Monday, October 12, 2015

On the Uber Rollercoaster: Narrative Tweaks, Twists and Turns!

A slightly abbreviated version of this post appeared in TechCrunch on October 9, 2015, and it is the first of series of three on the ride sharing business. In the second post, I value Lyft, the other ride sharing company, and also look at how ride sharing companies globally are being priced. In the third and final post, I look at ride sharing as a business and at four possible scenarios for its evolution. Those posts will also appear in TechCrunch over the next two weeks and will be reproduced here soon after.

It seems like ages ago, and perhaps even in a far-off galaxy, that I first valued Uber on my blog, but it was in June of 2014. One reason it seems like a lot of time has elapsed is that Uber has managed to be in the news, for good and bad reasons, almost all through this period. With each news story, the response is either rapturous or funereal, depending on the responder’s prior views on the company. Given how eventful this last year has been, I think it is time for me to revisit my estimates, eat some humble pie and redo my valuation.

A look back
I became interested in Uber after reading a news story in June 2014 that indicated that it had been valued at $17 billion in a venture capital round. I posted my first valuation of Uber in June 2014, viewing it as an urban car service company, with local (but not global) networking benefits. Assuming that it would increase the size of the urban car service market by about 40%, while preserving its low capital-investment business model, I valued Uber at just under $6 billion.

While some in the VC community were quick to dismiss the valuation, I will remain grateful to Bill Gurley for a post where he took me to task for having too narrow a vision of Uber’s business model. In his counter narrative, he argued that Uber was not just urban (it could create inroads in suburbia), not just a car service (it was in logistics & transportation) and that it was working with other businesses to create global networking benefits. Since Bill, as an early investor in Uber with access to its internal workings, clearly knew far more about the company than I did, I revalued Uber using his narrative and arrived at $54 billion as the value that reflected the narrative.

I was then taken to task by value investors who took issue with the value changing so dramatically from my assessment to his, and my response was that this was exactly what you should expect, early in the life of a company, where there is room for widely divergent narratives, and values that reflect these divergences. In December 2014, I tried to show this by creating a build-your-own-Uber valuation template, where I let readers choose Uber’s market (urban car service, all car service, logistics or mobility services), the effect it would have on that market’s size (from none to doubling it), the competitive advantages that would determine its cut of the ride receipts (from the existing 20% down to 5%), the networking benefits it would have (none, local, partial global, global) and business model (from its current no capital intensity to higher capital investments), and derived values for Uber, ranging from less than $1 billion to close to $100 billion.

The news keeps coming..
As I noted at the top of this post, it is an understatement to say that Uber has been in the news. Each week brings more Uber stories, with some containing good news for those who believe that the company is on a glide path to a $100 billion IPO, and some containing bad news, which evoke predictions of catastrophe from Uber doubters. For me, the test with each news story is to see how that story affects my narrative for Uber, and by extension, my estimate of its value. In keeping with this perspective, I broke the news stories down based upon narrative parts and valuation inputs.

The Total Market 
The news on the car service market has been mostly positive, indicating that the market is broader, bigger, growing faster and more global than I thought it was, even a year ago. 
  1. Not just urban and much bigger: While car service remains most popular in the urban areas, it is making inroads into exurbia and suburbia. The evidence for this lies not only in anecdotal evidence and the market capitalizations commanded by ride sharing companies, but also in the numbers that have been leaked by these companies. A presentation to potential investors in the company put Uber’s gross billings for 2015 at $10.84 billion. It is true that these are unofficial and may have some hype built into them, but even if that number over estimates revenues by 20% or 25%, that represents a jump of 400% from 2014 levels.
  2. Drawing in new customers: One reason for the increase in the car service market is that it is drawing in customers who would never have taken been in this market in the first place. While the evidence for this is still mostly anecdotal, another leaked report out of Uber indicates that ride sharing has created a market three to four times larger that the original taxi cab/ limo market in San Francisco, the city with the longest history with new ride sharing services. While San Francisco is unusual in terms of the high proportion of its population that is young, tech-savvy and single, the argument that ride sharing is increasing the size of the market elsewhere, though not to the same magnitude that it did in the Bay Area, seems to be a solid one.
  3. With more diverse offerings: The other reason for the jump in the size of the ride sharing market is that it is no longer just a cab service, but instead has expanded to include alternatives that expand choices, reduce costs (car pooling services) or increase flexibility.
  4. And going global: The biggest stories on ride sharing came out of Asia, as the ride sharing market has exploded in that part of the world, and especially so in India and China. That should really come as no surprise since these countries offers the trifecta for ride sharing opportunities: large urban populations, with limited car ownership and bad mass transit systems. 
The bad news on the car service market front has come mostly in the form of taxi driver strikers, regulatory bans and operating restrictions. Sao Paulo may be the latest city to restrict Uber, but it is part of a long list of cities or entire countries that have tried to ban or restrict ride sharing. Even that bad news, though, contains seeds of good news, since the status quo would not be trying so hard to stop the upstarts, if ride sharing was not working. In my view, the attempts by taxi operators, regulators and politicians to stop the ride sharing services reek of desperation, and the markets seem to reflect that. Not only have the revenues collected by taxi cabs in New York city dropped significantly in the last year, but so has the price of NYC cab medallions, dropping almost 40% in value (roughly $5 billion in the aggregate) in the last two years. While auto sales may not have been affected materially yet, there are worries that there will be a drop in sales in future years, as people replace a "second" car or even a "first" car with ride sharing services.

