By Hubert Horan, who has 40 years of experience in the management and regulation of transportation companies (primarily airlines). Horan has no financial links with any urban car service industry competitors, investors or regulators, or any firms that work on behalf of industry participants
Uber released its S-1, its IPO filing with the Securities and Exchange Commission, on April 11th. Press reports indicate Uber is seeking to raise $10 billion and achieve a valuation close to $100 billion. To succeed, investors will need to believe that despite losing roughly $14 billion in the last four years, Uber not only warrants a current valuation that would make it the second most valuable startup IPO in US history (after Facebook) and the second most valuable publicly traded transportation company in the world (after Union Pacific), but that its stock will continue to steadily appreciate once it becomes publicly traded.
Uber’s biggest IPO challenges are to convince potential investors that it has already made substantial progress towards reversing its recent massive losses, and that its businesses have the ability to generate strong, sustainable profits.
Uber’s S-1 not only fails to present any credible evidence about future potential profits, but its presentation of historical results is designed to mislead potential investors about recent improvements that did not actually occur.
Nobody believes Uber’s claim that it earned a billion dollar profit in 2018
Uber’s S-1 claims that it achieved a $5 billion profit improvement in 2018, moving from a $4.03 billion loss in 2017 to a $997 profit in 2018. Uber’s efforts to manufacture artificial accounting profits that gullible outsiders might not immediately see through first surfaced with the release of its first quarter 2018 results, when a claimed $3003m gain from the sale of its failed Southeast Asia operations to Grab converted a huge loss into a small profit. A fair number of the reporters actively following Uber saw through the original first quarter ambit  and the vast majority of reporters following the S-1 ignored the reported $5 billion improvement.
Many current reports have still misreported the 2018 profitability of Uber’s ongoing operations in ways that incorrectly showed meaningful year-over-year profit improvement  because Uber was deliberately making it difficult for outsiders to understand those numbers. But most realized that any reported profitability numbers needed to exclude items such as the Grab gain.
All previously released Uber P&L data were based on actual worldwide operations. Uber decided to recast all of its S-1 financial reporting to segregate results from countries (China, Russia, Southeast Asia) where it failed dismally and sold its remaining operations to the dominant local company. This would be reasonable if done transparently and consistently, as it might help potential IPO investors better understand the past performance and future profit potential of the parts of the business that haven’t failed.
If Uber’s intention were to help these investors, the S-1 would have included clear warnings that the 2018 “improvements” were the result of the decisions to abandon these failed markets. Uber didn’t do that, because they are trying to give outsiders the impression that profitability is on a strong upward trajectory.
Uber overstated its 2018 profit improvement of its ongoing operations by $5 billion; ongoing operations lost the same $3.5 billion that they had in 2017.
Uber’s S-1 P&L says that “Net Income From Continuing Operations” improved by roughly $5 billion, from a $4 billion loss in 2017 to a $987 profit in 2018. This is not true; the actual net income from continuing operations in 2018 was negative $3.5 billion, virtually the same as 2017’s result. Uber mischaracterized $5 billion in gains from discontinued operations as gains from continuing operations in order to mislead IPO investors into thinking that the performance of its ongoing business was rapidly improving.
When Uber abandoned its failed Chinese, Russian and Southeast Asian operations, the dominant local companies gave Uber equity and debt instruments to partially compensate it for providing them an easier path to market dominance. These non-tradeable instruments only exist because of Uber’s decision to discontinue operations, but Uber includes their $5 billion value in “Net Income From Continuing Operations”
The current accounting value of these assets is based entirely on Uber’s judgement as to what paper issued by companies currently losing massive amounts of money might be worth someday.  If one takes Uber’s judgements at face value, one could conclude that Uber’s only profitable activity is getting paid off for discontinuing staggering unprofitable markets.
In addition to misleading investors about the proper distinction between the results of discontinued and ongoing operations, Uber inappropriately combines the hypothetical future value of non-tradable paper on the “Other Income” line with items that legitimately reflect current year business activity such as interest income and foreign exchange impacts. Uber undoubtedly needed to assign and record a value for these assets somewhere on its financial statements. The problem is Uber’s failure to sufficiently highlight the problematic nature of the assigned values and to ensure investors clearly understood that they had nothing to do with the performance of its current marketplace activities.
The table below summarizes the differences between the 2018 Net Income for discontinued and ongoing operations shown on Uber’s S-1 P&L statement (roughly zero and positive $987 million) and the actual numbers based on other data reported in the S-1. The $5 billion year-over-year profit improvement misstatement results from understating 2017 ongoing profitability by roughly $500 million (because losses from discontinued Russian/Southeast Asian operations had been included with other ongoing operating results) and overstating 2018 by $4.5 billion. Uber’s desire to mislead investors about the actual split (and its failure to improve the profitability of its ongoing operations in 2018) is further illustrated by the fact the numbers for discontinued operations were spread across six different subtables and footnotes. 
