Description

I like Carvana a lot.  I think the consumer value proposition and experience is fantastic.  I think the management team is incredible.  I think the stock price allows for years of value creation.  Vroom pitches themselves as a “hybrid” or “asset-light” version of Carvana with nearly as good unit economics, and better capital efficiency.  If that doesn’t sound good enough, Vroom is run by Paul Hennessy who gets very favorable reviews from his previous positions as CEO of Priceline.com and CMO of booking.com.  I think Vroom is a short.

I think Vroom is a short for the same reason that Carvana has been so uniquely successful in both its customer experience and its financial performance: the level of vertical integration required in this business.  Between his background in auto retail and auto finance, as well as his technology influence from his time at Stanford, Ernie Garcia III had unique insight into the pain points in auto retail and why vertical integration was required for a compelling digital platform.  In particular, he knew how important it was to control the inspection and reconditioning process, the logistics network, and the finance platform.  Owning these areas of the model has allowed Carvana to uniquely address key consumer pain points, while scaling on a fixed cost infrastructure.  Ernie was also lucky that his family had the experience and checkbook with which to confidently back him.  Vroom, as with Shift and Beepi, are/were “Internet” companies who raised money from “Internet” investors.  This, ironically given the incredible scalability of nearly every Internet model, has made these Carvana also-rans inherently unscalable.

Unfortunately for Vroom, the choice of Paul Hennessy as CEO has only compounded this problem.  Paul has an impressive background in online marketing and scaling digital platforms.  However, key aspects of the auto retail model in particular make it very different from a more “typical” Internet company like Priceline/Booking.com.  I believe this structural difference is beginning to dawn on both the company and investors over the last two quarters.  The company has begun to talk about integrating further into its last mile logistics to try and address both customer experience and cost issues, and the stock has come under pressure as I believe some very good Internet investors who know Paul, have begun to realize this is a different type of ballgame.

While, as I mentioned, the stock has already start to come under pressure, I think that that two catalysts could drive significant underperformance over the next 12 months:

1)     IPO lockup expires at the beginning of December and I believe that there are shareholders who are beginning to realize the fundamental dynamic between Carvana and Vroom described above.

2)     Next year the street is baking in a huge ramp in GPU for Vroom.  I don’t believe it will come anywhere close.

On the second, I think it’s important to look at Carvana for context.  They are vertically integrated, and thus should fundamentally have more operating leverage, have had pretty much flawless execution on a consistent plan, and have seen approximately $600/year of GPU expansion over the past several years.  Vroom on the other hand outsources a considerable amount of their model (the flipside of their “asset-light” model is that it’s much more variable cost), the execution has been anything but flawless with awkward M&A (TDA), management changes, and now a slow pivot to more vertical integration, and yet the Street is expecting their e-commerce GPU to expand by approximately $800 in 2021.  And that’s with Q3 of this year being a tough comp given the stronger than expected pricing environment (at a Blair conference yesterday the CFO said this is what drove the upside to their original guidance, despite the release talking about improvements in reconditioning and logistics).  For what it’s worth the Street is also expecting material leverage in SG&A per unit despite the fact they are forecasting little in Q4 and have talked about challenges / bottlenecks in customer sales and service.

Below are a few key differences to consider when analyzing/comparing Carvana’s current and target GPU to Vroom.

ASP:

Vroom started at a much higher GPU bc it was trying to access better unit economics out of the gate through a niche in high-priced vehicles.  The bulk of market demand is in lower-priced vehicles, and GPU has begun to disappoint as they make this move.  Q2 results were instructive on this front, and while they have tried to backtrack since, the CFO admitted on the Q2 call that lower ASP cars have lower vehicle and product gross margins.  They still have a ways to go to get to Carvana levels.  This is a headwind to GPU.

Reconditioning and Logistics:

Carvana owns their reconditioning and transportation network.  Furthermore, they take a LEAN manufacturing approach to reconditioning where rather than industry standard of one high-paid tech detailing a whole car, they hire and train up to individual tasks (windshields, tires, etc.).  Vroom outsources these functions to legacy industry participants.  Vroom’s model is both higher cost per unit/mile, and does not scale because they do not own the infrastructure.

Cars sourced from customers:

Carvana sourced 56% of their cars from customers last quarter.  Vroom, despite offering more on the average car per my research, is stuck in the ~35% range.  Additionally, that’s with the benefit of having TDA be a boost to this number, and they don’t tell you how many sourced cars are through TDA vs the website.

Finance:

Carvana is an integrated dealer-lender – they do everything, originating, scoring, pricing, etc., except invest in the loan.  They are like Carmax’s CAF unit, only better bc the LTVs are better.  Vroom is just an originator.

These are just a few examples of the deficiencies and lack of leverage in Vroom’s model.  Additionally, all of the above leaves out the fact that I am just touching on the economic implications of the differences in these two models, but there are real consumer experience implications as well.  Vroom cars take longer to purchase with worse customer service, are more expensive and take longer to arrive – this is all because key functions are outsourced.  Their discussion of in-sourcing last-mile operations supports this.

Vroom had a great pitch: Goldman IPO of an Internet platform company managed by a proven and well-liked executive and following in Carvana’s success in a large and fragmented market.  I think the reality is that Vroom is a completely different model than Carvana, and that has material implications for both customer experience and unit economics.  Furthermore, while Paul Hennessy’s background is impressive, he lacks the auto retail industry knowledge to navigate a strategy shift to a more successful model (and if he did, that shift would still require a lot of capital and time).  I think investors have begun to sniff all of this out, and that Vroom’s share price underperformance will accelerate as IPO lockups expire and Vroom begins to disappoint on very high 2021 expectations and leave investors questioning if the company will ever be profitable.

Vroom and Shift like to tell people that this is not a winner-take-all/most market, and that you needn’t look farther than the current industry fragmentation to see that.  Don’t listen to them.  The whole point of what is going on in this sector is that an industry shift from a variable cost structure to a fixed one is slowly and steadily shifting the industry from a fragmented one to a winner-take-most/all industry, much like a lot of other areas of retail.  Vroom is not that winner.

Catalyst

IPO lockup expires.

Disappointing financial results.