Description

I believe owning Hertz presents an attractive risk-reward opportunity with substantial upside.

As one of the most heavily shorted US equities with 47% of float sold short, the base case appears to be that Hertz is going to zero. For more on this viewpoint, please see bedrock346's short-selling pitch from August of last year.

I believe the risk-reward at current prices makes going long a better bet than being short. To support my view, I will endeavour to show data that contradicts the short thesis.

The short-thesis on the auto rental names is centered on three key arguments: 1) ride-sharing companies will destroy them; 2) the used car market is at its peak and in the subsequent downturn, auto rental names will suffer from ballooning fleet costs; and 3) the future of transportation will be autonomous, and no one will rent cars anymore. For Hertz, there is also its history of severe mismanagement and the risk of failure in its current attempt at a turnaround.

I believe the primary argument for the short thesis is #1 -- that ride-sharing companies will destroy auto rental companies. I will focus on this argument in this pitch, and spend only a little time addressing #2 and #3. If ride-sharing companies are *not* going to destroy Hertz, then I believe the valuation and potential upside is great enough to offset the other bear theses and substantial execution risks.

Anecdotally, it certainly appears that the ride-sharing companies like Uber and Lyft must be destroying auto rental. For investment professionals who predominantly live in NYC, SF, LA, Boston, or other urban markets, and can pay for the convenience of an Uber, renting a car seems almost inconceivable (see bedrock346’s pitch). It is easy to extrapolate our own experience to the broader market and conclude that these companies will be destroyed by ride-sharing startups in the same way Blockbuster was destroyed by Netflix. In addition to our own experience, there are plenty of news headlines that support this view. For example, the cloud-based business expense company Certify has been publishing its SpendSmart report for some time, which has led to charts like this:

The headlines are dire. For example, “Startups like Uber decimated taxi companies. Rental cars are next.”

https://qz.com/1253717/turo-is-doing-to-rental-car-companies-what-uber-did-to-taxis-and-theyre-scared/

Or, “Uber and Lyft pound taxis, rental cars, in business travel market.” https://www.forbes.com/sites/michaelgoldstein/2018/02/22/uber-and-lyft-pound-taxis-rental-cars-in-business-travel-market/#3d4131e6b5e7

Or, “101-year old car rental business nearing twilight in Uber age.” https://www.denverpost.com/2017/08/13/hertz-uber-lyft-car-rental-businesses/

Or, “Ridesharing causing rental car industry collapse.” http://www.genfkd.org/ridesharing-causing-rental-car-industry-collapse

Or simply tweets like this from venture capitalist Hunter Walk:

https://twitter.com/hunterwalk/status/1018708784812720130

Indeed, the impending demise of auto rental is a widely held notion.

But does the data support these headlines and our own anecdotal experience?

Here is total US auto rental industry revenues over the past 10 years compared with GDP (auto industry revs in billions, GDP in trillions; sources Auto Rental News and the US BEA):

Over the last ten years, auto rental industry revenues have grown at a 2.9% cagr, and gdp has grown at a 3.0% cagr. Uber went live in SF in July of 2010, and focused on national expansion through 2012 when it launched its international expansion. From 2012 to 2017, auto rental industry revs have grown at a 3.9% cagr, vs. gdp at a 3.8% cagr. Is that what decimation looks like? In fact, auto rental revenues have outgrown gdp in five of the last seven years, with the exceptions being 2013 (3.3% vs. 3.6%) and 2017 (0.7% vs. 4.2%). And so far in 2018, industry revenues have again outgrown gdp. This is hardly consistent with the narrative that Uber and Lyft are “decimating” or “pounding” auto rental.

So what is the source of the discrepancy between the popular bearish narrative and actual industry results? Well first, corporate spending is just a modest fraction of overall auto rental revenues -- for Hertz it’s about 30% corporate, 70% leisure, and for Avis, about 25/75. Second, both Hertz and Avis have told me that corporate spending is where they are seeing the greatest erosion from ride sharing companies and that leisure travel has not seen the same erosion. One reason for this is that corporate customers are less price sensitive than leisure ones. In addition, and importantly, the Certify SpendSmart data captures transactions as measured by number of receipts, not revenue. So a single day on the road for a corporate customer could result in many reimbursement receipts for Ubers taken, but only a single receipt for a rented car. Furthermore, before Uber, multiple colleagues flying into the same airport would wait for each other and then ride in the same rental car. Now they often aren't waiting or coordinating, and are each taking his or her own Uber. This is multiplied if the business trip takes multiple days -- still only one rental car receipt but that many more Uber and Lyft receipts. So one can easily see how the number of expensed Uber trips is exploding versus the auto rental transactions they are replacing. Hence the comparison is not apples-to-apples, and the conclusion one naturally draws from the “Uber and Lyft crush taxis and rental cars” charts is quite misleading.