Uber’s S-1 provides no evidence that its international operations are anything other than financially disastrous
One of the major risk factors identified in the S-1 is that “Our business is substantially dependent on operations outside the United States, including those in markets in which we have limited experience” with 74% of total trip volume currently coming from outside North America. Footnote 2 notes that only 54% of Uber’s revenue comes from outside North America, suggesting that foreign market and competitive conditions substantially reduce average revenue per trip.
Uber’s S-1 claims operating losses of over $3.5 billion from two and a half years of operations in China and two and a half years of operations in Russia and Southeast Asia. Unbelievably, the S-1 claims the discontinued operations incurred significant expense but had almost no revenue. Footnote 15 explicitly says Uber’s had over $1.5 billion in operating expenses in China in 2016 but only $1 million (with an m) in operating revenue. 2016-18 operating results from Russia and Southeast Asia shown in Footnote 18 are similarly anomalous. Uber’s accountants apparently saw no need to explain why Uber was conducting a large scale business that had no revenue, but the reported data contributes to the appearance of large subsequent profit improvements.
The combination of these data points should serve as flashing neon warning lights to investors trying to understand Uber’s future growth and profit potential.
But Uber doesn’t present any evidence in the S-1 that would mitigate concerns about the international operations it hasn’t yet abandoned. These three markets might have been Uber’s worst, or they might be the only markets where a larger operator was willing to offer Uber a deal including non-tradable paper allowing face-saving claims of hypothetical future gains if they abandoned the market.
Unsubstantiated press reports suggest certain markets (such as India) are especially problematic, and that meaningful competition precludes the possibility of profitability anywhere. Most foreign markets enforce regulatory and safety requirements that Uber freely flouted or nullified in the US.
A further hint that a lot of Uber markets may have especially weak financials is the disclosure that 24% of all Uber trips occur in just five especially dense and wealthy cities (Los Angeles, New York San Francisco, London, Sao Paulo). There is no way to access whether certain type of markets offer Uber the path to profitability that it has clearly been unable to find broadly. Uber’s failure to provide any objective data about profitability by market suggests that it may not have any data that could address these concerns.
Uber’s S-1 does not provide any information that would help investors understand driver economics, or the relative profit performance of Uber’s different ongoing businesses.
All of the Uber economic and financial analysis presented in this series since 2016 has been based on published taxi industry data and the limited Uber P&L data released to the press. Usually the publication of a 300 page IPO prospectus would give outsiders a trove of new data permitting much more detailed and sophisticated analysis of the company’s economics.
That is not the case here. Uber’s S-1 data does not allow outsiders to evaluate the current profitability or the profit trends of any of Uber’s lines of business (car service, food delivery, scooters, etc.). It does not allow outsiders to evaluate whether observed revenue changes are due to pricing or demand or competitive changes, or to understand whether pricing or other marketing changes increased or decreased operating losses. As already noted, outsiders have no way of knowing whether there are significant differences in revenue per trip or profitability by country or by markets within countries, and no way of evaluating what factors might drive those differences.
Uber’s business model depends on a massive number of allegedly “independent” drivers. Uber’s S-1 provides no information (such as base and incentive earnings, turnover, driving patterns, utilization rates, costs of acquiring new drivers) relevant to whether they can continue to provide the capacity that Uber’s customers are paying for. Prospectus readers have no idea whether how driver take-home pay compares to minimum wage levels or alternative low wage jobs, or how much driver compensation has declined. Prospectus readers can’t tell how Uber changed base and incentive compensation in response to marketplace challenges (such as the terrible publicity Uber received in 2017, or local market share battles) or whether Uber has any potential to reduce the driver share of passenger fares going forward. Prospectus readers have no idea what portion of drivers work very long hours, what portion only work during demand peaks and what portion only work occasionally, and have no idea whether these driver patterns align with demand patterns.
Lyft’s IPO prospectus presented ridesharing unit revenue data but Uber did not. Uber only presented the sum of car service and food delivery trips, so prospectus readers couldn’t figure out what customers were paying for the two services separately, or how prices for the two services have changed. The combined data suggest that both growth rates and pricing was declining in the second half of 2018, but prospectus readers have no way to identify the underlying problems, or whether those problem are likely to get worse.
Uber’s S-1 provides absolutely no data on operational efficiency. Its claims about synergies and scale/network effects are completely unsubstantiated. It says, “Our strategy is to create the largest network in each market so that we can have the greatest liquidity network effect, which we believe leads to a margin advantage” and that synergies “across our platform offerings …effectively lower our costs and allow us to invest in a scalable way that becomes increasingly efficient as we grow with each new product or offering.” But Uber fails to provide any supporting evidence about productivity gains or actual margin improvements.
The S-1 highlights the importance of “technologies” that allow better matching of supply and demand and more optimal pricing. But it makes no effort to demonstrate whether these technologies are any better than tools used by other companies and presents no evidence showing that the alleged better matching of supply and demand actually reduced unit costs or how its sophisticated pricing systems actually increased unit revenues. Although Uber’s entire legal defense of its “independent contracting” model is that it is a software company and not a transportation provider, the prospectus does not even pretend that it is a software company.
