Carvana: Wax Wheels
A history of negative EBITDA on spiking unsecured debt raises some serious questions with the used market eroding.
CVNA flew too high in valuation, moved too far and too fast on expansion, and needed an anomalous used car market to remain distorted for much longer to sustain market favor.
Negative EBITDA, soaring debt, deterioration at the macro level, and troubled industry fundamentals will drive a major reset at CVNA.
Intensifying competition from healthy franchise dealers and a financially powerful incumbent in the used space will not make the other side of Chapter 11 any easier without a new strategy.
News reports indicated that Carvana (CVNA) was in talks with restructuring advisors and that a majority of unsecured bondholders (Apollo as the most exposed with PIMCO and Ares reportedly in the mix) were in dialogue also. This sets the stage for some major recapitalization discussions ahead. A Chapter 11 filing would not be a shock given the pain in the core business, the performance of the stock YTD, and the deeply distressed pricing on the bonds. It is not transparent at this point, but the flavor of the reporting was that bondholders want to avoid conflicts and coercive restructurings without a broad agreement. The situation at Carvana is complicated by super voting shares and a dual class structure that gives the Garcia family interests voting control. We got rolling on some used car market commentary recently where we discuss Carvana and CarMax (see Market Menagerie: The Used Car Microcosm) in the context of the broader used car ecosystem.
The numbers being generated by CVNA make the end game of a debt restructuring almost inevitable with materially negative EBITDA on a substantially enlarged balance sheet. The unsecured bond burden more than doubled in 2022 even as business conditions were deteriorating rapidly. Cash bleeding plus excess debt as part of a material expansion plan makes for a tough combination in a consumer durables sector (used cars) that had been behaving irrationally across 2021 and built some false hope into the stock valuation.
Used cars deflating helps CPI but proves a dagger for CVNA…
The used car market looks like the sector is going to get a lot more scrutiny ahead as online players have seen market collapses while franchise dealers hang in much better for a range of reasons we touched upon in the prior note. Dealers have well established business models, have stronger balance sheets and cash flow, and inherent competitive advantages in used cars (trade-ins, lease-roll-offs, historical customer retention). The franchise dealers are also pursuing multiple digital initiatives across new and used car sales (including sourcing) and higher margins Parts and Services operations, so the learning curve and investment in online capabilities keeps growing. Ongoing investments in EV training for employees will take spending and effort but will be crucial to maintaining a competitive edge. We will be looking more closely at the Big 6 public dealers in coming weeks, but the main takeaway is that they are not standing still in making inroads in the online space.
For Carvana, the days of reckoning on balance sheet overhaul are here now, and possibly a new or evolved strategic plan could come with that. The bottom line on CVNA right now is that EBITDA is negative, the debt burden is massive relative to the negative cash flow, the macro backdrop is bad, and the used car micro fundamentals are worse near term as the new vs. used price differentials revert to logical relationships on later model used cars. The competitive playing field is also intensifying and commands high levels of strategic focus from financially healthier competitors. That is notably CarMax in used cars but also the major public dealers in their digital game plans. The next stage of investment by the dealers is likely to see most maintain an elevated relative focus on used car expansion as seen in recent years. That commitment will also get tested in 2023 for some.
Key items to ponder as the CVNA restructuring unfolds…
Barriers to entry in digital investments is a checkbook: The move toward a more ecommerce flavor is seen everywhere in auto retail. The challenge is to do it well and within core competencies. The trade publications have been all over this topic for years with franchise dealers’ strategies getting a lot of focus. Digital retailing arguably is not a disruptive new business at this point but more about execution and investment by any combination of new players attempting disruption and more focus from established brick and mortar incumbents. The relative stock performance YTD tells a story around the market perception (see chart). The good news for the dealers is they have a lot of core competencies in auto new car retail and in parts and services. Digital figures heavily into both. They also are all expanding and refining their financing and insurance (“F&I”) business lines either through their own operations, through partners or through the OEM captives. Digital is a major factor there as well.
Used car fundamentals will get worse before they get better: Prices for used cars vs new cars are still seeing used car prices normalizing (a polite word for used prices moving lower). Used car indexes are down 14% YoY, and the suddenness is hitting inventories and margins. We plot the crazy used car inflation in last months CPI commentary (see CPI: The Big 5 Buckets and the Add-Ons), and there has been no shortage of price pressure and inventory risks on the lots of the dealers or on the order books of the many players in used vehicles. The artificial and distorted compression of used cars vs. new cars seen in 2021 is widening back out, and that means higher depreciation in the cash loss sense of the term. The other part of the used car fundamental picture is financing costs with average used car financing at 9.3% for 3Q22 but across a very wide range that can run into 20% handles for low end customer credit quality. Auto finance issues are for a different day.
Nonprime and subprime customers face more stress ahead: A majority of used car funding is still super prime (12.6%) and prime (43.3%) borrowers at around 56% of total used car financing. That quality mix is sometimes underappreciated with many assuming used car customers are heavily subprime. That does not change the fact that that over 41% are in the combined categories of near prime (21.6%) and subprime 19.7%. There is a minimal 2.9% in deep subprime (note: those stats sourced from Experian 3Q22 data). During 4Q22 and rolling into 2023, risky consumer credit exposure will be seeing higher reserving needs and higher charge-offs.
