Why Uber’s capital costs will creep ever higher

So Saudi Arabia has invested $3.5bn in Uber, the ride-hailing app, making it the largest single investment ever made in a private company.

Talk of war chests and global expansion abounds. But perhaps what the above really implies is that Uber’s famous capital-light model is about to get much more capital intensive — especially as it moves towards rolling out the much hyped self-driving fleet. If that’s the case, investors need to pay attention. Along with capital intensity come limitations to the exponential growth rates investors have come to expect.

So far, of course, the genius of Uber’s business model has been all about transferring capital costs to partner drivers. Uber terms and conditions do their utmost to ensure you the customer understands the service-providing relationship is between you and the driver, not Uber. Drivers utilise their own cars, pay their own costs including petrol, maintenance, parking fees/fines, insurance, licensing costs. They run their own risks too.

This indeed is how Uber gets away with calling itself an intermediary, not an employer.

Going against that, though, is Uber’s control of the prices contractors get to charge customers and its reluctance to offer exclusivity terms with respect to rivals (the driver network is notoriously open-ended). As a result, contractors haven’t the power to price their services according to demand, or even in ways that cover their cost of operation.

Once costs are properly accounted for, many drivers contend Uber’s pricing strategies and uncapped driver policy — which sees drivers added irrespective of market demand — result in take home pay which amounts to a minimum wage or less.

But there are other factors in play too.

Uber’s driver network, for example, has always been made up of vehicle owners and vehicle renters. The distinction is important because being able to draw on a large pool of pre-funded private capital (a.k.a people who own cars and want to put them to use) has been key to Uber’s price-cutting tactics in the market.

That pool of capital, however, is limited not least because the sort of people who can afford the quality cars Uber demands for many of its services are not the sort to be satisfied with minimum wage earnings for long.

As the number of car owners who look to Uber for a means to a more flexible type of employment or a means to bolster earnings runs dry, so too does Uber’s capacity to undercut the competition and keep growing.

Uber’s growth model instead becomes increasingly dependent on attracting the latter sort: vehicle renters.

If Uber’s driver network is to keep growing on that basis someone somewhere must put up the capital drivers can’t afford to put up themselves — a service that’s unlikely to be offered pro bono.

In that context Uber’s rates simply can’t stay low forever: the aggregate earnings of the network must cover the aggregate operating costs of the network inclusive of insurance, minimum wage and capital interest owed.

So what does a capital-light app that wants to keep its driver network growing without hiking ride prices to cover capital rental and interest costs to do?

Answer: deploy its reputation in drivers’ name so as to fetch them better deals with manufacturers and credit providers than they could fetch alone.

Uber is thus no longer a ride-hailing unicorn. It’s well on its way to becoming an auto-leasing company — complete with all the capital costs and credit exposures that come with it.

For investors who bought into the capital light model this is concerning, not least because Uber has little to no natural edge in the leasing market — especially compared to specialist asset-financing companies or banks. Presumably that’s why Uber’s leasing offer still depends on discounted access to third party lender deals (which in the past has included Santander, Westlake, Exeter), with Uber’s role simply that of payment collector.

*According to Bloomberg, the latest twist sees Uber’s Xchange leasing solution benefit from a $1bn credit facility from Goldman Sachs for the purpose of funding new car leases directly, linking the company’s own balance sheet and credit profile to driver earnings even more explicitly.

But as Matt Levine noted this week, at $155 per week for a Toyota Corolla, Uber’s lease rates don’t really exemplify the collective bargaining deal of the century.

There are only two logical explanations for why Uber’s rates look and feel so predatory. The first is that disrupting the leasing and credit market wasn’t as easy as Uber expected, meaning the company remains a price taker because lenders understand the risks involved. The second is that Uber’s business strategy has become so opportunistic, it’s failed to notice that gouging its own drivers on credit deals is self-defeating in the long-run because all it does is sabotage the potential for growth or force those costs to be passed on to customers directly. FWIW, we think it’s probably the first factor.

But there’s another point to note here too. Uber’s leasing deal also offers free maintenance to drivers, further indicating Uber has woken up to the merits of economic scaling and the power of large-corporate structures in negotiating favourable supplier deals.

Rather than having thousands of drivers negotiating maintenance and service terms on a private basis and being overcharged accordingly, Uber can pool those cars to be serviced under a single contract for the entire fleet.

The same also applies to fuel costs. Uber is, for example, now offering drivers Fuel Cards which entitle them to discounts at partner petrol stations and fuel credit.