In my December post, I noted the possibility that Uber could move into other businesses. The good news is that it has delivered on this promise, offering logistics services in Hong Kong and New York, and food delivery service in Los Angeles. The bad news is that it has been slow going, partly because these are smaller businesses (than ride sharing) and partly because the competition is more organized. However, these new businesses have moved from just being possible to plausible, thus expanding the total market.

Bottom line: The total market for Uber is bigger than the urban car service market that I visualized in June 2014, and will attract new customers, and expand in new markets (with Asia becoming the focus), and perhaps even in new businesses.

Networking & Competitive Advantage (Market Share and revenue sharing)
The news on this front is more mixed. The good news is that the ride sharing companies have increased the cost of entry into the market, with tactics such as paying large amounts to drivers as sweeteners to sign up. In the US, Uber and Lyft have become the biggest players, and some of the competitors from last year have either faded away or been unable to keep up with these two. Outside the US, the good news for Uber is that it is not only in the mix almost everywhere in the world but that Lyft has, at least for the moment, decided to stay focused on the United States in its expansion choices. 

The bad news for Uber is that, especially in Asia, the competition has been intense, and that it is fighting against domestic ride sharing companies that dominate these markets, Ola in India, Didi Kuaidi in China and Grabtaxi in South East Asia. Some of the domestic company dominance can be attributed to these companies being first movers and understanding local markets better, but some of it also reflects that the market is tilted (by local investors, regulation and politics) towards local players. There is even talk, though it may be just that, that these competitors will band together to create a “not-Uber” network that can share resources and users, though the news story, two weeks ago, that Didi Kuaidi and Lyft were going to create a formal partnership adds heft to the thesis. To add to the mix, all of these ride sharing companies have been able to access capital at sky-high valuations, reducing the significant cash advantage that Uber had earlier in the process.

No matter how this process plays out, one of the key numbers that will be (and in some cases is already) under pressure, due to this more intense competition, is the sharing of the gross billing, currently set at 80% for the driver and 20% for the ride sharing business. In many US cities, where Lyft is challenging Uber most aggressively, Lyft is already offering drivers the opportunity to keep all of their earnings, if they drive more than 40 hours a week. While the threat of mutually assured destruction has kept both companies from directly challenging the 80/20 sharing rule, it is only a matter of time before that changes.

Bottom line: The ride sharing market is becoming competitive, but as costs of entry rise and the capital requirements become intense, it looks like this will be a market with fewer players with larger regional networking benefits and more capital.

