Financial markets are a fascinating social mechanism. When they work well, large numbers of buyers and sellers come together and employ a diverse set of investment philosophies to debate an asset’s intrinsic value. If investors do this without bias, the prices that emerge are, at least on average, correct.

Prior to going public, Lyft was priced by one style of investor – the growth-oriented venture capitalist. VCs are a notoriously optimistic bunch, which often leads to valuations in the private markets that require a lot of things to go right as a company’s story unfolds.

Now that Lyft is publicly traded, a multitude of investment styles have been brought to bear on the analysis of the company’s intrinsic value. As of this writing, that judgement has sent the stock down 25% from its IPO. In this article, I highlight the critical factors that I believe should shape the debate.

Three critical factors that will determine value

A while back I came across a presentation by Aswath Damodaran, a professor at NYU’s Stern School of Business, that laid out a framework for valuing companies based on their users. While this model is not commonly used among the VC community, I believe it offers an alternative approach to valuation that clarifies the drivers of Lyft’s fundamental worth. You can find Professor Damodaran’s presentation here.

Evaluating Lyft from a user-based perspective allows us to identify and analyze three key components that bulls and bears must debate regarding Lyft’s prospects: (1) The unit economics of servicing existing riders versus obtaining new ones, (2) the benefit of Lyft’s network effects, and (3) the size of market available to Lyft considering both their future competition and competitive moat.

The cost of servicing a rider – probably the most important part of the story

Professor Damodaran’s approach starts with the idea that not all losses are created equal. Companies that are spending a lot to acquire new users – what should arguably be considered capital expenses – may show GAAP losses while the underlying business of servicing their existing userbase is profitable. Applying this analysis to Lyft generates some interesting insights.

Lyft’s gross billings per rider, or the amount the average rider spends per year on Lyft rides, came in at $492 in FY 2018 based on an average of 16.375 million riders throughout the year (I average the number of riders at quarter end to get the yearly average). Of this amount 26.8% was converted into revenue for Lyft, or $132 per rider.

Determining the amount Lyft spent on servicing this existing rider revenue requires some subjective judgements. For this analysis I allocated all general and administrative expenses and 50% of research and development to corporate overhead. I allocated 95% of sales and marketing expenses to obtaining new riders. The rest of the expenses – consisting of the cost of revenue, operations and support expenses, the remaining 50% of R&D, and 5% of sales and marketing – were allocated to existing riders. Based on these allocations I estimate the amount spent to service existing riders to be around 57% of total expenses, or 82% of total revenues. This allocation is obviously based on assumptions and is open to debate.

These numbers result in annual operating expenses of $108 per active rider in 2018, which leaves Lyft with an estimated pre-tax operating profit per existing rider of around $24 in the latest fiscal year. The remaining operating expenses (excluding the portion allocated to corporate overhead) can be attributed to obtaining new riders. This means that in 2018 Lyft appears to have spent around $127 to acquire a new rider.

Based on these numbers, Lyft seems to be earning just under an 18% pre-tax operating margin on its existing userbase, and just over an 18% pre-tax return on the capital it employs to acquire a new rider. Assuming a 22% tax rate, this results in a 14% return on existing riders after tax.

Why these numbers matter

These numbers are important for several reasons.

First, the profitability and value of the existing userbase is a crucial determinant of Lyft’s overall firm value. If you are evaluating any user-based company, in fact, you should ask yourself this question: If the current users are not generating enough value, and there is no credible path to improved profitability, then what is the point of acquiring more users?

Ostensibly, Lyft’s existing rider economics look good – an 18% pre-tax return on incremental capital is nothing to scoff at. I would argue, however, that the current valuation demands even higher profitability from its current userbase, especially when you consider other factors like rider churn. Lyft can improve profitability on existing riders through higher growth in billings-per-user, higher revenue share from gross billings, higher margins on existing riders, or some combination of all these. As Professor Damodaran has pointed out, ultimately you want growth to be driven by your existing users using the service more. As the intensity of existing users increases, the existing business hopefully scales and becomes more profitable. When this happens the benefit of acquiring new users–since a user is now worth more–increases.