But as Uber increasingly stands in the stead of drivers for collective bargaining, scaling and credit purposes, it moves ever closer to sabotaging the capital-light model investors know and love. The Uber workforce also begins to look ever more employee-esque, risking its hallowed contractor defence.

There are other risks too.

As already noted, in both of the cases above, credit and capital costs are deducted by Uber (alongside its standing commission) from drivers’ pay cheques at source, meaning drivers receive only the differential on pay day.

Those drivers who hitherto had over estimated their net take-home pay (due to accounting inexperience) may via the process wise up to how their pay and lifestyles compare to more established minimum wage jobs with benefits.

If it becomes widely known contractor earnings barely beat the minimum wage, Uber’s capacity to recruit new drivers could be affected — undermining its growth potential. If wages consistently fall below the minimum wage, Uber might even risk widespread driver drop-outs, defaults on lease obligations and a souring reputation.

Given all of the above, if the driver network is to keep growing as projected without Uber hiking customer rates, so too must the overall number of hours worked per driver — at least if drivers are to meet their lease payment obligations. But since that would dilute earnings per hour further, this too might just stir drop-outs and network stagnation anyway.

At this point Uber usually claims its uncapped driver policy is entirely justified by the soaring demand for its services. What we think this view misses, however, is the degree to which demand may be hit if and when Uber’s prices are forced to adjust to drivers’ true cost of capital plus minimum wage expectations.

As a result, we predict, Uber might eventually be minded to do the unthinkable: regulate or cap its driver numbers — with lease-contracted drivers being required to work as many hours as their minimum payment obligations demand or shuffled into coordinated shifts. In both cases, the only appealing part of the job – the flexibility – is now removed.

Though even this sort of stabilisation might not be enough to spare Uber from the risk of network stagnation.

What happens if drivers are unable to meet lease payments because they fall sick, are involved in accidents or bad weather conditions prevent them from driving? The current leasing deal forces drivers are required to pay back all outstanding sums owed before any pay is released to them at all. As Uber’s leasing FAQ notes:

If that doesn’t sound like a recipe for defaults and driver drop outs at the first whiff of a set-back, if not a human rights issue (working for for zero income just to pay off debts seems inconsistent with labour laws), we don’t know what does.

Since Uber’s interests are clearly more aligned with offering drivers forbearance than those of lenders (who have even been accused of concocting systems which lock drivers out of cars if they miss payments) … might Uber then realise it’s only by offering drivers minimum securities like sick-pay, medical expenses as well as a guaranteed subsistence wage that it has a hope in sustaining its network, let alone encouraging its ongoing growth? An irony which won’t be missed by anyone.

For those thinking, ah ha! Uber’s driverless fleet will soon solve all those problems — we’d add that an autonomous fleet only turns Uber’s capital-light rentier model into a capital intensive more quickly. It also, arguably, changes its business model entirely.

Uber would no longer be a taxi company but a car-rental company — somewhat problematic for a self-driving service given the extent to which rental companies already depend on fuel and insurance surcharging tactics for their profits (rather than rental fees, which are loss-leading).

For Uber, forcing clients to take out insurance on a per journey basis or agree to terms that require them to refill fuel en route at market rates or risk being charged per mile at penalty rates is simply not an option. Nor, of course, would Uber have a rental company’s advantage of being able to regulate the size of its fleet according to seasonal demand because selling off autonomous cars to private owners in the long run only sabotages the taxi business. (When everyone has their own car on call, why use Uber or any taxi?)

In an autonomous fleet world Uber’s only break-even option, we think, would be to price its services at levels that cover its core operating and capital costs or by withholding supply. Would those prices beat the affordability of conventional public transport options or taxis managed and operated by humans? We’re not at all convinced they would.

As transit consultant Jarret Walker’s excellent Human Transit blog has stressed for a long time, single-occupant cars whether autonomous or not are and always will remain the most inefficient (and thus costly) way of utilising scarce urban space. If taxi rates are cheap enough to encourage substitution from public transport, somewhere in the system there’s a costly inefficiency.

Hence why we expect Uber’s capital costs will keep on rising, and with it the company’s appetite for funding rounds.

*Paragraph added post publishing due to inadvertent loss in editing process.

Related links:
If and when Uber drivers unionise… - FT Alphaville
‘This house proposes that we nationalise Uber’ – The Long and Short
Leigh Day Legal Action For GMB Uber Drivers To Secure Rights On Pay, Holidays, Health And Safety, Discipline And Grievances – GMB, July 2015

Copyright The Financial Times Limited 2019. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.

Read next:

Read next:

Lookout, there’s a dollar crunch!

FT Alpha Tweets