How to Read Uber and Lyft's Purposefully Confusing Earnings Reports

For years, Uber and Lyft have claimed that profitability is just around the corner (or used accounting tricks to claim that they’re making money). Now, the ride-hail companies are back at it again: This week, Uber touted a positive net income in its earnings report while Lyft posted its first profit, but both claims come with a giant asterisk.

For the second quarter of 2021, Uber reported a net income of $1.1 billion thanks to unrealized gains in its stakes in Chinese ride-hail firm Didi (worth $1.4 billion) and self-driving startup Aurora ($471 million), which purchased Uber’s autonomous car division last year. However, on an adjusted EBITDA basis, which means earnings before interest, taxes, depreciation and amortization and excludes investments and one-time sales, it still lost $509 million―which is $150 million less than last quarter and $328 million better than this time last year.

Companies such as Uber and Lyft rely on EBITDA as well as commonly used generally accepted accounting principles, known as GAAP, with Uber noting that it uses the former for operational decision-making and believes it to be a “meaningful” supplement to GAAP. 

Companies provide a reconciliation between both measures for most line items (though Uber excludes forward-looking adjusted EBITDA from this), but it can be difficult to parse the wall of financial figures and adjustment carve-outs. 

"EBITDA is sometimes used by companies with very large fixed assets, large intangible assets (such as goodwill acquired after a major merger) or significant debt financing to give outsiders a crude sense of a company's ability to meet its outstanding financial obligations. Uber has none of these characteristics," writes transportation analyst Hubert Horan in an analysis of Uber's finances. "More importantly, Uber's reported ‘EBITDA’ numbers exclude billions of expenses other than interest, taxes, depreciation and amortization.”

In last year's second-quarter results, for example, Uber Eats doubled its revenue but still only improved its margins by $54 million on an adjusted EBITDA basis. That adjustment’s definition is a 225-word long laundry list of exceptions, suggesting―as the Financial Times concluded in an article at the time―that the “only takeaway you should have from this is that Uber Eats likely lost a lot more money without these apparently non-core adjustments.”

Even with adjusted EBITDA, Uber’s numbers are nothing to write home about. Over the first two quarters of this year, Uber’s “mobility” segment saw revenue nearly double from $853 million to $1.6 billion while its adjusted EBITDA profit for that segment actually decreased by $119 million. Uber’s “delivery” segment saw revenue jump from $1.7 billion to $1.96 billion while its adjusted EBITDA loss narrowed slightly from $200 million to $161 million. 

In 2019, Uber’s S-1 claimed that it improved its margins by $5 billion, from a $4.03 billion net income loss in 2017 to a $997 million net income profit in 2018. However, the lion’s share of this improvement came from two things that showed Uber’s core business was not any more profitable. The first was a one-time accounting entry in the first quarter of that year, where Uber reported a $2.5 billion profit after selling its failed Southeast Asia unit to Grab for $3 billion. The second was an active attempt to shuffle and segregate data across numerous tables and footnotes in a way that minimized the losses incurred by its failed operations in China, Southeast Asia, and Russia, while emphasizing non-tradable debt and equity given in deals with Didi, Grab, and Yandex.

Lyft has not tried to play these sorts of games in the past, largely because it did not try to roll out failed units overseas, although it has made inflated claims in its S-1 (like that the market opportunity for ride-hail is twice as large as healthcare’s) to present itself more positively to investors ahead of an IPO. This quarter, the company reported an adjusted EBITDA profit of $23.8 million, which chief executive Logan Green hailed as a "significant milestone for a business and our industry" in an earnings call.

Lyft is about as transparent with its GAAP numbers as Uber, and its numbers tell a similar story: without complex EBITDA adjustments, Lyft is still not profitable. Its GAAP total costs and expenses come out to $1 billion, but after adjustments similar to EBITDA falls to $771 million. Its GAAP Loss from Operations is around $240 million but after adjustments similar to EBITDA shrinks to $6.2 million. Its net loss using GAAP is $251.9 million but after adjustments falls to $18 million. It’s only in the EBITDA calculation that Lyft says it posted a profit of $23.8 million.

But none of this has stopped Uber PR, Lyft PR, or lots of the press from focusing on adjusted EBITDA as an accurate measure of profitability.

Lyft declined to comment

Correction: An earlier version of this article mischaracterized reconciliations in Uber's earnings report. Motherboard regrets the error.