|Shares Out. (in M):||70||P/E||0||0|
|Market Cap (in $M):||1,837||P/FCF||0||0|
|Net Debt (in $M):||0||EBIT||0||0|
|Borrow Cost:||Available 0-15% cost|
Summary: Cars.com is a structurally disadvantaged business that is over-earning and experiencing severe competitive pressure. Additionally, its affiliate agreements and M&A have been masking an accelerating structural decline in the company, and those effects should normalize over the next several months. Lastly, while there was initially a lot of bullish “special situation” attention paid to the company and the equity, I believe that sentiment is waning. Recommendation is to short the equity of Cars.com with a target in the mid tens.
Cars.com is one of two main legacy used automobile online classifieds sites (Autotrader, now owned by Cox Automotive, being the other) that were started in the late 1990s. Both companies were founded out of legacy print operations (Cars.com as division of Classified Ventures, a joint venture between eight media companies, and Autotrader as part of Cox Enterprises).
1) Structurally disadvantaged model in decline:
First, I believe that the market opportunity for Cars.com is largely tapped out. Management frames their market opportunity as 43k auto dealers in the U.S. (they have approximately 21k as customers) and the $37b automotive advertising industry (of which, the company touts, only 56% is online). In terms of customer count, despite being only ~50% penetrated into that 43k number, I believe that Cars.com has fully penetrated their addressable market. The auto dealer industry, and particularly the independent used auto dealer industry, is extremely fragmented. Cars.com’s model lends itself to franchise dealers with large used operations, or the larger independent used dealers. These businesses have larger budgets and are typically more willing and able to spend on digital media. The long list of smaller independent dealers not only have smaller budgets and less sophisticated marketing capabilities, but are much more expensive for Cars.com to reach and convert with their salesforce. With respect to the tailwind of offline to online spend, while the aggregate numbers show an opportunity, I think the opportunity has been largely played out within Cars.com’s target market. Multiple dealer conversations and survey work have lead me to believe that the shift to online has largely occurred already at most larger dealers and if anything, there is on the margin a share-of-wallet shift from classifieds to search and social channels.
Next, and most importantly, the industry structure for Cars.com, while historically very favorable, has gotten more competitive over the last ten years, and materially more so over the last five. Given the local monopoly position of their print media origins, the used auto online classifieds business was for years a very stable and profitable duopoly. Conversations with industry executives who ran these businesses in the 2000s provide useful anecdotes of how the business was able to push through price increases with impressively consistent size and frequency in the first ten years of existence. Over time, as the revenue and margin levels at Cars.com and Autotrader reached high levels, other players that had existing audiences (Kelley Blue Book, Edmunds, etc.) got into the classifieds listing business, making it marginally more competitive. The real game-changer though, has been within the last five years as CarGurus has entered the market. CarGurus’ entrance was unique in that while most entrants into this market have struggled with the chicken-or-the-egg problem with inventory and audience, CarGurus a) created its site to index extremely well for SEO (do a search for ten variations of make/model/year to see – I just did one as I’m writing this and CarGurus was #1 on SEO ten out of ten times), and b) they gave the product away for free allowing them to build inventory very quickly. Importantly, because CarGurus does not have a branded advertising business (which is punished in organic search), I believe their organic search prowess is a sustainable competitive advantage (cars.com has a ~$100m branded advertising business, that while in structural decline, is too profitable to jettison in order to try and help their overpriced listings product). Not to mention the CarGurus folks are just very good at SEO; the C Langley Steinert is from TripAdvisor which is widely regarded to be one of the best Internet companies at SEO.
Below are two charts showing how successful CarGurus has been in growing both sides (audience and dealers) of their network; they have the largest base of both, which I attribute to a) their SEO prowess, b) their “freemium” go-to-market strategy, and c) their best-in-class product (all the products in the market are pretty similar, but I believe CarGurus has the best functionality and that you can see how others have been followers of their attributes (like the “Great Deal, Fair Deal, etc.” functionality).
Slide wouldn't paste - go to slide #5 in the following deck - https://investors.cargurus.com/static-files/e2a1d5ce-bb22-438c-a761-5537717e19c3 - showing CarGurus audience size.
Slide wouldn't paste - go to slide #12 in the following deck - https://investors.cargurus.com/static-files/e2a1d5ce-bb22-438c-a761-5537717e19c3 - showing CarGurus paying dealer base size and growth.