For CVNA, the financing business is one factor that calls for reasonable credit health even if just for the ability to be judged as a viable counterparty in contractual relationships. That means auto finance businesses will face headwinds all over the industry, but you need to be in it directly or as a viable relationship partner to capture customers and be an acceptable counterparty. The counterparty risk topic includes repo lines in securitization interests or counterparty faith in sale-leaseback deals. Counterparty risks matter even if that is not the main event right now.
The relationships with financing receivables buyers and securitization structures are a topic that needs more transparency on what the Chapter 11 risk would mean in areas such as servicing and the rights of the various parties in a filing. I don’t have an answer to such questions at this point. That topic came up with a vengeance in framing Hertz ABS risks and will likely come up again here to make sure the role of Ally and ABS structures are clearly understood. Ally’s recent floorplan facility from late Sept 2022 requires monthly interest payments and 12.5% of the principal amounts owed to Ally to be held as restricted cash. That highlights CVNA credit risk and the uncertainties of collateral risk.
Hertz is a useful frame of reference but a very different situation than CVNA: The structural evolution of mobility markets whether by subsector (ride hailing, car sharing, car rental, subscription services, etc.) or related services markets (remarketing, auctions, ecommerce variants, data, etc.) will be ongoing. This structural change promises growth opportunities and that is what keeps more private equity and strategic players in the game looking for ways to gain scale and drive upside in valuations. That came up a lot with Hertz, but at least Hertz had a long history as a very well-established brand. Hertz had a case to make for solid longer-term profitability based on reasonable EBITDA generation and Hertz core end markets. The EBITDA assumptions had a path for Hertz to even get to a par recovery. The financial model rolled into the pro forma stock valuation exercise as the later stages of 2020 unfolded, the vaccines were delivered, and then the Hertz bidding process started to control the restructuring.
The CVNA bond recovery story will be a challenge and even more assumption-driven than Hertz: As a memory jogger, Hertz CDS cleared the auction at a 26 handle, but the unsecured bonds ended up at a par recovery by the time the final deal was struck. It is safe to say that the CDS auction vs. the eventual par recovery was dramatically different! My own recovery estimate in a prior role in the immediate weeks after the Hertz filing was for a par recovery on 1L and 2L and ballpark 50 on unsecured bonds once the moving parts of the restructuring could be worked through. Life got tricky when Hertz threatened to blow itself up (and the accepted industry ABS legal interpretation on the leases) with its legal gambit in challenging the ABS structure. That is a long story for those who were not deeply involved. Revisiting that memory lane can wait for another day when we dig into car rental.
The Hertz structural assumption from my side was for equitization of all the unsecured debt and a continuation of the ABS anchor for fleet financing. The idea was to find a reasonable target fleet size, EBITDA margin estimate, and then make assumptions on whether the “new world of mobility” and cleaner balance sheet would allow for a higher EV multiple to be assigned. Hertz sure had a wild process, but in the end, there was considerable logic in the numbers that could be traced to the past and the future evolution. CVNA does not have history to support healthy numbers. The question in CVNA’s case is what else will be rolled into its future (merger with a public dealer?) or what alliances (such as the Hertz vehicle remarketing program) to reduce the risks of used car sourcing (merger with a car rental operator?). This is where whiteboard guesswork starts to get into industry evolution finger painting.
The assumption in framing CVNA exit EBITDA rates will be interesting: One of the main problems in framing CVNA is that it lacks a multicycle history in its business model and does not have a history of profitability to confidently lean on. That was not the case with Hertz who had a long history with a high correlation to airport travel and leisure travel. There was considerable secular change in the car rental business (the rise of ride hailing and decline in business revenues for car rental companies), but there was more logic in the modeling concepts. We know bankers and owners can come up with defensible assumptions, but in this case the history will not lend emotional support!
In the case of Hertz, even low double digit EBITDA margins would make it much less profitable than the industry leader (Enterprise) in more or less the same core business (renting cars at a price to recover costs on vehicles all bought from the same suppliers). A 10% or 12% EBITDA margin was very defensible in Chapter 11. The challenge was fleet size and whether the equity would get any credit for a growth multiple after it emerged.
The CVNA model is about sourcing cars at a price and selling them for more. They moved quickly into the space when demand for such online services soared, but it is not like that is a proprietary asset as a strategy. Meanwhile, many others are in the same business, but they are not purely dependent on online sourcing. The worry for those sourcing vehicles in the current market is that “If you won the online battle for that used car, maybe you overpaid.” Meanwhile, a dealer just had the same car returned off a lease. Or the customer with some modicum of dealer and brand loyalty just dropped off a trade-in and you get to sell him F&I products (leaving wiggle room on the trade-in price). The franchise dealers are now in the more enviable position.
There will be a lot to flesh out on CVNA as more of what is going on “over the wall” sees some light of day.