Thesis: Ride-Share Financing Programs
I believe that Uber's Financing program is not being strategically used to maximize value for the company.
How does it currently work:
As a middleman between drivers who actively use their cars and lenders who are seeking to deploy capital, Uber is in a prime position to source great lending candidates by hedge downside risk for lenders. Uber's Financing Program works with third-party lenders who are willing to provide subprime loans to their drivers. Uber reduces risk for lenders by taking loan payments from a drivers’ Uber earnings before paying the driver.
Explanation:
Multi-sided platforms create defensible business models by their ability to
(1) create inimitable features
(2) create high switching costs where the marginal value-add of using your pre-existing network is greater than the marginal value-loss of engaging in multihoming behavior (when a customer uses services of two competing service providers in order to capture network benefits)
I believe Uber can create a more holistically durable business by improving their lending services.
Uber has so much data. You can see their disclosure of what they collect by looking at Section III of their Privacy Policy (Data collections and uses): location data, transaction information, usage data, device data, communications data, data from other sources like business partners and referral programs and financial service providers, and more.
As a general principle, there are a huge markets in using alternative data that only you as the service provider have which enables you to create stronger underwriting models to financially empower your customers to increase your bottom line.
Uber drivers can be viewed as their own independent 1099 SMEs (thought employment regulations may be changing this). The car takes up most of the balance sheet as a depreciating asset which enables consistent cash flow based on trackable use. Many of these drivers have already financed their cars, which enables Uber to take a stab at offering refinancing services to consistent, reliable, and loyal drivers that they suspect have high lifetime value. For those that need to finance a new car, Uber is of course in the position to use their historical data to give them offers too.
From my understanding, Uber's current model takes loans payments from an Uber driver's salary. While this may be a good way to hedge risk for a third-party lender, I'm not convinced that it incentivizes drivers to actually drive more or for good drivers to refinance with Uber, which could be a relatively lower-risk and moderate-reward strategy for Uber on just the lending play.
Well, Uber isn't a financial services company... right? They should focus on their core competency: connecting drivers to rider!
I say: wrong. Every tech company is a FinTech company. At scale, every technology company can see massive benefits by enabling financial services.
When Uber becomes a stakeholder in their drivers, they create true partnership.
Uber's buy-in to their drivers enables them to have a higher value than competitors by offering gamified services that competitors simply can't do that same customer because they don't have a huge loan on the line.
"Complete r rides, h hours, t transaction value of rides today/this week/this month and let Uber sponsor x% of your daily salary or y dollars to paying off your loan."
As a driver, I would work pretty hard to hit those milestones. Especially if the opportunity cost of using Lyft means that I won't be getting free payment towards my vehicle. If the driver does more rides with the platform, they get algorithmically-determined or proportionally better payment that is essentially subsidized by Uber. Uber already makes 25-42.75% comission with an average of 39.01% commission on each ride (Ridester). If they offered 5% of daily salary to pay off the loan, they could easily make it up by the sheer number of more rides they may get back from drivers who traditionally use competitors. And because car loans are taken on an average of 70 months (Credit Karma), Uber is actually investing in a strategy for long-term incentive alignment.
So what happens:
-->Uber makes more money on (1) increasing the number of rides on their platform and (2) making margins on their loans.
-->Drivers can accelerate their financial freedom by paying off their car quicker and/or refinancing to better terms. The incentives to get financing should be directly correlated with great customer service and ratings too...
-->Uber's competitors lose market share from customers that no longer engage in that darned multihoming behavior.
Wait but it's not done yet. Let's take a play from Pagaya.
At the end of the day, Uber should in fact focus on their core competencies.
Americans borrow an average of $41,665 for new vehicles and $28,506 for used vehicles, making an auto loan is one of the biggest financial decisions any consumer can make (LendingTree). For Uber to hold the balance sheet risk would be dangerous. In fact if this was done at scale, the company may trade more like a financial institution than a technology company. We would hate to see them move from the 21.77x EBITDA multiple of a software to a 17.17x multiple of corporate financial services (Equidam). And an honest gut check, ~17x seems absurdly high for a financial services company as well. We would also hate to go under if rates suddenly took a hike and suddenly the interest payments on car loans (which represented an extraordinary amount of their balance sheet) become untenable.
Cough cough, SVB.
Uber can securitize their car loans into an auto loan and auto lease asset-backed securities (ABSs) and then divvy up the risk to sell to institutions like hedge funds and banks. Obviously they'd take a percent of the origination fee as revenue, but also if Uber is confident in their ability to rate the credit, which they should be if they offer loans in the first place, then they their FP&A team can use some financial engineering to provide the company incredible levels of leverage on what they deem as the best loans.
Pagaya is a company that uses their AI network to evaluate applications for businesses to lend to their partners. On the backend, they have a pretty freaky marketplace for people to buy this risk.
These ABSs are pretty special because in this model Uber is co-paying and thereby backing some percent of the loan based on metrics that they themselves adjust for driver performance. It's just one step away from the parallel to airlines securitizing the liabilities of their reward miles programs and then artificially adjusting their balance sheet by changing the conversion price. As a lender, if you believe in Uber having strong network effects, you're upgrading the credit rating of the loan. Also, these ABSs are not just sitting around idle. They're basically out there being productive all day and operating as cash flow machines. In fact, because Uber has all the granular details about general use time (such as how often, in which terrain, etc), they have pretty precise real-time feedback on what depreciation schedules realistically look like.
Further thoughts:
Uber can choose to enable this service via third party lenders and then buy their risk back if they so desire or by choosing to use their own balance sheet first and then sell the risk to institutions. A little experimenting and forecasting would help them clarify.
Creating partnerships with select manufacturers to create preferable deals with financing specific cars
Using all that to provide better insurance policies, which they may already do
While this lending services play isn’t inimitable, it creates such a high barrier to entry (aka competitors also need to be able to lend ~40k to capture a marginal value add). Execution and arranging the right partnerships at an institutional level is also hard and makes it not easy to copy.
--
Notes to self:
I wanted it to feel like Byrne from The Diff with how I write, so if it reads like that to others it'll be a compliment. I don't habitually blog about my theses' on companies, usually it's just a quick text to Arham and forget about it, but I think it's valuable to write out the models to develop the skillset. This was pretty fun but also tiring -- talking through it with others makes it more engaging.