Cost structure
This is the area where the most bad news has been delivered. Some of the pain has been from within the ride sharing business, as companies have taken to offering larger and larger up-front payments to drivers to get them to switch from competitors, pushing up this component of costs. Much of the cost pressure, though, has come from outside:
  1. Drivers as partial employees: Early in the summer, the California Labor Commission decided that Uber drivers were employees of the company, not independent contractors. That ruling was further affirmed by a court decision that Uber drivers could sue the company in a class action suit, and it is likely that there will be other jurisdictions where this fight will continue. While ride sharing companies may be able to delay the effect, it is almost inevitable that at the end of the process, drivers for ride sharing companies will be treated perhaps not as employees but as semi-employees, entitled to some (if not all) of the benefits of employees (leading to higher costs for ride sharing companies).
  2. The insurance blind spot will be filled: The other shoe that is poised to drop is the cost of car insurance. Ride sharing companies in their nascent years have been able to exploit the holes in auto insurance contracting, often just having to add supplemental insurance to the insurance that drivers already have. As both regulators/legislators and insurance companies try to fix this gap, it is very likely that drivers for ride sharing companies will soon have to buy more expensive insurance and that ride sharing companies will have to bear a portion of that cost.
  3. Fighting the empire is not cheap: Earlier in this post, I noted that the status quo (the taxi business and its regulators) was fighting back and that its cause was hopeless. However, the fight will still be expensive as the amount of money spent on lobbying and legal fees will increase, as new fronts open up. 
The evidence that costs are running far ahead of revenues again comes from leaked documents from the ride sharing companies. This one, for instance, shows that Uber was a money loser last year and in this one, the contribution margins (the profits after covering just variable costs) by city not only reveal big differences across cities, but are uniformly low (ranging from a high of 11.1% in Stockholm and Johannesburg to 3.5% in Seattle).

Bottom line: The costs of running a ride sharing business are high, and while some of these costs will drop, as business scales up, the operating margins are likely to be smaller than I anticipated just over a year ago.

Capital Intensity and Risk
The business model that I assumed when I first valued Uber was minimalist in its capital investment requirements, since Uber not only does not own the cars driven by its drivers but invests little in corporate offices or infrastructure. That translated into a high capital turnover ratio, with a dollar in capital generating five dollars in additional revenues. While that basic business model has not changed, ride sharing companies are recognizing one of the downsides of this low capital intensity model is that it has increased competition on other fronts. Thus, the high costs that Uber and Lyft are paying to sign up drivers can be viewed as a consequence of the business models that they have adopted, where drivers are free agents who are not bound to either company. 

For the moment, there is no sign that any of the ride sharing companies is interested in altering the dynamics of this model, by either upping its investment in infrastructure or in the cars themselves, but this news story about Uber hiring away the robotics faculty at Carnegie Mellon may be suggestive of change to come.

Bottom line: Ride sharing companies will continue with the low capital intensity model, for the moment, but the search for a competitive edge may result in a more capital intense model, requiring more investment to deliver sustainable growth.

Management culture
Though not a direct input into valuation, it is unquestionable that when investing in a young business, you should be aware of the management culture in that business. With Uber, your responses to the news stories about its management team will reflect your priors on the company. If you are predisposed to like the company, you will view it as confident in its attacks on new markets, aggressive in defending its turf, and creative in its counter-attacks. If you don’t like the company, the very same actions will be viewed as indicative of the arrogance of the company, its challenging a status quo will signal its unwillingness to play by the rules and its counter attacks will be viewed as overkill.

In New York, I saw this come into play when the mayor and city council decided to restrict the ride sharing companies (and Uber in particular) from expanding in the city, using the argument that it was worsening traffic conditions in the city. In response, Uber struck back with a counter publicity blitz which included not only radio and TV ads, but also add-ons to the Uber app that left no doubt in Uber users’ minds about how these restrictions would affect their choices. I thought that the DeBlasio option on the Uber app was clever, and while Uber won this battle with New York city, I wonder whether the scorched earth policies that it used will come back to hurt the company down the road.

Bottom line: While there are a few indications that Uber might be trying to soften its image, there seems no reason to believe that the company will become less aggressive in the future. The question of whether this will hurt them as they scale up remains unresolved.

Uber: An Updated Valuation
In summary, a great deal has changed since June 2014, partly because of real changes that have happened to the ride sharing market since then and partly because I had to fill in gaps in my ignorance about the market. I think that the best way to capture the shifts in my valuation is to compare my inputs on key numbers in June 2014 with my estimates in September 2015.
Uber valuation (6/14), Uber valuation (9/15)

I was wrong about Uber’s value in June 2014, when my estimate of $6 billion was below the $17 billion assessment by venture capitalists then. Correcting for both my cramped vision and the changes that have occurred since June 2014, my estimated value today is $23.4 billion.
Über valuation Spreadsheet (Ignore the umlaut, spell check did it..)

I know my estimated value lags the $51 billion value that VCs are attaching to the company today. This may very well be a reflection that my vision is still too cramped to capture Uber’s possible businesses, but it is what it is.

