Uber Pretty Much Had an IPO

uber

Post-2012 IPOs are a scam. There is nothing enjoyable about being a public company. Between audited SEC filings, exchange regulations, and analysts in the peanut gallery, public offerings are all downside.

Emerging growth companies are dragged kicking and screaming into the public markets when their early investors finally want to cash out. And an institutional investor only wants to cash out if the emerging company isn’t showing much growth anymore.

Uber’s still growing, which means it’s cool to stay private. But it needs more investor classes to draw from, so it went straight to the retail investors. Okay, the retail investors are Goldman Sachs clients, which makes them the top 0.1% of retail investors, but still: this is the same investor class that would be subscribing to an Uber IPO.

Uber has made a general solicitation, it’s acquired retail investors — Uber has done everything but list on NASDAQ. According to JOBS Act requirements, Uber can have up to 2000 accredited shareholders of record. It could even have up to 500 non-accredited shareholders. But why would Uber want your pocket change when it can have the pocket change of Goldman Sachs clients?

I won’t speculate on when and whether Uber will file for IPO. Maybe someday growth will have steadied and better investment opportunities will have emerged for its existing shareholders. Maybe it will have been so many years that Benchmark and First Round step in and say, Listen, our LPs are long dead and their heirs really need to cash out to pay their estate tax.

And then there will be a road show where institutional clients get first dibs on the Uber IPO. And there will be much fanfare, and a huge pop, and all the investors to date will walk away with fistfuls of money.

And after that, after everyone else has cashed out, that’s when Uber sells its shares to you.

130529121040-nasdaq-620xa

13 thoughts on “Uber Pretty Much Had an IPO

  1. You haven’t written about Uber in a while. Any thoughts on whether it will manage to IPO before its valuation crashes? (It’s inevitable, since a company that owns capital equipment like robo taxis looks very different financially than a software company, and competition wise, all the car companies are coming out with their own robo taxi services by 2020…)

    1. Well, talking to R&D at Ford and G&M, they both plan to have their own taxi services by 2020 and level 4 autonomous cars available to consumers by 2025 (so I imagine the designs will be done by 2020). Although Toyota is working with Uber. Maybe Uber will work with these other guys as the dispatch service, but it might be easy for them to disintermediate Uber. And I know Uber put in that order for 100,000 Mercedes cars back in March. If they plan to own their own vehicles, they should be valued like Avis rather than Google.

      1. So Uber will be buying/leasing autonomous cars from auto manufacturers? If car ownership is completely replaced with autonomous ridesharing, then Uber might be valued more like General Motors than Avis. $60 billion is a little rich, although Uber’s access to China and India might be worth more than I give it credit for.

        1. My understanding is that GM and Ford may want to just operate the services themselves. Who needs Uber when you can make your own app, and you can control the branding, quality control, and liability? Uber’s value is the chicken-and-egg problem of having both a network of drivers and a network of users. Remove one side of that equation and things get a lot easier.

        2. I’ve heard Didi is making bank now that they’re a monopoly (and raised their prices.) So Uber could milk that just like Yahoo and Alibaba.

          1. Okay, if GM and Ford could just operate a network themselves by deploying a fleet of autonomous cars, then Didi’s position is in danger too, right? I don’t know how easy it is to deploy a network of autonomous cars. Maybe they’ll all end up merging, like the airlines.

        1. Good morning,

          Sometimes Daily Updates mushroom; that kind of happened today, about Uber naturally, so this may be a bit of a first draft on a topic that deserves more focus in the new year.

          For now, though, it’s nearly Christmas, and there are only three days left to buy Stratechery gift subscriptions here; details of how gifts will work are here.

          On to the update:

          Uber Losses (But China Gains?)

          New Uber financial numbers from The Information:

          Uber lost more than $800 million in the third quarter of this year, before interest, taxes and stock-based compensation, according to people familiar with its finances. But Uber’s departure from China in the middle of the quarter helped slow the growth rate of its losses to less than 25% year over year, down from 34% year over year in the second quarter.

          Surprisingly, the China exit didn’t dampen revenue growth. Net revenue grew to around $1.7 billion, up more than 240% from a year ago versus a 190% year-over-year net revenue growth rate in the second quarter. And the company booked a $2.2 billion net profit for the quarter, thanks to a big one-time gain on the sale of UberChina to Didi Chuxing in August.

          Gross revenue was affected by the China exit, rising 8% to $5.4 billion in the third quarter from about $5 billion in the second quarter. A year ago, gross revenue was more than doubling on a quarterly basis. But the net revenue gains show that Uber is keeping a higher percentage of gross revenue than it did about a year ago (31% in the third quarter of this year versus 18% in the second quarter of last year). Uber passes back much of its gross revenue to drivers in its network. The company is privately held so it doesn’t release financials to the public, but it’s made the figures available to some investors.

          The first takeaway from these numbers is confirmation of what we already knew: Uber was not making any revenue in China; whatever money came in was passed directly back to drivers as subsidies. This explains the big jump in net revenue (ex-driver payments) relative to gross revenue (total rider payments).

