Description

We believe all available KPIs point to a clear inflection in CARS’ fortunes. With undemanding assumptions, we see at least 100-150% upside over the next 12-18 months with very little downside in a bear case scenario. We start wondering whether we are missing something here and are therefore grateful to hear the bear case on the stock. For background info on the business and on its separation from Tegna, please see mip14 long write up from July 2017 and lars excellent short write up from November 2018. As all information about the business and the industry can be found there, we’ll keep the write-up short and sweet, focusing only on the investment thesis.

Summary investment thesis

Following 3 very difficult years since the separation from Tegna, CARS appears to have finally put the house in order. As predicted in lars excellent write up, the stock cratered some 75% since the 2018 highs as the negative thesis on the stock played out:

1.   A mature industry disrupted by CarGurus led to price erosion and loss of customers for CARS

2.   Affiliate agreement conversion that were masking an erosion in the company fundamentals were not enough to offset underlying headwinds

3.   The stock was a “special situation” hedge fund hotel with many event driven funds playing the spin-off re-rating or the Starboard led campaign to get someone to buy the company

Fast forward to March 2020 and all the above headwinds are finally turning into tailwinds:

1.   CarGurus is slowing down and CARS is regaining market share

2.   The last affiliate conversions occurred in 2019 and from Q1-20 the business is fully direct

3.   CARS is now a hated, orphan stock poised for substantial re-rating as fundamentals stabilise

Traffic stabilising

The single most important leading indicator in a 2-way online platform is traffic. CARS sells a subscription service to car dealers that try to earn a positive ROI on their subscription costs by converting leads into profitable transactions. While we have no information on leads nor on the quality of the leads that CARS provide (conversions), it’s reasonable to assume that these correlates nicely with traffic. The higher the eyeballs coming to a website, the higher the number of monetizable leads for dealers. In turns, higher leads should lead to higher ROIs, justifying the cost of the service. It is common knowledge that under Tegna, CARS was milked for cash and under-invested in its platform. As a result, the older platform was clanky and was penalised by Google in organic searches. CARS difficulties emerged at a time when CarGurus organic search prowess thanks to their SEO capabilities made them the largest and fastest growing online classified platform for 2nd hand cars.

Between 2016-17, as CarGurus was growing into a powerhouse, CARS traffic turned negative with obviously negative consequences on the ROI CARS generated for its dealers. CARS dealers responded with increased cancellations. Adjusting for affiliate dealers’ conversions, beginning in Q1-18, CARS exhibited 7 straight quarters of declines in number of dealers.

Things started to improve though on the traffic front in 2018. CARS management invested in a replatforming effort that finally delivered the expected results. Traffic growth consistently accelerated since Q1-18:

The replatforming clearly worked and CARS is now the fastest growing online classified in the industry. CARS reversal in traffic is even more impressive if we compare it to CarGurus:

Since Q1-19, CARS is now growing its traffic faster than CarGurus. This is not a coincidence. In March 2019, Google made a major change to its algorithm, affecting players that overly relied on aggressive SEO capabilities, such as CarGurus. The change in CarGurus traffic post Google change was material:

CARS’s ability to outperform CarGurus, and other competitors, post Google algorithm changes should not come as a surprise. CARS still has the strongest brand awareness of all used car classified properties. This is confirmed by the fact that it has the highest proportion of direct traffic of all competitors:

Players with good brand awareness playing by the Google rules did well post Q1-19. Players that in the past abused the system (e.g. CarGurus) were penalised. Based on Comscore data, CARS is massively outperforming all peers in terms of traffic growth since the replatforming took place and Google changed its algorythm:

The above data comes from Cosmscore and is very similar to the data showed by CARS in its latest presentation:

In Summary: we think that the most important KPI, traffic, has now inflected positively. CARS will continue to gain share over CarGurus and may soon overtake it in terms of overall viewership.

Number of agents stabilising

The most important KPI in a subscription business is probably the number of subscribers. As discussed above, beginning in Q1-18, CARS consistently lost dealers. Their cancellations were simply running way ahead of new customer wins. In our view, there were 3 reasons for this:

1.   Traffic – CARS was losing share in traffic for the reasons described above. Dealers responded cancelling their subscription. As discussed above, this has now fully inflected and traffic is growing rapidly

2.   Affiliate conversions – In the 3 years to Q3-19, CARS lost over 3,000 dealers. We reckon that over 1,500 of these were previously affiliate dealers that were lost during the affiliate conversion. Customers that were previously followed by the likes of Tegna or tronc, found themselves overnight without a sales rep, or at least without a familiar one. Out of the c, 7,000 affiliate dealers, over 20% jumped ship in the affiliate conversion process which was significantly mismanaged by CARS management. However, the conversion is now 100% complete and churn should naturally fall therefore

