Carvana (NYSE:CVNA) update (CVNA – $187) – Originally I was going to keep this brief. And I apologize to those who are tired of hearing about what a great short CVNA is only to watch the stock relentlessly levitate higher. But once I started re-digging into CVNA’s Q2 numbers and the footnotes in the most recent 10-Q, I believe CVNA ultimately will prove to be one of the best shorts in the next 12 months.
Recall that Ally Financial (NYSE:ALLY) announced earlier this month that auto delinquencies and charge-offs jumped considerably more in July and August than its internal risk-control model led management to believe would occur. “Higher than its risk-control model” projected. That is a key phrase.
Risk control models at sophisticated algorithms designed to predict the percentage of loans that will fall delinquent, thereby enabling management to determine its willingness to take lending risks plus how much to charge (loan rate) for each given tier of risk. When the actual outcome jumps “outside” of the guard rails established by the model, it means that a fundamental shift occurred in the ability or willingness of borrowers to make loan payments. Whatever that shift was, in all probability, it’s worse than ALLY’s management represented.
I bring this up because ALLY is the bank that finances the loans underwritten by Carvana. CVNA has a “Master Purchase and Sale Agreement” with ALLY in the amount of $4 billion which enables CVNA to sell the loans it originates to ALLY, subject to certain underwriting criteria.
Given the disclosure by ALLY, it’s a good bet that a material percentage of the loans sourced from CVNA is part of the problem for ALLY. The current agreement expires on January 10, 2025. It’s also a good bet that ALLY likely will tighten the underwriting standards on the auto loans it is willing to buy from CVNA. In turn, this will adversely affect CVNA’s sales volume.
Further to this point, not all of CVNA’s loans (“finance receivables”) go to ALLY. It has also issued $14 billion worth of Asset Backed Securities (ABS) – in this case bond trusts securitized by the CVNA auto loans. In aggregate, the delinquency rates for the loans in these ABS trusts are beginning to soar.
The graphic on the next page is from CVNA’s Q2 10-Q. It was prepared by and posted on Twitter by @Bluechip, someone with whom I share ideas and analysis.
I believe he’s a professional investor and analyst. He’s also a fellow CVNA bear. As it turns out, and not surprisingly, the delinquency rate on CVNA underwritten auto loans is escalating:
12.6% of the total loan balance in these trusts is in some degree delinquency. The total 31-day+ delinquency rate jumped from 7.71% in June 2023 to 8.71% in June 2024. The 61-day+ delinquency rate jumped from 2.39% at the end of 2022 to 3.373% at the end of 2023 to 3.97% as of the end of June. The big jump in all delinquencies means that the 61-day+ delinquency rate will be considerably higher by year-end. CVNA has balance sheet exposure to these trusts in the form of variable interest entities. If enough of the tranches above CVNA’s investment in equity tranche become distressed, CVNA’s interest will be wiped out and it will have to recognize the loss on its income statement.
In its latest ABS pool, the weighted average interest rate is nearly 14% with an average remaining term of 71 months.
The average principal balance is $24,504. But I noticed something interesting. The average selling price for a CVNA vehicle YTD in 2024 is $23.7k (per the latest 10-Q):
Because the average loan balance in the loans that went into CVNA’s latest ABS deal are below the average selling price of the vehicles, it means there’s a meaningful percentage of these loans that were made with negative equity.
In other words, the car buyer was underwater on the trade-in and CVNA loaned enough to buy the new used car and pay off the loan on the old used car. Those are also likely loans extended to the bottom half of the FICO range in this loan pool (576 to 894), which means the loans carry an interest rate substantially higher than the average rate for the pool (13.69%). It also means that a material percentage of the loans will eventually default.
As with ALLY, as the rate of distress and defaults rise, the market will be less willing to invest in CVNA ABS bonds, which means CVNA will either be forced to retain a higher percentage of the loans that it dishes off to ALLY and the ABS market or curtail the number of loans it underwrites. The latter issue will adversely affect sales.
Here’s also why that matters. In Q2 2024, buried in its income statement (but disclosed in its footnotes), CVNA recognized non-cash $173 million in gains on from the sale of finance receivables. It’s hard to tell where CVNA stashes this number but it is supposed to be an entry below the operating income line item. It’s probably in the “other expense (income), net” line. Regardless CVNA’s net income before taxes was $49mm in Q2.
See the problem? Excluding the non-cash gain on loan sales of $173mm in Q2, CVNA’s NIBT would have been a $124mm loss. Moreover, given that the “other expense/income net” line item was $124mm lower than the gain from loan sales, it likely was because CVNA had to write-off bad loans of that amount or close to that amount.
But there’s more. Recall that CVNA “restructured” its debt and replaced a large amount of cash pay bonds with PIK bonds (payment-in-kind, meaning the interest is paid with more bonds). The bonds PIK for two years then revert to cash-pay. The restructuring temporarily reduced CVNA’s total debt outstanding but if CVNA PIKs the bonds for the fulld two years, it will be saddled with the same amount of debt as before the “restructuring” and the interest rate on the bonds will be even higher (12-14%). The bonds revert to cash pay in 12 months.
CVNA would have trouble making cash interest payments right now on that debt. As the economy gets worse and used car sales decline, CVNA will have no ability to make cash interest payments on that debt.
The bottom line is that CVNA remains insanely overvalued. It trades at 54x trailing earnings and 123x the consensus 2025 earnings. Those metrics are based on just the market cap. Including the debt, CVNA is trading at an enterprise value of 34.2x the last six months’ operating income annualized. But CVNA’s operating income likely will decline over the next six months. For point of comparison, Carmax – which is also overvalued – trades at a trailing P/E of 29x and a forward P/E of 21x.
An argument can be made that CVNA is not even worth the amount of debt on its balance sheet, which is $5.4 billion right now and will be over $5.6 billion when the PIK bonds go cash-pay. Using the Street estimate for 2025 net income, which invariably will turn out to be too high, that “p/e” for the debt is 34x. Using the last six months’ net income annualized that debt is 52x net income. But CVNA’s net income declined from Q1 to Q2. Using the Q2 net income annualized, the debt is 75x net income. Compare those valuation ratios to the ones for KMX above. When CVNA eventually has to restructure its debt, the shares theoretically are thus worthless because that debt will be worth far below par value in bankruptcy.
The short interest as of mid-September has declined by a considerable amount. For most of the last few years, CVNA’s short interest has been over 30% of the share float but it has declined to 12.6% of the float. That’s still fairly high but it makes it more difficult for Street trading desks to engineer short-squeezes, particularly since the founder, Ernest Garcia II has been a habitual seller of shares, including dumping over one million shares this month.
CVNA’s move higher is truly puzzling. It’s reminiscent of the trading action of the dot.com stocks in late 1999/early 2000. If the pattern remains consistent, it would appear that CVNA is getting ready to sell-off down to the $130-$140 area. The next earnings report is at the end of October. I think CVNA will move lower ahead of earnings, particularly if there’s any type of material move lower in the stock market. This is likely why Ernest Garcia II has been one of the most prolific insider sellers in the entire stock market.