The Automotive News Top 100 Used-Car Retailers report shows the results for car dealers based in the U.S. and these are the top eight:
Image Source: Automotive News Top 100 Used-Car Retailers
Investors must focus on the view through the windshield and parts of the above view through the rearview mirror are incomplete. Given the way Carvana (CVNA) is growing, they should pass Lithia Motors in retail units sold in 2019 or 2020. Penske is domiciled in the U.S. and they are listed as #2 above but a large percentage of their sales are outside the U.S. As such, the big three to watch in the U.S. market for the foreseeable future are CarMax (KMX), AutoNation (AN) and Carvana. While announcing their first store in the Rio Grande Valley, CarMax mentioned that about 90% of their purchases start on their website or app. My thesis is that the size advantage in the U.S. market enjoyed by CarMax, AutoNation and Carvana is becoming increasingly important as the web plays a larger role in used vehicle purchases. The combination of size and changing buying habits offers tremendous opportunities for these three companies in the coming years.
AutoNation sells both new and used vehicles but this article is looking at their used vehicle sales. The used retailing efforts for these companies tend to exclude the low end of the market as cars with values less than $5,000 or mileage more than 100,000 are generally wholesaled. This article focuses on retail sales as opposed to wholesale although CarMax does have a large wholesale business.
Here are the average retail and wholesale prices for used vehicles for each company:
 Unlike the other numbers that come straight from 10-K filings, the AutoNation wholesale per unit is estimated based on $315.7 million wholesale revenue from the 10-K divided by 62,014 wholesale units from the above Automotive News Top 100 Used-Car Retailers report.
The above table is key because it shows that the typical wholesale vehicle for all three companies is sold for less than $6,000. Here is what CarMax says about wholesale in their 10-K:
The typical vehicle sold at our wholesale auctions is approximately 10 years old and has more than 100,000 miles.
The "100,000 miles" distinction is significant and that number might become 125,000 miles or more in the future as electric vehicles capable of additional mileage become a meaningful part of the market.
It is important to remember that gross margins per vehicle are like comparing apples and oranges with AutoNation because of the way they classify reconditioning in the separate parts and service bucket.
The world is changing with respect to the way people use online capabilities to feel more comfortable with car purchases. Buyers expect online resources to be available such that time can be saved at dealerships. A 2017 CDK Global research study focuses on new vehicle purchases but it applies to used vehicle purchases as well:
In a recent study, we found that 86 percent of retail shoppers prefer a dealer that offers "buy online" capabilities, and 68 percent of in-market shoppers already expect these features to exist today. Once again, millennials are leading the charge, with 90 percent preferring to structure their deal online. Ultimately, this new experience comes down to two things: simplicity and transparency. When asked why they prefer online buying capabilities, 52 percent indicated that it provides them with more information up front, helping them feel more at ease. In addition, 79 percent said they would use the tool because it would save them time at the dealership.
Common sense tells us that the above words “simplicity” and “transparency” apply not just to the price of the vehicle itself but to the financing as well. The CarMax July 2003 Background Paper talks about the fact that there is no negotiation anywhere in the CarMax Auto Finance [CAF] process.
Doing searches on carmax.com, I see 50,017 matches when changing the location drop down box to “Nationwide.” Looking at autonation.com, we see 18,464 used plus 4,842 certified for a pre-owned total of 23,306. Carvana.com shows 13,805 and they are growing much faster than CarMax and AutoNation.
Living in the Los Angeles area before the age of the internet, I shopped for cars by visiting various dealers and getting quotes. Things are different now, people do much of the legwork online and large companies with a wide selection and practical shipment/delivery options are advantaged.
Not only can the big companies offer more selection than competitors, they can optimize their websites faster. The sites with the most sales and the most traffic have more data available to enhance the user experience. Smaller companies with less active websites need to let their A/B tests run longer in order to reach acceptable sample sizes.
Risk of New Entrants
New entrants are always a concern. Amazon Prime (AMZN) is fantastic for repeat purchases but most people let a tremendous amount of time pass by between car purchases. I don’t think infrequent vehicle transactions are a great fit for Amazon but one never knows how they will choose to expand. If a large web company like Amazon enters this space in a more meaningful way than their present involvement then all the incumbent companies will face challenges.