3.   CarGurus – As just mentioned, between Q3-16 and Q3-19, CARS lost over 3,000 dealers. Over the same period, CarGurus added over 10,000 net new dealers. It’s clear that many CARS dealers jumped ship and moved to CarGurus. The heightened churn for CARS was also the result of CarGurus aggressive growth. CarGurus offered CARS an attractive proposition:

a.       Higher traffic / growth – as discussed above, CarGurus was the fastest growing player, overtaking CARS in terms of traffic. This is no longer the case though; CARS is now the fastest growing player in terms of traffic and is about to overtake CarGurus

b.       Cheapest offering – CarGurus had an average price per dealer significantly lower than CARS. CARS dealers that jumped ship had an immediate cost saving in the move. This is no longer the case. Over the period discussed (3 years to Q3-19), CarGurus average revenue per dealer increased 71% or 20% pa, effectively erasing CarGurus competitive advantage

As the 3 headwinds described above turned into tailwinds, the number of dealers for CARS finally started to grow in Q4-19. Even adjusting for new dealers buying only Dealer Inspire, CARS added c. 150 net new dealers in the quarter and confirmed that both in January and in February 2020 they saw net new dealer additions.

In summary: number of new dealers has finally turned positive. We think the inflection observed in Q4-19 will be followed by further growth ahead. We conservatively assume 50 net new dealers per quarter going forward:

ARPD to stabilise

Q4-19 results showed the positive inflection in number of dealers but also exhibited an unexpected weakness in reported ARPD, down 5% YoY and down 2% sequentially. Management explained that the weakness was predominantly a function of adverse mix from lower ARPD affiliate dealers that got converted in H2-19. This mix effect will only affect reported ARPD in H1-20 and should stabilise in H2-20. Furthermore, management explained that ARPD will stabilise in H2-20 also thanks to the introduction of new features, like Fuel, that are ARPD accretive. While given management track record we tend to be rather sceptical, in this case we believe that ARPD can really stabilise in H2-20 and from 2021 onwards. The other reason for believing that ARPD will stabilise is again CarGurus.  Given the huge price discrepancy between CARS and CARG in the past (in 2016, CARS was 2x more expensive than CARS), it was impossible to increase prices for CARS. In fact, on the margin, CARS was forced to offer discounts to disgruntled customers. However, as per charts below, the headline ARPD gap has now nearly disappeared:

We believe CarGurus will have to continue to increase its prices significantly if it wants to continue growing. In the last quarter, CarGurus added only 298 new dealers in the US, an annualised 4% growth. If CarGurus wants to continue growing double digits in the US, it will have to continue raising its prices double digits going forward (+19% in Q4-19). It won’t take too long before CarGurus will be, even on a headline number, more expensive than CARS. Furthermore, CarGurus needs to have higher prices if it wants to achieve its margin target of 30% from c. 15% today. CarGurus prices are simply too low today. We view the market evolving from a fierce competitive battleground with CarGurus disrupting the incumbents to a cosy oligopoly where all players will need / want to maintain price stability as there is no more room to grow in terms of number of dealers.

In Summary: we believe ARPD will stabilise in the near term. Weakness experienced in the quarter was due to a mix effect and CarGurus competition is changing from a headwind to a tailwind.

Summary modelling assumptions

For simplicity, we divide the business in 3:

1.   Marketplace revenue – simple subscription business where revenue is derived from number of dealers’ times ARPA. All metrics discussed above refer to marketplace revenue

2.   Dealer Inspire Revenue – this is effectively a SaaS business for dealers that was purchased in 2018. It’s growing in the mid-20s and we expect acceleration in 2020-21 thanks to the GM certification that “brought” with it 800 customers today and up to 4,000 in the future

3.   “Other” revenue – this includes OEM brand (display) advertising and other revenues

Our summary revenue forecasts are the following:

1.   Marketplace revenue – we expect ARPD stabilisation in H2-20 and very moderate growth in net new dealers. Revenue should stabilise in 2021 but we see upside optionality here if they keep adding 100-150 dealers per quarter

2.   Dealer Inspire revenue – revenue slowed down to 20% due to delayed certifications but should re-accelerate in 2020-21

3.   Other – this business is cyclically depressed. We estimated -16% in 2020 but flattish thereafter. It could cyclically rebound in 2021

Regarding margins, the company guided for very low EBITDA margin of 25-27% in 2020. This is primarily due to 2 one-off issues that will pass in H2-20:

·         C. $12m in affiliate revenue share for older affiliate contracts

·         C. $10m in platform / tech cost to onboard GM dealers on Dealer Inspire in H2-20

Because those 2 items affect H1-20, margins in H2-20 will be above 25-27%, trending towards 30%. We see no reason why EBITDA margins should be lower than 30% in 2021. Resulting EBITDA in 2021 will be c. $185-190. At that point in time (2021), CARS will be a growing business again, having reduced leverage to 2-3x and should then regain investors’ confidence. Applying a very low EV / EBITDA multiple of 7x (the lowest ever in CARS history) we get a share price in the high teens for at least 100% upside. We wouldn’t be shocked to see a sponsor snatching this asset in the $20s for a near triple from here.

Vice versa, if the company fails to grow from here remaining a value trap, it would still generate c. $90-100m in fcf per year over a market capitalisation of just over $500m. It’s hard to see much downside from here.

Catalyst

• Continued net dealer additions in Q1-Q2 2020
• ARPD to stabilise in H2-20