Importantly, CarGurus is still significantly cheaper than Cars.com. First of all, Cars.com reports their revenue per dealer statistic as their direct retail plus wholesale revenues divided by average dealer count. I believe this is misleading because it a) includes the amortization of non-cash revenue that was created in 2014 when they stepped up the value of their affiliate agreements on their balance sheet, and b) the conversion of their affiliate agreements where they realize 60% of retail rates to retail agreements. In the first quarter, management cite this metric as up 6% YoY, implying nice pricing power / incremental product penetration. When you adjust for the aforementioned factors (taking out the non-cash revenue and then grossing up affiliate revenue to retail to get an apples-to-apples comparison), the metric was flat. Most importantly though, is that when you make those adjustments to reflect cash retail rates, their average annual revenue per dealer is approximately $26k, which is double CarGuru’s current average annual revenue per dealer of $13k. I believe this is why Cars.com’s dealer count (legacy, not including acquired Dealer Inspire dealers) declined in 2017 at 1.3%, declined in 1Q18 at 5%, and declined in 2Q18 at 5.8%; I believe that as CarGurus raises their price / penetrates dealers with higher-priced product / functionality, they are stealing wallet share from high-priced legacy listings providers like Cars.com and Autotrader. While right now that is manifesting itself at Cars.com in flat pricing and dealer count declines, I think that over time they will not only lose dealers, but they will also see pricing converge to CarGurus. Given that CarGurus is increasing its AARSD, I think it’s reasonable to think that rates will converge around the $20k level in a few years. From a high level, this makes sense; it was a duopoly for a very long time that experienced a lot of pricing power, it slowly got more and more competitive, and then over the past few years a very disruptive entrant has emerged and is rapidly driving excess returns out of the industry.
2) Affiliate agreements and M&A are masking an accelerating decline in fundamentals:
Cars.com wholesale revenue reflects the company’s affiliate agreements. In the brief “Background” section at the beginning of the write-up I mentioned how Cars.com began within Classified Ventures, a JV between several traditional media companies – the affiliate agreements reflect sales where the Cars.com product has been sold through the sales forces of those traditional media companies in their geographies. These affiliate agreements were set to expire in 2019 and 2020, but the company has been negotiating early conversions to expedite this process. Affiliate agreements are basically revenue recognized at 60% of retail rate, so as they are converted to direct sales, it increases revenue for the same sale. While this is hardly an unknown phenomenon, it does obscure an apples-to-apples comparison of business trends over time, and management doesn’t provide much clarity, in fact as I mentioned previously they report average revenue per subscribing dealer without normalizing for the change (i.e. the 6% they might report is not indicative of the pricing power in the business). In order to get a clear picture of what the business trends are, one needs to remove the non-cash amortization of the contract liability from the affiliate revenue line, and then gross up to retail rates. Once you do that you see a picture of an accelerating dealer decline (down 1% in ’17, 5% in Q1 and 6% in Q2) accompanied by a flat to negative (down 3% in ’17, slightly down in Q1, and slightly up in Q2) trend in pro forma direct revenue per dealer.
The margins of the business are similarly obfuscated by this affiliate to direct shift, as well as the purchase of Dealer Inspire earlier in the year. One of the tenets of the bull thesis on CARS is that this shift from affiliate to direct will come at a very significant positive impact to profitability. While current management supports this idea, I’ve spoken with members of the prior management team who are skeptical that this will occur, due to the necessity of hiring and maintaining sales resources in the affiliate regions. I believe that the lack of leverage in this shift is already showing up, but it’s being obfuscated by 1) the Dealer Inspire purchase, and 2) the shift in non-cash amortization of the contract liability from the revenue line to a contra-expense in the affiliate revenue share line. On the first, one can pull the Dealer Inspire contribution that is disclosed in the Qs out of the Marketing & Sales line to see that the core expense is rising as the company hires new staff to support the affiliate regions. While one could argue that this increase is less than the discount to the retail rate, I would say that 1) this expense is just beginning to build, and 2) some is still in the affiliate revenue share line. That brings me to my second point, the affiliate revenue share line. As contracts are converted to direct, the contract liability that was established in 2014 shifts from a non-cash benefit to Affiliate Revenue to a non-cash benefit to Affiliate Revenue Share as a contra-expense. While the total Affiliate Revenue Share expense line looks fairly flat YoY, when you pull out the growing non-cash contra-expense portion, you can see that it’s been rising at a good clip over the past few quarters. This is because as the company converts the McClatchy regions, it has to pay marketing support payments to McClatchy. I don’t believe this is a temporary expense but that as that arrangement ends, that expense will shift to the Marketing and Sales expense line and continue to grow in the form of additional sales resources to replace the support from McClatchy.