The Washington Post has reported that the FTC plans to place Amazon on a closer watch with respect to antitrust matters. It is probable that Amazon would think carefully before raising their profile by substantially stepping up efforts in this space. At this time I don’t think it would be in Amazon’s best interest to dramatically expand into this industry where customers make infrequent purchases. Still, there is always a chance they could buy a company like Carvana which would be a nightmare for CarMax, AutoNation and just about all the other companies in this industry.
Electric Vehicle Risk
Some time ago my friend let me drive his Tesla Model S and it was a tactile experience getting my hands on the wheel and feeling my head slamming backwards towards the headrest as my foot pressed on the accelerator pedal. Recently that same friend arranged a tour of the Tesla factory in Fremont and it was impressive seeing some of the robots like Wolverine and IceMan in action. Electric vehicles are becoming more mainstream in California and the rest of the country should follow eventually but I think it will take a long time.
Having significantly fewer parts than internal combustion engine vehicles, electric vehicles are projected to allow more mileage per vehicle even with issues of battery degradation. A thought experiment with a small town allows us to see how this can impact the future of used vehicle sales. In this town we have 100 internal combustion engine cars that last about 150,000 miles each such that approximately 15 million miles are driven. Now let's suppose that at some point in the future these vehicles are replaced with electric vehicles that last around 300,000 miles each. Now we only need 50 cars to provide the same 15 million miles. We talked about the fact that the typical wholesale vehicle at CarMax has more than 100,000 miles. In the first case two-thirds of the time we had cars in circulation with less than 100,000 miles on them. In the second case only one-third of the time did we have cars in circulation with less than 100,000 miles. If the “100,000 miles” distinction remains constant for CarMax wholesale then the second part of our thought experience is not good for their retailing business. However, we have to ask how this distinction could change over time. If a car that is only supposed to last 150,000 miles has 100,000 miles on it then it has depreciated substantially as there are only 50,000 miles left. On the other hand, if a car that is supposed to last 300,000 miles has 100,000 miles on it then this car still has a large amount of value. This car that still has 200,000 miles left on it may be retailed instead of wholesaled well past its 100,000 mileage mark.
The 1Q19 deliveries for Tesla show that demand for the Model S and the Model X fell dramatically:
Image Source: Wall Street Journal
The electric car distribution from Tesla and others will increase but the process will take time.
There is also an indirect electric threat that is more significant for AutoNation than CarMax and Carvana. All three companies boost the bottom line with restoration work and there are more parts and restoration possibilities with gasoline engine cars than electric cars. But there is also the service business and I believe AutoNation in particular is more vulnerable here as electric vehicles requiring less service slowly become more mainstream.
In summary, electric cars pose a direct threat to the used car market, especially as battery degradation gets addressed such that they are capable of significantly more mileage than internal combustion engine cars. However, I think it will be a long time before gasoline and hybrid cars are replaced nationwide.
Current Ridesharing Risk
A 2017 Moody’s Analytics study says ridesharing will hurt the used car market as it becomes more widely used:
if ride-sharing eventually dominates the market for private transportation, cars will, on average, pass through far fewer owners during their usable lifespans. Indeed, in such a world, I see no plausible situation where a drivable car would be sold to another party. The net result is a lower volume of used vehicle transactions and thus a significantly smaller used vehicle industry overall.
When it comes to keeping customers happy over the years, taxi cabs have been an egregious failure. The fact that we can use smartphones as remote controls to summon ridesharing vehicles and watch their arrival progress is astounding. For many, ridesharing is a much better option than taxi cabs. But I think the current risks of ridesharing disrupting the used vehicle market in a significant way are overstated. My brother recently flew in from out of town. We looked at the numbers and decided it was cheaper for him to rent a car than to use ridesharing. As ridesharing grows it isn’t getting all its increases from car ownership decline. For example, we saw a play in San Francisco the other day and we took three different Uber rides after parking our car. Were it not for Uber, we likely would have done one of these short trips on BART and the other two on foot.
Ridesharing is growing but it doesn't seem to be replacing car ownership at the pace some forecasts said it would. Looking at the Uber 1Q19 release, much of their revenue growth in total dollars and percentages was from Uber Eats as opposed to ridesharing. Ridesharing revenue went from $2,180 million in 1Q18 up to $2,376 million in 2Q18 for an increase of $196 million or 9%. Meanwhile Uber Eats revenue went from $283 million in 1Q18 up to $536 million in 1Q19 for an increase of $253 million or 89%!