To put some numbers on the current situation, there were 619.4 million rides taken on Lyft in 2018, a 65% increase over the 375.6 million rides taken in 2019. Of this increase, however, only 30% came from existing riders riding more, while the remaining 70% came from new riders moving onto the platform. Since Lyft’s revenue stems directly from the number of rides it provides, we can infer from these numbers that most of Lyft’s revenue growth is being driven by acquiring new riders. I think it is fair to say that the market is questioning whether these riders will be profitable enough to justify the money being spent to acquire them.

If you are bullish on Lyft at its current valuation, then, you must believe that there will be a dramatic improvement in operating profit per existing rider as the company grows. This improvement will be a function of (1) the growth in rides, along with any ancillary services, among its existing userbase and the prices it can charge these users, (2) its overall revenue share from ridesharing (a function of how much it must pay its drivers), and (3) how well its expenses to service existing riders scale as the business grows. The first two of these will be determined by the degree of competition the company faces as the market for ride sharing evolves.

What we can infer about Lyft’s network effects from the cost of acquiring a new user

A key benefit from Lyft’s much-touted network effect is the ability to acquire new users over time at a lower cost than potential entrants can match. If Lyft is benefiting from such a network effect, we would expect its historic customer acquisition cost to be higher and reflect a decline as the company has grown.

Lyft doesn’t publish customer acquisition costs in its financials, but one way to estimate the number is to look at the total amount the company has lost from its inception on a per-user basis. Lyft shows an accumulated deficit of $2.945 billion on its FY 2018 balance sheet. With 18.6 million active riders at the end of 2018, that equates to around $158 per rider acquired on average since inception. The same figure as of year-end 2017 is $161.

Comparing these numbers to the $127 customer acquisition cost we just estimated, it appears that the cost of acquiring a customer has indeed come down over time. This is positive for the bulls, although they should keep a close watch on how this trend evolves – the less the cost of acquiring a new rider scales, the more important it is for the profitability of existing riders to improve.

How big and contested will the market ultimately be?

Another reason investors should pay attention to the user economics is to understand the incentive for entry into the ride sharing / car service market. A key part of the bull thesis for Lyft is that the market for ride sharing will be much larger than the historic market for car hire services, both because the ride sharing experience opens the historic market up to new users, and because ride sharing is becoming an alternative to car ownership. The bull case is a big, growing market story.

Large growing markets are good in the sense that the size of the ultimate pie is increasing, but they are bad in the sense that potential entrants can operate at an efficient scale with a smaller portion of overall market share. As retired Columbia Business School professor Bruce Greenwald has said, growing markets can be the enemy of profitability.

Consider the incentives for potential entrants into the ride share / car service market. Based on the current enterprise values for Lyft and Uber, the market is valuing a rider at around $670 for Lyft and around $550-$600 for Uber (depending on how you value Uber’s holdings in Didi, Grab, and Yandex). If we believe that the cost of acquiring a new rider for the incumbents is around $130, then even at three to four times that number there is an incentive for competition to enter the market. This competition could come in many forms – especially given the unknowns surrounding autonomous driving. In addition to Lyft and Uber, companies like Tesla and Alphabet are lining up to try and take their share. The key for Lyft will be the durability of its economic moat. That will be determined by how sticky the userbase ultimately is.

A lot of unknowns

Warren Buffett has a box on his desk marked “too hard” for all the investment ideas that are too difficult to get comfortable with. When you pull out the factors that will ultimately drive the valuation of companies like Lyft, it becomes obvious to me that many of these unknowns belong in that box.

The probability implied by the current market price that the company will be able to dramatically improve its operating profit per user, in my opinion, is far too high. The implied growth in billings per user, again, is too high for a conservative valuation. That said, by focusing on the range of values the company can have under different scenarios surrounding these factors, we can start to get a sense of what kinds of valuations might provide a margin of safety. Facebook got to that point. In time, perhaps Lyft will as well.

~source