Fire away
When teaching and writing, my objective is to evoke interest and excite my audience, and failing that, to provoke dissent and incite argument, but the reaction that I dread the most is boredom. The very fact that you are reading this post still is good news, but if you are in complete agreement with my valuation of Uber, I have failed. I would rather that you fall into one of three groups: that you think my value is too high, that you feel it is too low or that you believe that I have no business even valuing Uber.
  • If you think I have over estimated the value of Uber, it should not be because it is losing money (given its growth trajectory, you should be suspicious if it did not), is trading at a high multiple of revenues (it should) or because your stable growth dividend discount model gives you too low a number (it is the wrong tool for a growth company). It should because you think the regulatory roadblocks will make the ride sharing market smaller than I have forecast it to be, and that competition will be much more intense (reducing market share and operating margins). 
  • If you think that I have under estimated the value of Uber (again), don’t blame DCF models for being biased against growth companies. The fault lies with me, and it has to be somewhere in my inputs, i.e., that I am not foreseeing other markets that Uber could enter, that its networking benefits are far stronger globally than I predict them to be (giving it a higher market share) or that the operating margins will bounce back to much healthier levels once they navigate their way through the growth phase. 
  • If your view is that I have no business valuing Uber because I am not a tech person, that I am not an expert in the ride sharing business and/or that I have not made money as a VC, my responses are guilty as charged, you are right and without a doubt. I don’t have a tech background, don’t work on the right coast and know technology only as a user, but I don’t think any of those are prerequisites for investing in technology companies. I have tried to incorporate what I have learned from technology companies in the market into the valuation, in the form of easier scaling up, larger networking benefits and bigger market effects, but I might not have done it well enough. There are many who know a great deal more about ride sharing than I do, and while I have tried to learn about the inner workings of this business from Harry Campbell and about the Chinese growth potential from Drake Ballew, my ride sharing know-how is limited. Finally, if we did restrict writing about the valuation of young companies only to venture capitalists who have been consistently successful over long periods, the list of potential writers would be very short, and most of the people on the list would be too busy investing in these startups to write about them. 
Whatever group you belong to, rather than complain about my mistakes (which are too many to count) or bemoan my limitations (which are legion), please take my Uber valuation and make it yours, putting your superior knowledge and experience into the numbers. In fact, let’s give this a crowd valuation twist and get a shared Google spreadsheet going, where you can post your results.

To be continued..
Notwithstanding the dressing down that I got from some in the technology/VC space for my first valuation of Uber, or perhaps because of it, there is no company that I have enjoyed valuing and talking about more during the last year, than Uber. Thecompany illustrates all the broad themes in valuation that I have returned in my blog posts and teaching: that uncertainty is part and parcel of valuation, that narratives drive numbers and that the pricing and valuation processes can yield different numbers. As a teacher, I am constantly on the look out for learning and teaching moments, and few companies have offered me more than Uber. 

To show you how much Uber has found its way into my every day thinking, I will end with a personal story. Towards the end of last summer, my youngest son, who is fifteen had a friend over for the afternoon, and when it was time for the friend to leave, I looked out at the driveway, expecting the “Mom car", the typical mode of transportation for a 15-year old in the middle of suburbia (where I live). When I saw a strange sedan with a bearded man in the driver’s seat, I was taken aback, until I was told that it was an Uber car. Since this happened only two months after my valuation of Uber in June 2014, where I labeled it an urban car service company, my first reaction after I got over my surprise was that I needed to revalue Uber afresh. Of such small actions are obsessions born!

YouTube Version



Anonymous said...

Well the VC's valuation of $51Bn is biased, especially if you've paid @ $17.

The more important thing to realise here is how any "amazing" business which has a +ive economic value add (ROIC > Cost of Capital), attracts new entrants and competition. Technology is both the advantage and disadvantage and it will drive all returns across this industry to the opportunity cost of capital.
Whether you look at Groupon or KING entertainment, Zynga. They all followed the same path.
Uber will inevitably follow the Amazon style of using every $ of cash for what management think of as growth. (Future) Shareholders value will never be seen in the traditional sense. Consumers may benefit if their business model enables cost savings, but personally, I don't see this. A traditional taxi - involves a car, a drive, and the fuel, marketing costs to get customers (?). Uber has the same operational costs. Ride sharing splits the cost per customer obviously, but we are yet to see if that takes off. If it does anyway, new entrants come in, and the cycle begins again!

Yehuda Fruchter said...

Great writing. But, this valuation post is a perfect example of the effects of anchoring bias. It is precisely this bias, that VC's have been able to exploit to drive the valuations of these unprofitable companies to insane levels.

Aswath Damodaran said...

