3%). Therefore, our total pre-tax income comes out to $3.8-$4.7bn in this scenario. These estimates seem very reasonable, but it is important to understand that not even the most knowledgeable industry participant will get this right. This total market in the United States may come out to be much larger if the consumer preference towards used vehicles intensifies. Or, the online model might prove revolutionary at scale, although I believe this is a remote possibility. CarMax may start a roll-up strategy which can dramatically change the 14% ROIC we are seeing today in either direction (management has clearly stated they don’t plan to use this as a growth strategy, but management has changed hands many times in the past). The international market might be up for grabs after 360 stores- to my knowledge, I am unaware of another similarly sized company with a large foothold in any international market.

Despite these uncertainties, let us take this midpoint $4.3bn figure for pre-tax income. How long will it take to get there? CarMax is adding 15-20 dealerships a year, and at this rate, it will reach 360 within 7-10 years. Of course, there is reason to believe this increase will intensify as the company achieves more scale. At this earnings figure and future growth profile, it is reasonable to assume at the time that the company will be trading at ~$50 billion, or ~$284 per share based on the company’s current diluted share count. Of course, during this time, the company will be spending ~$600m-1bn a year buying back stock, as its been doing for the past several years. If the company buys back shares on the order of ~2% a year (currently running at ~3.5% per annum), share count will decrease by ~22% in 10 years, so our per share valuation will be about $346 which comes out to a 14.7% annual return over 10 years if our assumptions are realized (based on a current share price of ~$88. Seems really high but the assumptions of operating income are entirely reasonable. If the company is instead trading at a $40 billion valuation, our annual return drops down to 12.2%, still a very formidable figure.

Another way to look at the situation is as follows: the analysis here shows that KMX is better than its competitors in terms of profitability, growth and competitive position within the industry and therefore should trade at a higher multiple:

Does this mean everything is priced into the market already and our analysis was incorrect? Not necessarily. While all these companies are in CarMax’s market, I’ve shown how they are not apples-to-apples. AN, LAD and PAG all have a presence in the new car market in addition to the used car market. PAG derives a large part of its operating income from truck rentals while the other three do not operate in that segment. They are different by virtue of their size and how they obtain financing (floorplan financing for AN, LAD and PAG while CarMax has relatively little debt in its dealership segment). Finally, these companies fare very differently in recessions. While cars are big ticket items that are particularly hurt during recessionary periods, higher margin services like parts, repairs, and financing offer complementary stable revenue streams. For example, during the period 2007-2009, Ford witnessed a 42% drop in sales, but less than a 10% drop in Ford Finance income. Moreover, while AutoNation experienced a 40% drop in new car sales during this same period, revenues from its parts and service segment declined only 10%. CarMax’s used car segment is relatively more stable than its competitors in recessions; in contrast, large write-downs in its CAF segment do occur (normal default rates are on the order of 2-2.5% but spiked to 4% during 2007-2009). AutoNation experiences significant declines in its new car segment, but has an extremely stable, high margin and more sizable parts and services segment relative to CarMax. Finally, avenues of growth are also very different for these companies and have different FCF implications. AutoNation and Lithia grow through acquisitions and roll-up’s while CarMax builds new dealerships, leading to a subdued FCF during this high growth period. PAG has a large international presence in Australia that similarly affects growth dynamics. For these reasons, I wouldn’t place as much weight on a comparable valuation for these companies.


It is clear to me that CarMax is a great company. However, the price is another story. The company has a robust growth profile, but it does look optically expensive relative to peers; nonetheless, I am leaning more towards the 12-14% IRR calculation when personally looking at the company.

During my discussions with people in the industry, there was a clear admiration for CarMax as a business and what the company has accomplished over the past two decades. This post does read very “pro CarMax” but that’s what the facts tell me. I would love to hear a dissenting opinion. After completing my analysis, I saw that the price had increased from $55 to $95 which hurt psychologically but didn’t change many of the facts presented in this piece. During the same period, Carvana increased from $40 to $90. Carvana officially is more valuable than CarMax despite selling only 25% of the cars. Congratulations to the Garcia family (yes, the father plays a bigger roll than people think). AN is also up 36% over the past year, LAD a whopping 93%, PAG 25%, ABG 67%, and SAH 120%. A rising tide lifts all boats.

1 thought on “Carmax and the Struggle of Small Dealers”

  1. Thank you for another informative blog. Where else could I get that type of information written in such an ideal way? I’ve a project that I am just now working on, and I’ve been on the look out for such info.

Leave a Comment

Your email address will not be published. Required fields are marked *