An analyst at Bloomberg recently explored the Certify data and expanded upon the mistaken conclusion it suggests in an article with this very long title: “No, rental cars aren’t about to disappear: They face challenges but aren’t being wiped out by Uber and Lyft just yet. Here’s the story of a chart that was too stunning to be true.” https://www.bloomberg.com/view/articles/2018-08-02/uber-and-lyft-are-not-making-rental-cars-disappear

The Bloomberg analyst finds that rather than a dramatic 33% loss of corporate market share over four years, the actual loss in revenue was likely closer to 13% over that time period. See the following two charts from his article:

Versus:

Certify has recognized that its data presentation was leading to faulty conclusions, and has decided to stop issuing this report, with the CEO noting that the reports “didn’t sound quite right.”

Additionally, there’s the likelihood of selection bias in Certify’s customer set, as noted by the Bloomberg analyst -- Certify’s corporate customers are more likely to be more tech-oriented than for example, its competitor Concur, which is owned by SAP.

So, industry topline data certainly does not support “decimation,” and on closer inspection, corporate spending data, such as Certify’s, leads to the conclusion that a portion of corporate spending is being lost, but not as much as commonly reported.

Indeed, Hertz and Avis have said for some time that their average transaction is much longer than the average Uber ride, and the overlap of similar rides is relatively small. I spoke with both companies to get an update as to their best estimate for how much of their business is vulnerable to the ride-sharing threat. They reiterated the following stats:

For Hertz, the average transaction is for 5 days and 588 miles.

For Avis, the average transaction is 4 days and 450 miles.

Using Ubers for the average auto rental transaction would be much more expensive than using a rental car. If on average an Uber ride costs $2/mile, using Ubers would cost almost $1,200 for the average Hertz trip. Meanwhile, Hertz makes about $41 in revenue per day per car, or about $200 for the whole transaction. Adding in ~$70 for fuel (588 miles/25 mpg*$2.84 ave price of gas), plus taxes and ancillaries, and you still have a cost that is around one quarter of the cost of using Ubers, plus you have the guaranteed availability of your vehicle. Hence, there remains a significant economic advantage to renting. While this may not matter to investors on this idea exchange, it does matter to many consumers and businesses. Furthermore, some portion of this economic difference is structural, because for an Uber, you will always will have the extra expense required to compensate the driver, and the lack of guaranteed availability of a car.

If the data shows that ride sharing isn’t decimating auto rental, what is its actual impact?

The portion of business that Hertz considers most vulnerable to ride-sharing are those trips that are less than 1 day or less than 50 miles. For these trips, renting a car may have a cost disadvantage, since you’re paying for a whole day when the trip itself may cost less. If an Uber is readily available, renting a car will also have a significant time and convenience disadvantage as well. As a percent of volume, 1-day trips for Hertz are about 5%, and less-than 50-mile trips are also about 5%, with 1-2% overlap. Hence, about 8% of Hertz’s volume is what the company sees as most vulnerable to losses to ride-sharing. However, these shorter trips are also the least profitable for the company, since they require the same level of interaction with the customer and vehicle as a longer trip. Hence, the actual percentage of profitability that is at risk is lower than 8%.

Avis cites similar stats but adds some additional detail. The company told me that this is the third year they have tracked and bucketed transaction data to try to determine where and how much they were losing to ride-sharing companies. What they have found is that the number of days doesn’t matter, but rather number of miles. They found that the dividing line is 75 miles. Transactions shorter than 75 miles are declining, while trips longer than 75 miles are growing. Six and a half percent of Avis’s volume in 2017 was for trips less than 75 miles, about half a percent less than it was in 2016. Like Hertz, Avis said that these shorter trips were less profitable than longer ones, or sometimes not profitable at all. In their most recent quarterly call, Avis talked about its efforts to actively extend the length of transaction, which in some cases means they decline one-day reservations. In plenty of other cases, Avis told us they take the 1-day rentals because they have to, not because they want to. So the companies are trying to give these less or un-profitable short trips to Uber. It would literally be good for Avis and Hertz if Uber took all those 1-day trips from them!

A valid question is whether this erosion stabilizes after some level of losses, or whether it accelerates as Uber and Lyft densify and spread. I believe that the fundamental difference in economics between an Uber ride, whose natural price floor requires an adequate return to the driver, and that of an auto rental means that even as the ride-sharing networks expand and become more available in regions where auto rental is the only option now, there will remain a value proposition to renting, thus leading to a natural equilibrium in markets where ride-sharing is fully penetrated.