          Secondly, with the caveat that it is not clear how to square “Uber lost $800 million” with “the company booked a $2.2 million net profit for the quarter”, it sure seems like Uber not only received 17.7% of Didi Chuxing but was also paid around $3 billion for the privilege. If that’s true Uber China achieved an even more amazing outcome than it first seemed: I’ve already said that a 17.7% stake of Didi Chuxing is a great return for the ~$2 billion (or more) that Uber spent in China; if the company got its cash back too that is really remarkable. Again, this is presuming that the $800 million loss is referring to operating losses, and that the “net profit” includes financial transactions — if I’m wrong I suspect I’ll hear about it and will let you know in a future update.

          Beyond that, what do these results tell us about Uber’s viability as a company and prospects going forward?

          Well, nothing, because we don’t know the unit economics.

          Uber and Unit Economics

          A series of posts from the site Naked Capitalism have made one of the most in-depth cases to date that Uber is a bad business. They are quite long but in my opinion worth reading:

          • Part 1: Understanding Uber’s Bleak Operating Economics
          • Part 2: Understanding Uber’s Uncompetitive Costs
          • Part 3: Understanding False Claims About Uber’s Innovation and Competitive Advantages
          • Part 4: Understanding That Unregulated Monopoly Was Always Uber’s Central Objective
          • Part 5: Addressing Reader Comments and Questions

          At the beginning of Part 4, the author Hubert Horan summarizes his case:

          The first article presented the evidence that Uber is a fundamentally unprofitable enterprise, with negative 140% profit margins and incurring larger operating losses than any previous startup. Uber did not achieve any meaningful margin improvement between 2013 and 2015 while the limited margin improvements achieved in 2016 can be entirely explained by Uber imposed cutbacks to driver compensation. Uber’s ability to capture customers and drivers from incumbent operators is entirely due to predatory competition funded my massive investor subsidies — Uber passengers were only paying 41% of the costs of their trips, while competitors needed to charge passengers 100% of actual costs.

          The second article provided a breakdown of the taxi industry’s cost structure, and demonstrated that Uber was the industry’s high cost producer, with a significant cost disadvantage in every cost category except fuel and fees where no operator could achieve any advantage. It also explained that Uber could not “grow into profitability” because there were no significant scale economies related to any of these cost categories.

          The third article debunked a range of claims about potential sources of Uber competitive advantage that might explain its ability to drive incumbents out of business; none were based on actual evidence of industry economics, and none of the claimed sources had ever had ever produced major competitive impacts in any other industry.

          The conclusion in part 4 was that Uber’s goal was to wipe out all taxi companies and become a monopoly provider that could exact rents significant enough to generate a meaningful return to investors, and that given this, in Horan’s opinion, was the only plausible positive outcome, Uber was on net destructive to overall economic welfare and constituted a planned “transfer of wealth from consumers and suppliers to Silicon Valley billionaires.”

          I first encountered this series when only Part 1 was posted, which I immediately dismissed due to two significant errors, one factual and the other analytical:

          • First, Part 1 insisted that the financial figures excluded Uber China, which was incontrovertibly false. That’s a pretty big problem considering the scale of Uber’s losses in China! Not only did that error make the all of the numbers (very) wrong, it also raised serious questions about the analysis as a whole, particularly given the fact it remains uncorrected.
          • Secondly, Part 1 analyzed Uber’s profit and losses at the company level. While that is certainly an important data point — as should be obvious, in the long run all companies need to be profitable — it’s not particularly helpful when it comes to a company investing heavily in growth, particularly if they are investing in pursuit of an opportunity governed by not just network effects but also scale effects. What is required is an evaluation of unit costs: how much does Uber earn or lose on an individual ride? And, per the previous point, how did that number differ by city, level of market maturation, etc?

          To understand why the second point is critical consider Uber’s financials for a new city:

          • Uber headquarters’ costs that apply broadly should have no impact when it comes to evaluating any individual market; R&D, corporate G&A (General & Administrative costs, which includes Uber’s legal fees), and some marketing spend fit here. Yes, those things need to be paid for, but in theory as the company grows these costs can be spread ever more thinly over an ever increasing number of rides. An individual city also has its own fixed costs: G&A is the most obvious one, but just like corporate costs this should not be included when determining unit economics, as its impact on any individual ride will decrease with scale.
          • Marketing costs are much more challenging to calculate. Most expenditures fit in the same category as G&A costs: they apply to all rides, and their impact on any individual ride will decrease with scale. However, rider subsidies apply to individual rides and should be a part of the unit cost calculation; what is tricky is determining how much. After all, you can argue that subsidies increase the likelihood a rider will use the service in the future (i.e. there is a lifetime value payoff) and should thus be excluded from unit cost calculations; this is in theory valid but in practice is all too often a very good way to run out of money.
          • The most obvious and easy-to-calculate costs are the ones directly implicated by a single ride, which is primarily credit card fees.

