Rental Cars & Car-Sharing in an Autonomous Future
This memo was written for the Stanford Graduate School of Business course GSBGEN 569: The Open Road taught by Keith Hennessey.
Introduction
Existing Market Landscape
Rental car and car-sharing companies differ significantly on product, business model, and price point. These companies rent out cars (from either owned fleets or individual owners) on an hourly or daily basis for the user to drive. At a high level, there are “traditional” daily rental car services and newer car-sharing services which have a diverse set of business models.
There has been significant consolidation in the traditional rental car market. Today, three companies represent 95% of the $30B market in the US: Enterprise (Enterprise, National, Alamo), Hertz (Hertz, Dollar, Thrifty) & Avis (Avis, Budget, Payless). Traditional car rental companies have seen increased competition from ride-sharing and car-sharing start-ups; the number of rental cars in service in the US has stagnated since 2015 (though revenues have grown).
While Zipcar (founded in 2000) remains the largest car-sharing service in the US as a subsidiary of Avis, car-sharing has recently exploded. A mix of start-ups, OEM subsidiaries, and local governments offer services. Business models differ widely. Some rent cars hourly, some daily, some both. Some services are station-based, while some are “free-floating,” similar to dockless scooters. A few companies (Turo, Getaround) are peer-to-peer platforms for individual car owners to rent out cars.
Car Rental & Car-sharing Economics
The unit economics of a traditional car rental company are relatively simple. First, the company purchases cars from OEMs at a bulk discount. Most often, these purchases are funded on debt with the car itself as collateral. Some of these cars are “program” cars, which the OEM agrees to buy back at a set price after a set period. Then, the company rents the car out to its customers while managing the operations (utilization varies but averages ~70–80%). After 1–2 years of use, the company pulls the car from its fleet and sells it at auction or back to the OEM.
The economics of car-sharing companies differ widely. Platforms like Turo earn a percentage of each booking, but do not own or manage vehicles. Fleet-based car-sharing companies have similar economics to car rental companies, though they spend less on fixed operating assets like storefronts and tend to own and operate their cars for a longer period of time. Many car-sharing services operated by OEMs are not profitable and are seen as longer-term investments in understanding “Mobility as a Service” via operating data.
The Future: Autonomous Vehicles & Mobility Use-Cases
Despite the wide range of ways that consumers use cars today, the conversation about autonomous vehicles largely centers on a single business model: robotaxis. This is the business model that most major AV companies are focused on (Waymo, Cruise, Zoox, Uber, Lyft, etc.). However, an Uber-like robotaxi will never be ideal for all consumer vehicle use-cases, like running errands with multiple stops and shopping bags, taking a weekend trip to the mountains, or road-tripping across the country. This gives rental car and car-sharing companies (along with traditional car manufacturers and dealers) hope that autonomous vehicles will not completely disrupt their businesses in the short term.
However, as consumers learn to trust robotaxis, demand will grow for autonomy to fulfill all vehicle use cases. There are at least three prevailing thoughts as to how this need will be served:
1. Robotaxis expand to service these use cases — As ride-sharing companies have diversified into micro-mobility, they could offer longer-term dedicated autonomous rentals. Similar examples exist today: ReachNow has both car-sharing and ride-hailing integrated into one app and micro-mobility company Lime has begun rolling out LimePods, a free-flowing car-sharing service. This would help fulfill the ideal of “mobility as a service,” disrupting not only just rental car companies but also the modern concept of car ownership.
2. Rental car and car-sharing companies port existing model into AVs — Simple, but these companies do not have the technical capabilities today to build AVs, so this must develop via partnership.
3. Individual AV owners or small fleet owners rent excess capacity via a platform like Turo — For this scenario, consumers must grow trust in peer-to-peer car-sharing and individual ownership of AVs must proliferate.
In the very long-term, the first scenario is most likely, but there are many opportunities in AVs for rental car and car-sharing companies in the near and mid-term.
Existing Assets & Capabilities: Why Car Rental and Car-sharing Companies Could Win
Purchasing: Relationships with Car Suppliers & Financing
Car rental companies are the largest individual purchasers from OEMs², and thus have deep, long-standing customer relationships. They get bulk discounts and have existing financing and distribution relationships in place. However, many AV companies are owned by or have strategic partnerships with OEMs, and most have deep pockets, likely negating these advantages.
Platforms like Turo and Getaround have existing relationships with car owners. If individual AV ownership proliferates, this would be a strategic asset in launching P2P AV sharing. However, Uber and Lyft also have relationships with car owners (drivers), so they will still be advantaged in a robotaxi model.
Ownership: Accounting Considerations
Some commentators have argued that simple accounting positions rental car companies for success. Hertz and Avis hold $14B and $12B of depreciating car assets on their balance sheet, charging $2.8B and $1.9B in annual depreciation expense respectively. The argument goes that while AV companies could afford this, they would prefer to use their money elsewhere. However, a simple holding company could own the cars at a lower cost to AV companies. Additionally, companies like Alphabet clearly have an appetite for capital investments.³ In the near term, AV companies will purchase cars directly with low-cost financing, using the vehicles as collateral, to reduce the cashflow impacts of the purchases.