Hertz told me that in some markets like San Francisco, NYC, and DC, where they have seen the greatest share loss, and also where it’s hard to imagine Uber achieving much higher penetration, they have already begun to see revenues flatten, and in some cases rebound and grow faster than airplane enplanements, a sign that in those markets they have indeed stabilized. This anecdotal evidence supports the industry data showing a rebound in revenues in 2018 versus gdp growth, and also points to the improving health of the industry.

The data that Hertz and Avis have shared about the type of transactions where they are seeing erosion versus growth -- the 1-day or sub-75 mile ones -- provides a sense of the potential size of long-term competitive losses. Of course, the companies are losing some business in the longer trips as well, but estimating what might have been is difficult, and even the companies themselves aren’t sure how to precisely measure the counterfactual.

Nevertheless, if the companies are still growing in over 90% of their business by volume, and an even greater percentage of their business by profit, I believe we can reject the argument that Uber and Lyft are “decimating” these companies.

Is there an offset to the market share auto rental companies will lose to ride sharing?

While it’s clear that ride-sharing is having an impact, though a much smaller one than commonly thought on auto rental companies, there is more to the story. Hertz has taken the lead in actually renting to ride-sharing drivers, and Avis is now following suit with their recently announced partnership with Lyft. At present, 5.5% of Hertz’s fleet is devoted to ride-sharing drivers, and growing rapidly. These drivers rent from Hertz at a lower rate than Hertz charges normal rental customers, of around $210 a week or about $30 a day versus Hertz’s average revenue per day of $41, but these rentals also have much higher utilization and lower opex because they are for weeks to months at a time, and Hertz only needs to see the vehicle once a month to inspect and maintain it. Hertz told me that contribution margin on these rentals is very high because of the higher utilization and lower expense. In addition, depreciation is much lower because Hertz is using vehicles at the end of their normal rental lives, with 30k+ miles, when they otherwise would have sold the vehicle, and when the depreciation curve is much shallower than for a new vehicle. Hertz believes it can fully offset its competitive losses to ride-sharing by renting to ride-sharing drivers. I think this is aggressive, but at 5.5% of their fleet, they are already offsetting a material portion of their likely losses, and the company describes this as a “robust and growing channel” for them.

There’s a lot of room to grow for this channel because the addressable market is very large. There are over 1 million Uber/Lyft drivers in the US. Assuming roughly 25% are full-time, and hence the vehicle they use is primarily for the purpose of ride-sharing, it likely makes sense for all of these drivers to rent their car from a large fleet manager like Hertz rather than buy and maintain the car themselves. Indeed, for everything involved in the ownership of their car, Hertz has an advantage. Hertz can buy the car cheaper, it can service and maintain it more efficiently, and it can resell it at a higher value.

Thus, the primary bearish argument is not supported by data. The overall industry revenue data shows that so far, ride-sharing has not decimated the auto rental industry. Corporate spending data shows that the revenue share that has been lost is less than commonly reported or perceived. Furthermore, the transaction data that Hertz and Avis report shows that the types of transactions that are being lost are a small fraction of the total. Finally, Hertz’s early experience in renting to ride-sharing drivers, shows that there is at least a partial offset to the competitive losses that are happening.

The second main bearish argument is that with near peak used car values and a wave of cars soon to come off lease and hit the used car market, residual values will plummet and fleet costs will skyrocket. This may indeed happen. If the auto rental industry operates rationally and keeps fleets correctly sized for demand, it historically has been able to pass fleet cost changes on to its customers. The main reason the industry has had pricing challenges in recent years, particularly in 2016, has been due to Hertz’s mistakes in buying the wrong fleet, and then being forced to put it on rent to recoup whatever they could. Even Hertz’s public competitor, Avis does not believe they intentionally were operating irrationally, but that they were simply inept. Hertz took their pain selling off the wrong parts of their fleet in early 2017, and since then we’ve seen much healthier pricing conditions for the industry. From every comment Hertz and Avis have made on their earnings calls, both companies are buying with the intention of maintaining an appropriately sized, somewhat tight fleet for the demand, which should support the industry’s ability to pass any changes in fleet cost on to customers through price.

In addition, Hertz has been making great progress in improving the prices it receives from selling its used cars, by increasing the number it sells direct rather than through auction where they receive the worst residual values. As noted in their last earnings call, they now sell less than 25% of their cars via auction, while their dispositions through dealer-direct and direct-to-retail channels grew 16 percentage points year over year. These direct sales return hundreds of dollars more per vehicle.  

