Carvana reported Q4/full-year 2021 numbers on February 24th after the close. The Q4 net loss was wider than forecast at $1.02/share vs 41 cents/share a year ago. Revenues “beat” estimates, as unit sales were 113k vehicles vs 72k in Q4/2020. The operating loss for Q4 and the full-year was $104 million.
Where the numbers really get interesting is in the Statement of Cash Flows. CVNA’s operations burned $2.5 billion in cash during 2021 vs $608 million in 2020 and $757 million 2019. The Company also spent $557 million on capex. Total cash burn in 2021 was $3.15 billion. This was funded by a $3.5 billion increase in debt. Carvana uses debt to fund its massive cash flow burn and to fund Ernest Garcia II’s colossal extraction of cash from CVNA. This by definition is Ponzi Scheme on a scale that would make Bernie Madoff blush.
More shocking than CVNA’s financial performance, the Company announced the acquisition of ADESA’s wholesale physical car auction business. ADESA is subsidiary of KAR Global (KAR – $18.94). CVNA is paying $2.2 billion, all of which is funded with debt. CVNA is also taking down another $1 billion in debt to fund $1 billion in “improvements” across ADESA’s 56 auction sites.
I pulled up ADESA’s numbers as broken out in KAR’s 10-K. First, over the last three years, ADESA’s revenues have plunged from $2.1 billion in 2019 to $1.5 billion in 2021. In 2021 ADESA squeaked out a $21.4 million operating profit. This compares to a small operating loss in 2020, but was down from an operating profit of $264 million in 2019 and $307 million in 2018. On this basis, CVNA is paying 102x the 2021 operating profit.
But CVNA is taking down a total of $3.2 billion in debt for this deal, including an additional $1 billion to fund capex improvements. The total amount of debt being used for this transaction is a whopping 149.5x operating profit. Regarding the $1 billion borrowed to fund improvements, KAR had planned just $115 million in capex for 2022 for ADESA.
While I believe using the operating income multiple is the “cleanest” number on a GAAP accounting basis, using ADESA’s 2021 EBITDA of $194 million, CVNA is paying 16.8x EBITDA. This is a “nose-bleed” multiple for a business with thin profit margins and a shrinking revenue based.
Using an EBITDA multiple to justify the amount of debt used in a buyout is a problematic gimmick popularized by Drexel Burnham Lambert. Based on ADESA’s historical numbers, the amount of D&A and capex is similar each year, which means that in order to avoid cannibalizing the asset base, the cash flow represented by D&A needs to be reinvested. The amount of cash represented by D&A therefore is not available to service debt or pay out dividends to shareholders.
CVNA is thus paying an absurd price for a car auction business with a shrinking revenue base and tiny operating income. What’s worse, it’s funding the entire purchase with $3.2 billion additional debt. This is on top of the $3.5 billion in debt needed to fund its cash flow burn in 2021.
When this deal settles, CVNA will have $6.4 billion in long term debt. It also has $2 billion outstanding under its short-term revolving facilities, which I believe is secured by vehicle inventory (floor financing primarily). Nevertheless total debt outstanding will be $8.4 billion. In 2021, CVNA’s interest expense was $176 million. A good estimate for 2022 interest expense is that it will be close to double to $352 million. Unless Moses appears and parts the Red Sea for CVNA in 2022, CVNA will be borrowing more money in 2022 to fund interest payments.
This leads to the question of what bank would possibly allow CVNA to borrow $3.2 billion to fund the ADESA acquisition, given the numbers as laid out above. JP Morgan and Citi are funding the debt. I don’t know if they will be syndicating part of the loan to other banks or to hedge funds until this deal closes and the 8-K is filed. As part of the acquisition, CVNA will assume KAR’s $2.5 billion in finance receivables plus $579 million in PP&E. The debt will be fully secured by these tangible assets and likely anything else unencumbered on CVNA’s balance sheet (Ally’s warehouse loan facility is secured by vehicle inventory).
The motive to fund this cesspool entirely with debt is profits. Until the docs are filed, we won’t know the rate of interest on the JPM/Citi loan. But in all likelihood the rate of interest will be higher than the rate paid by a company like Carmax on its secured debt facilities. My best guess is that JPM/Citi are going into this with an asset package that, at least on paper, over-collateralizes the amount of debt.
JPM/Citi will earn a huge fee on the transaction plus they will earn a big spread on the rate charged to CVNA vs the near-zero cost of capital for these banks. I’m also guessing that the two banks will lay off part of the credit risk via credit default swap transactions with hedge funds. From this standpoint, the loan package at least cosmetically entails low risk relative the amount of fees and interest the two banks will skim.
Why would CVNA want to do this deal and why would CVNA’s Board of Directors approve it? As for the Board approval, I don’t know. Maybe someone can write Dan Quayle (he’s a director) and ask him. Recall that CVNA outsources several operational functions to DriveTime (Carvana’s Cash Burning Ponzi Scheme). This is a corrupt, related-party arrangement used by Earnest Garcia to suck $100’s of millions in cash out of CVNA.
It looks like this deal was put together in order to give Garcia’s DriveTime business the ability to mine huge fees securitizing and servicing the ADESA finance receivables. In addition all of ADESA’s used car reconditioning needs, similar to Carvana, will presumably be outsourced to DriveTime.
CVNA is a house of cards riddled with conflict of interest, securities fraud (the finance receivables business) and financial fraud (the DriveTime relationship). Its operations went from burning $35 million in 2015 to burning $2.5 billion in 2021. CVNA requires outside investor money from the issuance of equity and debt – mostly debt – to fund the cash burn. It has been doing this for seven consecutive years.
Eventually this company will collapse and the shares will be worthless, but not before Ernest Garcia has extracted many billions from this scheme – sale of his stock plus the cash sucked out by DriveTime. In a different era (30 years ago), CVNA would have been shut down by the regulators similar to Drexel Burnham Lambert. If the bear market takes the SPX down another 30-40% in the next 12-24 months, CVNA will hit the wall.
This analysis is from the February 27th issue of my Short Seller’s Journal. At the time CVNA was trading at $152. My subscribers and I cleaned up shorting and buying puts on CVNA. In every issue I provide put option ideas to accompany my short ideas. I am still long puts on CVNA because I think it will go below $10 within the next 24 months. You can learn more about my newsletter here: Short Seller’s Journal