          Figuring out the revenue side is similarly tricky:

          • The gross revenue is the amount collected from the rider.
          • The net revenue is gross revenue minus the amount paid to the driver. This isn’t completely straightforward though, because Uber often subsidizes drivers in a variety of ways, including sign-up bonuses, discounts for accepting a certain number of rides in a specific time range, guarantees earnings, etc. As I understand it the reported net revenue for Uber excludes most of these subsidies (thus, for example, the aforementioned disparity in growth between Uber’s gross and net revenue thanks to the massive subsidies given to Chinese drivers).
          • For the purpose of determining unit economics, however, at least some of the subsidies should be included, particularly those calculated per-trip. You can make the same lifetime value argument I noted above to only include a percentage, but the discount should be extremely slight in my opinion; by the same token some portion of lump sum bonuses should be applied as well, particularly those based on a finite set of pickups (i.e. everything but sign-up bonuses)

          The ultimate equation looks something like this:

          Gross Revenue – Driver Payment – Driver Subsidy – Credit Card Fees – Rider Subsidy

          If this number is negative — as it assuredly was for Uber in China — your business is in big trouble; if it’s positive, then it’s not only acceptable to be generating overall losses, it can in fact desirable if there are the aforementioned network and scale effects to be had (subject to other factors I didn’t get into here, the most important of which are customer acquisition costs and churn).

          The issue with analyzing Uber is that we — including the author of these blog posts — simply don’t know what the company’s unit economics are; that the author attempted to make pronouncements about the company’s financial viability anyways was, in my estimation, an error just as significant as falsely claiming that the numbers did not include Uber China.

          Reconsidering Uber

          Part 2 is far better, and in many respects redeems the series: here Horan does get into the unit costs of pre-existing taxi companies and makes the compelling case that Uber has the highest costs in the industry because car depreciation and maintenance will always be higher for individuals than it will be for fleets. The implication — which I suspect is valid more often than Uber would admit — is that whatever advantage Uber has stems from its drivers not properly calculating their costs (and by the time they do some are locked into new car payments “helpfully” arranged by Uber). Horan goes further, accusing Uber of not just taking advantage of drivers but of leveraging venture capital to engage in predatory pricing to drive companies with better cost structures out of business with the goal of charging monopoly rents that would otherwise be captured by incumbents.

          I don’t agree, at least not completely: I think it is fair to say that Uber has leveraged both bad driver math and consumer subsidies to achieve critical mass quickly, thus giving them leverage to rework decades-old regulations that artificially limited taxi service to the benefit of incumbents. That was a necessary condition to create the level of liquidity that would drive more usage, increasing driver revenue (and costs, to be clear), further increasing liquidity in a positive feedback cycle. Indeed, there is significant evidence that in Uber’s oldest cities this is exactly what has happened — there is far more demand for Uber than there ever was for traditional taxis. It is bizarre — and perhaps telling — that Horan flatly denies this market expansion in Part 3.

          I’m not going to argue that Uber wouldn’t like to be a monopoly; of course they would, and clearly the current plan is to spend Lyft into the ground (in a price war that was initiated by Lyft). But I don’t think it’s correct to say that investors put money into Uber with the express goal of funding predatory pricing.

          • First, I still believe there are supply-side network effects in car-sharing; it follows that venture capital is appropriately used to fund all of the costs excluded from the unit economics calculation above.
          • Second, Uber’s full potential is to replace personal car ownership, at least in some areas; that means the company needs to spend not (just) to kill taxi companies but to overcome the sunk costs of the cars users already have (that most potential users don’t have cars is one of Didi’s big advantages).
          • Third, self-driving cars obviously change the economics of car-sharing completely, and Uber is by far the best-placed to take advantage of those economics. In this case venture capital is providing the capital to not just develop the technology but also to prepare the market to apply the technology.

          I’m going to stop there; this update is already far longer than I intended, and should probably, along with a summary of Uber’s San Francisco self-driving adventures,
          be refined into a Weekly Article (I suspect it will be soon). Ultimately, while I spent most of my time pointing out where I disagreed with those Naked Capitalism posts, the reason I addressed them here is that it represents an argument there about Uber’s viability that is worth taking seriously — and it’s especially important for me to take it seriously, given I’ve generally been an Uber bull. After all, back when I first started writing about Uber in-depth most observers couldn’t grok how the company could be valued at $18 billion; now that the company is worth nearly 4x it is more than fair that the burden of proof start running in the opposite direction.


          The Daily Update is intended for a single recipient, but occasional forwarding is totally fine! If you would like to order multiple subscriptions for your team with a group discount (minimum 5), please contact me directly.

          Thanks for being a supporter, and have a great day!

  2. Thanks for sharing that. Maybe next quarter’s results will give more clarity on what the subsidies are without China. I don’t think the author makes a convincing argument that Uber is best placed for rolling out the self driving cars though versus their competitors. I think fleet operators like zipcar and car companies like Ford have some advantages.

Leave a Reply