The third main bearish argument is that in the next few decades, all driving will be autonomous, and there will be no need to rent a car. Hence, these businesses will simply go away. While predicting that far into the future is a fool’s game, I am comfortable with this risk, because I believe that the companies developing self-driving car technology do not also want to enter the boring, yet necessary and specialized work of managing fleets, and that they will partner with companies that have this expertise. https://www.technologyreview.com/s/608176/why-waymos-partnership-with-avis-makes-sense/

We are already seeing evidence of such deals in Waymo’s partnership with Avis from last summer to manage over 600 vehicles in Phoenix. This partnership has recently been expanded to include other locations. Hertz just signed a deal this July to manage a fleet of 75 self-driving cars for Aptiv in Las Vegas, with plans to expand from there. (Aptiv is a spin-out from Delphi that’s focused on autonomous driving technology).  http://www.autonews.com/article/20180731/OEM10/180739904/aptiv-hertz-autonomous-las-vegas-q2-earnings

While it’s too early to place a value on these ventures, the business case for self-driving technology companies to partner with the companies that have the real estate, technology, and know-how to manage large fleets, makes a lot of sense. Hence, should the self-driving vision of the future become a reality, I believe that the auto rental companies will still have an opportunity to exist and play a valuable role.

Execution risk:

If the external threats represented by the three main bearish arguments aren’t likely to kill Hertz, Hertz still could blow itself up, as it has seemed hell bent on doing over the last several years. For example, they flubbed the Dollar/Thrifty acquisition, they moved headquarters from NJ to FL and lost a lot of key middle management, they bought the wrong fleet in 2016 and hurt industry pricing, and inflicted a lot of other damage upon themselves and the industry. However, it appears that so far, the newest CEO, Kathryn Marinello, is doing the necessary blocking and tackling to fix the business.

To see an indication that Hertz is on the right track with regard to execution we can simply look to the trends in key metrics shown in its most recent earnings report:

Revenue, volume, pricing, and utilization have all been trending positively while fleet costs have been declining. Pricing growth is even better than it appears because the growing ride-sharing business has pushed down the average revenue per unit in recent quarters. Improvements in utilization indicate a fleet that is more appropriately sized for demand. In short, Hertz has now been making progress on its turnaround for at least a year.

The car rental business is not rocket science. With decent management, which Hertz appears to finally have, the company should be able to generate reasonable returns.

Valuation:

If Hertz isn’t dying because of external threats, and can return to some normal level of profitability with decent management, the stock has huge upside. Hertz management has indicated that 2018 and 2019 will be years of elevated investment in the company’s turnaround, with a major focus on rebuilding the company’s 30-year old technology platform. While expenses will remain elevated during this year and next, the company expects the positive trend in ebitda over the last year to continue. After this period of investment, the expenses associated with the turnaround should fall off, and the company expects a return to a “competitive margin profile” which the company described to me as at least Avis-level margins. Structurally, there is no reason Hertz can’t achieve similar margins as Avis, which guides to roughly 9% adjusted ebitda margins this year, and targets 13-15% by 2021. The large private competitor in the auto rental industry, Enterprise, is believed to operate with high-teens or better margins, but Enterprise has a structural advantage with its large off-airport insurance replacement business. Avis and Enterprise show that decent margins can be earned today in this industry.

For my modeling, I assume low single-digit revenue growth over the next several years, and a return to Avis’s current margins of ~9%, but not Avis’s target long-term margins. This results in roughly $9.7 bn in revenues in 2021, $870 mm in adjusted ebitda, and around $4.40 in earnings per share. At $19.50 a share today, Hertz trades at 4.4x this estimate for 2021 eps. If Hertz traded at a market multiple then -- and I’m not arguing it should -- it would trade at roughly 12x 2021 earnings, or 170% upside from its current price.

My point is not to actually estimate what they will earn in 2021 or what multiple they will earn because there remains too much uncertainty in their turnaround to have confidence in a precise number. My point is that if the bear thesis that Hertz is dying is incorrect and the company regains some normal level of profitability, one can put a huge haircut on earnings / multiple / timeline and still end up with significant upside. I put a nearly 50% haircut on the above bullish case and end up with a placeholder estimate of fair value that shows around 90% upside from present levels. I believe this kind of upside balances the execution and competitive risk that remains with the company.

Bottom line: if Hertz doesn't die as the world seems to think it will, you should win.

Bonus stock pitch: I think (and have bet) Avis is also undervalued due to the same macro reasons Hertz is. They are simply at a different spot on the risk-reward continuum.

Catalyst

Continued industry revenue growth.
Continued improvement in key metrics of price, volume, utilization.
Continued growth in ride-sharing rentals.
Execution on technology platform rebuild in 2019.
Continued margin improvement.
Expanding partnerships with autonomous vehicle technology companies.
Short covering.