2) “Special situation” sentiment is wearing off:
There has been a lot of interest on the long side in Cars.com because it was a spin-off, and because it is a small-cap, high-margin, capital-light business where the largest private competitor has historically changed hands at attractive valuation multiples. On the last part, I don’t believe that the 2010 and 2014 Autotrader transaction multiples are relevant to Cars.com. Those transactions occurred in a pre-CarGurus environment, and speaking with Cox executives will give you an idea as to how much that business has deteriorated since then. I have been told that Cox is devoting no resources to that business and is basically letting it run off, and I think this is somewhat reflected in the complete focus of their M&A strategy on dealer-related software over the last few years. I think this is why after rounds of rumors that Cars.com was an acquisition target and that they had hired JP Morgan to sell themselves, no transaction has occurred.
Financials & valuation:
I think that the downside to a short in CARS is a 12x EBITDA multiple, or mid-high single digit FCF yield on consensus numbers. Consensus numbers reflect much of the bull thesis around the affiliate conversions, yet that still doesn’t change the fact that CARS is largely tapped out in its addressable market with an inferior, high-priced product. That downside is 20-25% to a stock price in the low $30s.
I think that the upside in a CARS short is 8x an EBITDA number closer to $200m, which would yield a stock price in the teens; substantial downside. That $200m is built through 1) continued dealer declines at the current rate (-5%), 2) flat to slightly down revenue per dealer (on an apples-to-apples basis / adjusted for the conversions), and 3)what I believe will end up being the realistic level of Marketing & Sales expense (~$275m – includes the build out of resources for all conversions, so the current ramp, as well as replacing the McClatchy support payments when the Affiliate Revenue Share line disappears). What I think is most compelling about this build is that I believe the $200m is still a melting ice cube. Right now CarGurus is driving a massive shift in wallet share to themselves from legacy providers (Autotrader, Cars.com, KBB, Edmunds, etc.) with a better product, a larger and still growing audience, at a much lower cost to dealers. Not only do I believe that this will result in CARG continuing to take wallet share from CARS, but I think that it will end up saving the dealers money as well. In other words, I believe that CARS business will likely shrink by some combination of dealer loss, as well as having to meet CarGurus in the middle on price, and there’s a lot of negative leverage for them between $26k per year per dealer (where CARS currently is on a retail / cost to the dealer basis) and $13k per year per dealer (where CarGurus currently is). If in a few years, if CARS and CarGurus are similarly-priced at $22k per year per dealer, Cars.com is a much smaller business with EBITDA margins in the low 20s. That seems like a plausible scenario to me for an undifferentiated Internet company in a mature and highly competitive market. I think that scenario is why we haven’t seen a sale of the business, as has been speculated in the media and among bulls on the stock.
Acquisition: There have been sale rumors in the past, but they have not culminated in a transaction, and the activist shareholder (Starboard Value) has begun selling, per SEC filings.
Continued M&A: Continued M&A could buoy #s, however the DI & LDM deal was done at a 12x multiple (was pitched as 9x, but all 3 turns of improvement are from revenue synergies). In my opinion, CARS does not have anything unique to offer potential targets versus the likes of Cox, CDK, etc.
1) Dissapointing paying dealer numbers.
2) Dissapointing cost / margin performance.
|Subject||SEO vs. SEM|
|Entry||11/12/2018 03:11 PM|
Lars - as a follow up to Mencken's comments, you mentioned CARG's strength in SEO. Is it SEO or SEM? I think of the two very differently. I'm looking at the financials for CARG vs. CARS - in the last quarter, CARG spent 71% of revenue on marketing vs. CARS at 33%. I agree that CARG spending these levels will be a negative for CARS, but is it a moat any more than TRVG's SEM spend was in hotels (90% of revenue which for a brief moment in history drove +90-100% revenue growth)? Again trying to assess the moat - companies can grow rapidly as a vertical-specific lead gen company if they are willing to bleed SEM $ and buy traffic. That will drive more rapid revenue growth (and pressure competitors to spend more to hold share) but won't drive value unless the customers they are acquiring are sticky and will return directly to the site for their next purchase (skipping the need for CARG to spend on SEM to re-acquire the customer). Given car purchases/sales are so episodic, I wonder if the customer's next purchase/sale again begins in Google search. Thanks for any thoughts.