Another aspect of ridesharing forecasts that seems questionable is the fact that the locker convenience of car ownership is minimized. Among other things, I often keep golf clubs, gym equipment, running shoes and hiking gear in my car. Serving as more than a transportation device, my car is a locker as well. If I want to drive from the office to the golf course then it isn’t a big deal. On the other hand, if I gave up my vehicle and had to take an Uber home from the office to grab clubs before hitting the course then things would be more complicated.
I’m not terribly worried about ridesharing dramatically hurting the used car market in the near future.
It will eventually be interesting when we see a large number of ridesharing vehicles that are electric and self-driving. The challenges for used car retailers will be especially high if these vehicles don’t change hands often. I don’t think the amalgamation of all these factors in a mainstream environment will be a reality anytime soon. A November 2018 CNET article talks about Waymo head John Krafcik saying widespread self-driving is far away:
It'll be decades before autonomous cars are widespread on the roads -- and even then, they won't be able to drive themselves in certain conditions, the chief executive of Waymo said Tuesday.
Company Specific Risks
In addition to the general risks above, these companies face specific risks.
Carvana is growing rapidly but there is a chance they won’t be very profitable even after reaching scale.
As AutoNation continues to have more of an online presence, the various dealer names might be puzzling. The average web customer might get confused when the local store is called Allison BMW of Mountain View instead of AutoZone. Another source of confusion for AutoNation web customers might be their on again off again implementation of no-haggle pricing. In early 2019 AutoNation appointed Carl Liebert to succeed longtime CEO Mike Jackson so there is some uncertainty in terms of the landscape under new management.
CarMax might be too slow disrupting the current in-store model as customers prefer to make increased use of the omni-channel approach. If CarMax isn’t fast enough then they might be disrupted by more companies than just Carvana.
Investors who see the online potential should take a closer look at these companies. Those who agree that some of the short term risks are overblown ought to find them interesting.
I tend to hold companies for many years and I think it is more important to bet on the right companies than the right prices for my style of long term investing. When it comes to valuation there are many parameters and the odds of getting everything precisely right are low. However, I need to be directionally correct with the big picture factors and accurate enough to hit the side of a barn.
When it comes to valuing a company, what ultimately matters is the amount of cash that can be pulled out from now to judgment day relative to the market cap and overall enterprise value [EV].
Carvana does not make money at this time and it isn’t clear when this will change nor how much we can expect them to make in the next five to ten years. As such, they are in my “too hard” pile in terms of how much the company is worth.
CarMax and AutoNation have both been profitable for decades and I think they have attractive merits in terms of the amount of money they should make in the next five to ten years relative to the price of each company.
AutoNation has a trailing twelve months [TTM] P/E that is a little above 9 while CarMax has a P/E of around 16. But the P/E ratio is only part of the story. The TTM P/E ratio provides limited information as the GAAP accrual earnings are in the rearview mirror and they can be less representative of the future bottom line economics than adjusted free cash flow [FCF]. And the P in the numerator excludes debt which is especially important when comparing CarMax and AutoNation. Excluding CarMax non-recourse debt, AutoNation has a significantly higher debt to market cap ratio than CarMax. CarMax has net debt of $1.6 billion relative to a market cap of $13.3 billion while AutoNation has net debt of $2.4 billion relative to a market cap of just $3.7 billion.
When we look at EV relative to cash flow generation we have debt in the numerator so we can add interest on debt back into adjusted FCF to keep the numerator and denominator consistent. Therefore we use adjusted adjusted free cash flow to the firm [aFCFF] in the denominator. The EV/aFCFF ratio for CarMax is 15.5 [$14,922 million/$961 million] and it is 14 [$6,070 million/$433 million] for AutoNation.
Based on past earning power, AutoNation is statistically cheaper than CarMax but we need to take a look back over the years as we digest what history can tell us about future prospects:
Image Source: CarMax Financials From Morningstar
Image Source: AutoNation Financials From Morningstar
Morningstar shows that CarMax has increased top line revenue 143% over the last ten years from $7.5 billion to $18.2 billion. Meanwhile AutoNation has increased top line revenue 95% from $10.8 billion to $21.1 billion. CarMax has reduced the share count 21% from 222 million to 176 million while AutoNation has reduced it 49% from 177 million to 90 million. CarMax EPS has increased 280% from $1.26 to $4.79 while AutoNation EPS has increased 288% from $1.12 to $4.34.