Fleet Management: Experience vs. Algorithms
While margins are slim, car rental (and to a lesser extent car-sharing) companies have decades of experience optimizing fleet management and pricing. However, these functions are similar to the algorithms developed by ride-sharing companies to load balance their services, and ride-sharing companies have more engineering talent to build these complicated fleet management capabilities.
Parking: Real Estate Assets
Rental car companies have both airport and retail locations, which serve as operational hubs. Airports have a limited number of “slots” and generally charge a fee (2–20%) on every rental. These slots could be valuable to robotaxi fleets to improve pick-up times at airports, but so far ride-sharing companies have tried to solve airport issues using different methods. Retail locations are valuable as maintenance hubs for fleets — they tend to be distributed throughout cities and are already fitted for car maintenance. Enterprise is particularly well-suited here, as they have significantly more city-based retail locations than Hertz or Avis. Lastly, car-sharing companies have parking assets, either dedicated spots in private lots, or public parking agreements with cities. These would be assets for AV companies to minimize idle driving time.
AV companies today are buying up real estate in key cities for the storage and operation of their fleets, so this is a key consideration and an area where strategic partnerships with rental car companies could create value for both parties. However, a concern is space — for example, how much space would an Enterprise location have available to dedicate to a partnership with Cruise? Other companies like parking lot operators or REITs could potentially offer similar partnerships, but they likely do not have the same maintenance infrastructure or expertise. Additionally, while a distributed maintenance footprint would reduce AV idle road time, it would also add to operational complexity and may not be worth it, especially in the short term.
Maintenance: The Dirty Work of AV Fleets
The most commonly cited asset of rental car companies is their experience maintaining fleets of vehicles. Rental car companies have partnered with AV companies to perform maintenance, but in small trials. Some industry experts see these as PR stunts, but others see them as a step in the right direction towards adapting business models to autonomous technology. The key question is value capture — how much would AV operators be willing to pay Enterprise to clean and maintain their cars, and how does that business model compare to the economics outlined above? Other companies like car repair shops also have expertise in this area and could compete for contracts.
There also are alternative models for car maintenance today that may be a model for AV fleet maintenance. Getaround has partnerships with local vendors that will clean, maintain and refuel cars on their platform (available to larger fleet owners in select cities). You could also imagine a program similar to the “juicer” model for scooters, where individuals earn money for cleaning and/or recharging AVs at their home.
Predictions
Near-term (0–5 years)
Rental car companies will continue to see pressure on their business model from ride-sharing. P2P car-sharing platforms will buck this trend due to increased consumer adoption on both the supply and demand side. Start-ups in car-sharing, particularly peer-to-peer car-sharing and free-flowing car-sharing, will develop strategic partnerships with (or be acquired by) ride-sharing providers, as transportation network companies (TNCs, e.g., Uber & Lyft) expand their multi-modal transportation offerings. Getaround has previously partnered with Uber for Uber Rent, but the service is no longer available.
As AV deployment grows, rental car companies will continue maintenance partnerships like what we’ve seen already, but the scale will be small (pilots not major contracts).
Mid-term (5–25 years)
Both rental car and car-sharing companies will see an increase in core market size, as cheaper L4 robotaxis serve most day-to-day needs, allowing more people to forgo car ownership. Thus, for trips that do not fit into the robotaxi model (long-distance, multi-stop), rental car and car-sharing companies will steal “market share” from car ownership. If TNCs expand to offer car-sharing, they may “steal” this growth, leaving traditional car rental companies with grim prospects. Expect rental car and car-sharing companies to bring AVs into their portfolio, especially if OEMs commercialize individually-owned AVs.
Robotaxi fleets will experiment with different operational maintenance models to determine which is most profitable. Three models will exist: 1) centralized “depot” maintenance; 2) distributed “retail” maintenance; and 3) decentralized “contractor” maintenance. Rental car companies, in particular Enterprise, will win partnerships looking to develop a “retail” system, but are not particularly well-positioned for the other models.
Long-term (25 years +)
Predicting this far out is an exercise in science fiction but at some point, true L5 autonomy will proliferate. At this point, rental car companies (if they still exist) will exclusively rent out AVs. The most likely scenario is that dominant robotaxi operators will expand their mobility products to include AV rental/sharing to serve the widest possible set of transportation use-cases.
Thank you to all individuals who contributed to the writing of this memo
[1] Hertz is an uncharacteristically poor performer, with operating losses in 3 out of 4 years. Hertz stock has declined 82% since February 2015. In contrast, Avis has consistently had operating margins of 5–10% over the past decade, but they have a more diversified business, making unit economics harder to ascertain
[2] The three major car rental companies purchase over 1 million cars annually. As a risk-saving measure, each car rental company attempts to limit the proportion of purchases from any one OEM, so they have many large relationships; in contrast, Waymo made big news with an order of only 62k Chrysler Pacificas
[3] Alphabet’s PP&E totaled $42B in 2017, and their depreciation expense was $6.1B, mostly from depreciating infrastructure for cloud services, admittedly a much higher margin business than AVs