Uber claims it benefits from significant scale economies saying “we believe that the operator with the larger network will have a higher margin than the operator with the smaller network” but presents no evidence showing that size is the primary driver of profitability. In reality, other factors such as the ability to maximize the revenue utilization of assets and labor against complex demand patterns are far more important drivers of transport profitability. Since Uber is already orders of magnitude larger than any previous urban car service operator, if size was the critical driver one would think these powerful margin effects would have shown up by now.
Uber’s claims of powerful network effects are nothing more than the assertion that some drivers can carry both passengers and food deliveries and that “Uber Eats is used by many of the same consumers who use our Ridesharing products.” Uber has no evidence that its platform increases the loyalty of either drivers or customers. The most likely explanation for the rapid recent growth of Uber Eats is not that customers are locked in to an app they like, but that customers got the same massive subsidies that drove the early growth of Uber’s car service. Airline passengers also buy hotel rooms and rental cars and restaurant meals But airlines understand that the synergies of combining these products within a single smartphone app owned by a single corporate entity are trivial.
Uber has internal data that could address all of these efficiency, pricing and margin questions in great detail. One can reasonably presume that they chose not to include any of it in the IPO prospectus because it would raise serious doubts about the company’s future growth and profit potential.
Since the prospectus fails to make a legitimate, compelling case for strong near-term profit improvements, its other shortcomings can be ignored.
Numerous other sections of the prospectus fall somewhere on the scale between “should raise red flags” (Uber has over $1 billion in San Francisco real estate commitments) and “argument too ludicrous to take seriously” (Uber’s true growth potential is defined by the worldwide market for urban travel). Again the question is not whether Uber might someday reach breakeven, but whether investors see enough robust long-run revenue, profit and equity growth to justify making Uber the second biggest startup IPO in US history and the second most valuable publicly traded transportation company in the world.
Uber correctly notes that all of its expansion opportunities involve industries with much more challenging economics and much more efficient incumbents than ridesharing. The IPO is clearly dependent on investors who think that years of powerful revenue growth can drive equity values even without profits, and the prospectus strongly emphasizes that Uber has yet to achieve a 1% share of its near-term market potential. But Uber fails to tell investors how much capital will be required to fuel this long-term growth (it already has $7.5 billion in debt), or how long it will take to earn returns on that capital.
Uber notes that it has already invested $475 million in autonomous car development, but doesn’t say how much more investment will be required, how its approach offers greater promise than what its competitors are doing, how long it will take before it believes autonomous cars can begin to produce meaningful revenue, and makes no effort to explain why Uber’s investments are likely to succeed given the huge technological, legal and competitive obstacles. There are various unsubstantiated claims about “synergies” between scooters and taxis, but absolutely no explanation of the current economics of scooters or how they might someday become a profitable business. The prospectus briefly mentions Uber’s investment in flying cars, but seems to be hoping nobody notices.
Growth beyond Uber’s core taxi business requires being able to profitably produce services substantially cheaper and more convenient that existing options such as private car ownership and public transit. Since Uber is still billions away from being able to produce profitable taxi service, there is little need to spend time evaluating Uber’s long-term growth claims.
 The Uber S-1 can be downloaded at https://www.sec.gov/Archives/edgar/data/1543151/000119312519103850/d647752ds1.htm
 Can Uber Ever Deliver Part Fifteen–Uber’s Q1 Results – Reporters Show They Aren’t Up to Reading Financials (May 24th, 2018)
 Most news reports, including the New York Times, Washington Post, CNN and Fox Business have reported a headline $1.8 billion loss. This is a non-GAAP EBITDA contribution measure, not a GAAP profit measure. A couple others reported Uber’s $3.0 billion loss from operations, which is also not a true profit measure.
 Uber financial results through mid-2017 are documented in Hubert Horan, Will the Growth of Uber Increase Economic Welfare? 44 Transp. L.J., 33-105 (2017). Available for download at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2933177. Subsequent P&L data can be found in other parts of this series including Can Uber Ever Deliver? Part Thirteen: Even After 4Q Cost Cuts, Uber Lost $4.5 Billion in 2017 (18 Feb 2018) and Can Uber Ever Deliver? Part Seventeen: Uber’s 2018 Results Still Show Huge Losses and Slowing Growth as IPO Approaches (16 Feb 2019)
 The S-1 notes that neither the 2016 Didi transaction nor the 2018 Grab transaction have been approved by the local competition authorities. Rulings that these deals unacceptably exploited the elimination of competition would presumably reduce these hypothetical paper values to zero.
[6} There are several discrepancies in the data Uber reports. Footnote 3 says the gain from the Yandex transaction $1234m; footnote 9 which says the value was $954m. The $2328m Grab number in the S-1 is inconsistent with all previous releases of 2018 Uber P&L data which said the Grab value was $3003m. Presumably Uber’s accountants could explain these discrepancies, but they demonstrate that those accountants were going to great lengths to make it difficult for S-1 readers to figure out the actual financial split between discontinued and ongoing operations.