Carvana: Prisoner’s Dilemma, Used Car Getaway

Carvana looks to execute on a coercive exchange and collateral sleight of hand to buy time and gain some much needed liquidity.

“Do you own the 2025s or the 2030s? Purely discretionary, of course.”

• With everyone fresh off discussions of the prisoner’s dilemma for depositors at regional banks, we now get to shift to the old school zero-sum game of “take from bondholders and give to equity holders.”

• The Carvana example has a more interesting twist with insider control, but the governance questions and Class B stock put some cards in the hands of bondholders that might inspire them to not blink as they often do.

• Bondholder recoveries would be better if the company filed Chapter 11 tomorrow, equitized the debt, and started a bidding war for an asset with way too little cash flow and way too much debt.  

Today Carvana (CVNA) propped a coercive exchange trade at the same time they designated their ADESA US subsidiary as an unrestricted subsidiary. That ADESA unit is no longer a guarantor of the existing notes and will not be a guarantor of the new secured notes proposed in the exchange.  The consideration on the exchanges run from 80 handles on the 2025s and 79 on the 2030s to low 60s on the 2029s. It has been a painful ride for bondholders and shareholders. Now management wants bondholders to take a voluntary haircut.

The new secured notes will be 9%/12% Cash/PIK Toggle Senior Secured Second Lien Notes due 2028. The company is looking to exchange $1 bn in notes with a $500 mn minimum. That is a structure favored by extreme LBOs, and that parallel works here. Given the explicit family control at Carvana (88% of voting per 10K), the current exchange terms are in the spirit of an LBO – even if just to buy time.  

The headlines today indicate that Apollo and PIMCO were against the deal, and they are the largest bondholders. Apollo has a strong interest in the space broadly, so their role as a massive bondholder and aggressive restructuring player thickens the plot.

We would reject the exchange terms for numerous reasons including the fact that this company may very well end up in Chapter 11 anyway and the residual for 2L interests would be highly suspect at that point.  That Chapter 11 risk cuts both ways. On the one hand, bondholders do better if they organize to fight this. On the other hand, the zero-sum game has almost always worked, and this company is heading for Chapter 11 in the absence of an extraordinary shift in financial performance.

So, the impulse for at least $500 mn worth of bonds could be the trigger for some to “get on the structural seniority bus now.” This company might wish to divide the debt interests on the way to its next cash crunch. The fly in the ointment is that usually the “fix is on” before these sorts of deals get launched.

The 2025s appear to be the main target to run up the score with the high coupon 10.25s of 2030 next on the list. Both of those bonds are over 10 points up from the other three bonds (2027, 2028, 2029s).  This is going to be one to watch closely in the coming days as to where the price and trading volume action is moving for the 2025s and 10.25s vs. the pack.  

Governance, insider control, and the dynamics of coercion…

The disproportionate control is mostly a function of the family holdings of Class B common stock. The plug-and-play nature of coercive exchanges is off-the-shelf legal fees over the years. They design them on the way in and get repeat business on the way out. That means the real work will need to be handled by the bondholders at the “voting booth.”

Coercive exchanges were so prevalent during and after the credit crisis that law firms could just as easily delegate it to the summer intern with a template. Here we are again.

We had discussed this coercive scenario in our recent credit profile on CVNA (see Carvana: Credit Profile 3-5-23), and this is a logical move in the playbook when a company is badly in need of new money financing to delay the inevitability of a severe liquidity crunch. As we stated then:  

“Realigning the balance sheet to offer relief would not be likely without some governance conditions attached. Equity deals have ended badly. Bond deals have ended badly. In the aftermath of the credit crisis back in 2009-2010, this was when the liability management advisors might tag team (private equity?) and would trot out the coercive exchange handbook. That would seem to be a weaker hand in this case when management has insider control that they would lose if they followed through on the end game of coercion – which is to threaten to file Chapter 11. Management wants to keep control, so that defangs the coercion. In theory.”

The logic for bondholders in a perfect world is they invoke the “just say no” campaign, and then the unsecured bondholders would hold the cards after the asset based lenders get refinanced in Chapter 11.

In Chapter 11 the debt holders could get equitized and there would be a “financial do-over” under the leadership of a private equity group or an auction of the valuable assets. The Class B shares would get derailed and possibly canceled. This is an unfriendly market for online retail asset valuations right now, but selling the physical assets and taking time to maximize the value of the estate for bondholders would be the driving force rather than keeping control in insiders’ hands. At least this is one potential outcome.

Life as an asset-lite services operation…

The asset protection is pretty skinny at CVNA after the asset-based lending, so an asset-lite services operation only has so much asset protection to offer without a very sound plan and good stewardship of the company. CVNA has to contend not only with other online players but also the 800-pound gorilla of used cars (CarMax) who is also looking to be a major factor in omnichannel distribution. Meanwhile the legacy franchised auto dealers are going to eat from the outside and work their way in across cycles while the financially healthy and well-capitalized KMX sits at the table as a used car incumbent. That is a lot of heat on CVNA.

Negative EBITDA and a massive debt burden are obvious enough and has been for a while at CVNA, but the cash interest expense bill over $600 mn is the dagger that is going to be hard to hide from. If this exchange process goes as planned, bondholders will get a material haircut, those not in the exchange will be further layered down the food chain in terms of recovery rate risk, and the family interest will still be sitting on a majority control via their voting power in their class of stock as detailed in our earlier notes on the company (see Carvana: Wax Wheels 12-8-22). Per the 10K, they control 88% of voting rights.

The idea of bondholder group successfully overhauling an asset lite services operation with outside money and operating skills stepping in to save the day can be seen in Hertz. Hertz surprised a lot of people with its franchise value, but at least Hertz had a much better history than CNVA in earnings (a very low bar to clear). Hertz was at least part of a well-established oligopoly. Hertz had tough competition, but they were all companies that bought all their cars from the same companies, served the same airports, metro locations, and travelers.

The car rental group also all had plays on the next generation of auto technology stories (EV rentals) and in fleet management services operations. CVNA has elements of the next-gen buzz in auto retailing, but that has lost its edge in this market.  The upside just did not arrive yet other than in the 2021 speculative binge and valuation excess.

In a common theme, Hertz was overly aggressive in debt-financed acquisitions and executed badly. CVNA has some of those same ills (including bad timing on ADESA), but without any history of profitability. Insider stock interests bet the ranch using debt and now want the bondholders to redistribute to them in what is essentially a one-way “Enterprise Value Toggle” from debt holders to equity holders.

1Q23 update…

CVNA also gave a provisional update on where the company stood this quarter with another negative EBITDA ready to get printed but not as bad as feared. The volume pain and declines in revenue were partially mitigated by cost cutting and better unit margins. Retail units were down by over 26% and revenue by over 28%. The market is still not friendly for used cars even if the inventory adjustments were heavily absorbed before the 1Q23 period.  Used vehicles have hung in better of late after swinging into deflation mode.

The macro picture raises some clouds as well. Credit contraction theories bandied about include tighter credit for vehicles. Coming off a regional bank whirlwind of the last two weeks, it is hard to get very bullish around discretionary consumer sectors. The same is true for companies serving that sector that have negative EBITDA and massive debt. It’s just an idea.