@the10thman1: Carvana $CVNA Deep Dive

Episode Summary

Executive Summary: This episode features an anonymous investor, “10th Man,” explaining his path from sell-side energy research to buy-side investing and his expectations-based approach to analyzing Carvana. He argues Carvana’s logistics, data, and customer-experience advantages are real, but his base-case valuation still lands below the market price because he assumes lower market share and tougher competition than bulls expect.

Main Topics: 10th Man’s investing background and philosophy: He describes moving from law-school aspirations to finance, then from sell-side equity research to buy-side investing. His philosophy centers on expectations investing: identifying where market expectations are too optimistic or pessimistic, especially in forecastable, changing industries. Why Carvana was an interesting idea: Carvana was attractive because used-car retail is fragmented, disliked by consumers, and underpenetrated online. He initially studied CarMax, then Carvana, because the latter had more obvious disruption potential and was already a polarizing debate stock. Carvana’s origin and relationship with DriveTime: Carvana emerged from DriveTime, with Young Ernie Garcia pitching the concept and leveraging DriveTime resources early on. The company scaled quickly using DriveTime’s infrastructure, then gradually weaned itself off that dependence. Hub-and-spoke logistics as a competitive edge: He argues Carvana’s centralized inventory/reconditioning model enables national selection, better route density, lower per-unit logistics costs, faster delivery, and a better customer experience than dispersed dealership lots. Customer-friction removal and product-market fit: Carvana’s online presentation, quality control, return policy, instant online selling, and no-haggle model reduce friction versus traditional dealers. These features help convert buyers who dislike dealerships and make the digital purchase process workable. Unit economics, scale, and growth drivers: He breaks down Carvana’s margin improvements into market-network effects, economies of scale, better source mix, higher IRC utilization, better logistics utilization, and lower days-to-sale. He sees these as continuing tailwinds, though many low-hanging gains are already being harvested. Valuation, dilution, and risk discussion: Despite liking the business, his 10-year model values Carvana around $190-$195 per share, below the then-current market price. He views future capital raises as manageable and mostly expected, but sees the biggest risks as lower-than-expected unit growth and intensified competition.

Key Arguments: Carvana operates in a fragmented, consumer-frustrated market where online penetration is still low, leaving room for share gains. The company’s hub-and-spoke model lowers logistics costs and improves delivery speed by concentrating inventory and reconditioning in large IRCs. Customer trust is built through transparency, inspection standards, and a return policy, which are essential for online car buying. The ability to sell a car quickly online and have it picked up at home is a major differentiator and hard to replicate at scale. Data on buyer/seller behavior lets Carvana arbitrage inventory across geographies and source vehicles where they are cheaper or more available. Gross profit per unit is helped materially by sourcing more cars directly from consumers rather than wholesale auctions. Scale should reduce advertising, technology, and headcount costs per unit, while mature markets should benefit from higher utilization and more word-of-mouth. Wholesale is still a small part of the business, but first-party wholesale auctions could meaningfully improve margins if Carvana reaches enough scale. Customer financing, service contracts, and gap coverage are important profit contributors, but should grow more in line with unit sales as they mature. The biggest fundamental risk is not a fraud narrative, but whether e-commerce adoption and Carvana’s share of it are high enough to justify current valuation.

Data Points: Carvana retail share of U.S. used-car market (assumption in base case): 4% in 10 years - Used in 10-year valuation model as the largest used vehicle retailer Carvana share of U.S. used-car market today (at time of discussion): 0.65% - Referenced as the starting point for growth assumptions CarMax share of U.S. used-car market today: 2% - Used as benchmark for industry leader E-commerce penetration in used-car retail: ~1% last year - Cited to show how early the online channel still is Top 100 dealership share: <10% - Used to illustrate fragmentation of the market Independent dealership competitors: 30,000-40,000 - Estimate of the fragmented competitive landscape Retail units sourced directly from consumers (Carvana): ~60% in 2020 - Showed increasing importance of direct sourcing Retail units sourced directly from consumers (Carvana, peak quarter): 80-90% in Q3 2020 - Quarter with record GPU GPU uplift from consumer-sourced vs wholesale-sourced vehicles: $600-$1,000 higher - Estimate of gross profit per unit advantage Days-to-sale improvement: Reduced by 60 days - He links this to roughly $600 GPU improvement GPU improvement per 10 days reduced: ~$100 - Rule of thumb discussed in the episode Ad spend per unit in mature cohorts: Sub-$1,000 - Older market cohorts had lower customer-acquisition costs Ad spend per unit in 2020 cohort: ~$1,700 - New-market cohort with heavy growth investment New markets opened in 2020: 120 - Illustrates aggressive expansion Increase in new markets in 2020 vs prior year: ~80% - Used to explain elevated ad spend and growth Wholesale share of revenue: ~8% - He described wholesale as a small portion of revenue Wholesale share of gross profit: ~5% - Wholesale is currently minor in profitability mix Retail gross profit per unit on consumer-sourced vehicles: $600-$1,000 higher than wholesale - Key driver of margin expansion Current broad retail GPU outlook: $2,200 base-case by model - Part of his valuation assumptions Implied market price expectations: ~5% share by 2030, $2,400 retail GPU, 8.5% EBIT margin - What the current stock price seemed to imply to him Capital needs over next few years: $1-$2 billion additional external financing - He expects this before self-sufficiency Carvana financing penetration: ~75% of retail units financed - Customer financing is a major profit contributor Interest rate on consumer loans: ~8%-9% APR - Average APR on originated loans Securitized loan yield spread: ~6%-7% lower than APR - Explains gain-on-sale economics Base-case fair value: $190-$195 per share - His estimate using a 10-year model and 8% discount rate Potential downside to valuation: ~30% below then-current share price - He believed the stock was overpriced relative to his base case Implied share gains in model: 25% CAGR in units sold for a decade - To become the largest U.S. used-car retailer Market share assumption for e-commerce used-car retail: ~25% share for Carvana of 15%-16% channel penetration - Implicit assumption in his model

Pivotal Quotes: "If you can't explain to a fifth grader, then you probably don't quite understand it yet." — Host: Discussing how writing about a company is a useful investing filter "I think the number one reason, at least originally for starting the 10th man, was to crowdsource the devil's advocate." — 10th Man: Explaining why he began publishing his research publicly "There's a known, like unknown." — 10th Man: Summarizing the management/reputation risk around DriveTime and the Garcia family

Implications: The discussion suggests Carvana may be a real operating improvement story, not just a meme stock, but the market could still be overestimating how big and how fast the opportunity becomes. Investors must weigh strong unit economics against valuation, dilution, and competition.