Description

Stand-alone auto lenders today trade at 6x NTM EPS. Franchised auto dealers trade at 5-8x NTM EPS. But if KMX is any indication, a concoction of the two is apparently highly synergistic, for the stock trades at 25x estimates that we think are way too high!

Bulls would object to our tongue-in-cheek intro and remind us that franchised dealers are exposed to the more economically sensitive new car industry and are overearning on new vehicle GPU, with KMX deserving a premium multiple as (1) GPUs are in line with history, (2) industry unit demand is at a cyclical trough, and (3) KMX is making no money on an ex-CAF basis (CAF is its financing segment). Not to mention KMX is an e-commerce-driven disruptor / category killer, right? Bulls throw out a $5 EPS power a few years out as a reasonable mid-term normalized EPS  

We think KMX is a short and see way more attractive risk/reward profiles elsewhere in the downstream auto space. We would counter with the following points:

1. More than 100% of KMX’s earnings power on a run-rate basis is from CAF, which is meaningfully overearning. To frame this point, consider retail comps in FY22 were down 7% vs. 2019 but CAF income is 45% above. CAF profitability has started to roll hard, with 4Q22 CAF income only up 11% vs 2019, (albeit still comfortably higher vs retail comps down 17%). We see high visibility into a sharp reset given NIMs today are still running in the 6.0-6.5% range (benefitting from record spreads during the CY20-21 period, for reasons we can discuss in Q&A) vs ~3-4% NIMs implied by its 4 most recent ABS issuances. Mark-to-market would suggest net interest income falling ~33%, with CAF EBIT falling even more as provision for loan losses continue to rise. We model a more tempered reset (20-30% declines in FY23) as KMX’s accounting will flatter its underlying earnings power, but see pressure continuing into FY24 as FY20-21 loans are flushed out of its receivables
2. We think 2019’s ~40.2M industry used unit sales is more likely peak demand than not normalized/mid-cyhcle demand. Demand has reset >20% but there are both supply and demand dynamics that are working against KMX for the next few years and will make a return to 2019 industry unit sales challenging:
   • Demand is impaired because affordability; although used car prices are off 10% vs peak, they still remain up >50% vs pre-pandemic (per Black Book). g
   • On the supply side, used car availability is expected to remain constrained, with projections by Cox for wholesale auction volumes to be ~20-25% below 2019 levels in FY24 (vs 30-35% below in FY22), arguing for stronger for longer used car prices (and limited relief to affordability). Specifically for KMX, its vehicle sweet spot is 1-4 year old cars; given ~17M SAAR in 3-4 years pre-pandemic and the ~14M avg SAAR over the ’20-22 time frame, this rolling car parc is expected to be ~15% below 2019 levels in 2024.
3. KMX’s historical valuation paradigm should no longer apply going forward. Its competitive moat / value proposition has been eaten away by price transparency (given the advent of third party aggregators) and by its commitment to maintaining a narrow band around ~2,200 retail GPU (which is entirely strategic, but is now resulting in share losses as peers let their excess GPUs normalize). This dynamic, coupled with a lower unit growth algo (4% vs 8% a decade ago), warrant a more pedestrian multiple
4. The main reason KMX earnings are under pressure is because of structural fixed cost investments related to KMX’s omni-channel investments, not cyclical pressures. Compared to pre-pandemic total retail sales including new store openings are down LSD% vs SG&A (incl. D&A) up ~20%, with most being fixed in nature as KMX tried to maintain e-comm leadership vs digitally native peers. The SG&A detail provided by KMX supports this view, with other overhead accounting for the lion’s share of SG&A growth. KMX has started to find religion on cost control, but SG&A is still growing for ~LSD% KMX in FY23... this is an improving trajectory, but far from commendable in retail in a contracting market
5. Used vehicle demand has not troughed, yet Street assumes some underlying recovery as we move through the year.  Estimates for 2023 imply retail comps ~9% below FY19, but comps -pre-SVB were still running =>HT/low 20% vs pre-pandemic. We expect KMX’s lending partners (30-40% of unit sales) to tighten lending standards and hurt KMX comps, similar to what happened in CY17 when Santander Consumer pulled back. KMX comps should also come under pressure as it mixes away from Tier 3 (deep subprime) type customers. Used auto monthly payments are also on the rise with both APRs and used car prices headed higher.

We think comps contract MSD%+ in FY23 and that coupled with CAF pressures and modest SG&A growth yield EPS closer to $1.90 – 2.15 vs Street’s ~$2.75, leaving KMX trading at 35x vs its constituent parts (auto dealers and lenders) at single digit P/E multiples. We see fair value at $50/sh, or 25x trough EPS. We think the squeeze following Q4 results offers an attractive entry point, as estimates for FY23 only moved down a little bit and are still (conservatively) 15-20% too high.

Looking out to FY24, we think KMX comps have to be running 8-10% above ‘’19 levels for KMX to be run-rating $5 EPS earnings power given our assumptions for CAF and SG&A. Should that prove feasible we see KMX fairly valued at best as it would need to trade at a HT multiple on an undiscounted basis for the stock to look attractive here.

Catalyst

Deteriorating weekly vehicle sales and auto credit trends as full effect of soft tax refund season and APR hikes impacts demand further