Over the last five years CarMax has done a better job than AutoNation increasing top line revenue and earnings per share. Both companies have excellent opportunities moving forward but I understand CarMax better and I’m a little more excited about their prospects.
I think people can get into trouble when they go long company A of a promising industry and short company B. Over the next five years I predict these three companies will do better than smaller competitors in terms of relative revenue growth. Carvana is still in my “too hard” pile with respect to their future per-share bottom line economics. I like the relative per-share free cash flow prospects of CarMax and AutoNation over the next five years.
Adjusted Enterprise Value:
mkt cap ($79.89 * 166,187,221 shares)
long term debt
financing and capital lease
cur. portion of debt
cur. portion of financing/capital lease
mkt cap ($41.24 * 89,208,853 shares)
long term debt
cur. portion of debt
Spreadsheet sources: CarMax 2019 10-K; AutoNation 1Q19 10-Q.
*Share prices are from 6/7/19.
*Non-recourse debt is excluded for CarMax.
*Floorplan debt is excluded for AutoNation.
Adjusted Free Cash Flow
I treat Trailing Twelve Months [TTM] Adjusted Free Cash Flow as net cash provided by operating activities minus stock-based compensation expense minus purchases of property and equipment. In the case of CarMax I start with adjusted net cash provided by operating activities which is essentially moving the issuances of non-recourse notes payable line and the payments on non-recourse notes payable line from net cash provided by financing activities to net cash provided by (used in) operating activities.
Once the above is done I try to separate growth capex from replacement capex and make an adjustment such that growth capex does not end up being subtracted.
CarMax aFCF not adjusted for growth capex
= $1,054 mn - $75 mn - $305 mn
= $674 mn
CarMax aFCF adjusted for growth capex
CarMax capital expenditures totaled just $22.4 million during the financial crisis in the fiscal year ended February 28, 2010. This was because growth capex was paused during that time. These days they have more stores so maintenance capex is likely closer to $75 million. I think that all but about $75 mn of the above $305 mn capex is for maintenance so I add $230 mn back in to get an adjusted FCF of $904 mn.
AutoNation aFCF not adjusted for growth capex
= 1Q19 + FY18 - 1Q18
= [$259.7 mn - $13.7 mn - $60 mn] + [$511 mn - $25.5 mn - $387 mn] - [$198.7 mn - $14.3 mn - $101.5 mn]
= [$186 mn] + [$98.5 mn] - [$82.9 mn]
= $201.6 mn
AutoNation aFCF adjusted for growth capex
The growth capex at AutoNation is harder to quantify. Their capex declined during the financial crisis but it seemed to slow earlier for them in the crisis than it did for CarMax. It was just $90 million in 2008 and $75 million in 2009. Things like dealer incentives are somewhere between maintenance and growth capex. They increase overall capex but they tend to show up in the financials as lower cogs as opposed to increased revenues. Their overall TTM capex was $345.5 million. I’m guessing that these days about $200 million of this capex is for maintenance. Adding $145.5 million back into FCF, I get about $347 million adjusting for growth.
Adjusted Free cash flow to the firm [FCFF]
We can use adjusted FCFF which adds debt interest back to adjusted FCF such that the numerator and denominator are consistent with respect to the EV/aFCFF ratio.
The 2017 Tax Act lowered the federal tax rate from 35% to 21%. Including state taxes, the overall effective tax rate for CarMax and AutoNation has been close to 25% in recent periods.
CarMax had $76 million in 2019 interest. Adjusting for taxes, we add $57 million back to adjusted FCF to get adjusted FCFF of $961 million.
Excluding floorplan interest, AutoNation had TTM net interest of $114 million [[$27.8 mn - $0.2 mn] + [$119.4 mn - $1.1 mn] - [$32.3 mn - $0.2 mn]]. Adjusting for taxes, we add $86 million back to adjusted FCF to get adjusted FCFF of $433 million.
= $842 million
= 1Q19 + FY18 - 1Q18
= $92 mn + $396 mn - $93.7 mn
= $394.3 mn
The CarMax TTM P/E is around 16 or $13,277/$842 mn.
The AutoNation TTM P/E is a little above 9 or $3,679 mn/$394.3 mn.
Disclaimer: Any material in this article should not be relied on as a formal investment recommendation. Never buy a stock without doing your own thorough research.
Disclosure: I am/we are long KMX, AMZN, UBER, VOO. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.