Which of my anchors is causing the bias? In fact, I am shifting all my anchors (total market, market share, profit margin).

Marco Weihert said...

Hey All,

I totaly agree with Yehuda. It`s a big part of the VC business to do priming. Especially, as long as Uber is a privat company and not publicly traded every statement coming from management, supervisory board, investors, consultants etc. needs to be regarded with super suspicion, because they have really strong incentives to inflate valuation. (Inflated valuations don't hurt investors as much as the original founders due to liquidation preferences) Furthermore, we all know or can imagine that Uber is a big marketing&pr juggernaut.. that means all the news about Uber has to be regarded with suspicion as well. So, even if Mr. Gurley has some good company insights, he is biased and is a successful investor maybe because he is a master of priming! Every successful VC needs to be that!

But, I also agree with Aswath.. it's the nature of startups (young innovative companies working in absolute uncertainty) that valuations are closely tied to the narratives and vary extremely. One of many reasons for that is that terminal value can't be calculated seriously and is highly imprecise, because nobody knows what will happen in the future.. concerning Uber that means we don't know if the service will be banned in other countries as well (I come from germany where it's is banned), how long it will take that regulators react. etc... there are just to many unknown variables and variables we actually don't know that they exist or have an impact etc.

So, in terms of Uber it's nothing more than a bet if Uber can make an impact in the industries it is aiming for. If so, than it has to be crystal clear to everybody that it won't survive if it doesn't aim for a big junk of these markets.. that means if you think Uber can make an impact than you must expect that it gets uber huge.

Yehuda Fruchter said...

Thanks for your comment.

The VC's created the anchor by throwing some valuation number basically out of thin air. This anchor influences all future valuations, both consciously and unconsciously. It is almost impossible to ignore this anchor bias in future valuations, especially as many people attempt to "rationalize" the original valuation anchor.

For example, you say: "I was wrong about Uber’s value in June 2014, when my estimate of $6 billion was below the $17 billion assessment by venture capitalists then." Why were you wrong then? Simply because the VC's valued it at $17 billion? And then this valuation must influence your future valuation of the company.

The valuation methodologies used by VC's in many of these private start-ups now is very suspect. It's really just a ponzi scheme of sorts, ala Minsky. Virtually nobody analyzing these companies has audited financials. It's more of a trust me situation, with big numbers thrown out to support some fairy tale narrative.

Personally, I use Uber all the time, and it's great, but without access to any audited financials, I don't think it's possible to value this, and my gut, considering the many flaws in the business model, just tells me that the current valuation is simply absurd, especially when compared to other stable companies that are valued at $20 billion+.

Aswath Damodaran said...

I now get your point but that is not why I said I was wrong. It is because my valuation of the company has changed. That may reflect the pricing by VCs, but if you look at my history in valuing young companies, I have tried (though not always succeeded) in not letting the price feedback into my value. Thus, I valued Twitter around $20 when it went public and stuck with that value as the stock premiered and went even higher. With Apple, my valuation stayed around $600, as the stock gyrated higher or lower. Thus, if I revalue Uber next year, and the cost structure remains untenable, I would hope that I will have the courage to lower my value, even if VCs are pushing the price up. That is less an issue of anchoring and more one of faith in value.

The Oriole Way said...

One risk I think missing from this analysis is the sustainability of revenue (this is the same as its cost base, but from the other side of the ledger). Uber is obviously growing fast, provides a fantastically valued service, etc. But part of why it is 1) growing so fast and 2) perceived as so great is that its fares are so low. What happens if Uber has to raise fares by 50% or 100% in order to cover its costs, regardless of whether or not those costs are imposed upon it by regulators? I know that if my regular $10 rides were $15 I wouldn't ride nearly as often. It's not entirely for nothing that taxi fares are set at the levels they are.

Lankan Markets said...

Prof. in terms of competition I'd like to highlight a very small first mover in a very small market like Sri Lanka.

Lankan Markets said...

Prof. in terms of competition I'd like to highlight a very small first mover in a very small market like Sri Lanka.

Anonymous said...


The problem with all of these DCFs and arguments about valuations is that they are irrelevant (at least for the VC/Growth investors). All of these investments are some form of convertible preferred equity. Preferred Equity have several unique covenants that are unfamiliar to common stock investors: most importantly, Liquidiation Preferences and PIK accruing dividends (paid in preferred shares). these covenants allow investors to choose whether they want to convert their preferred shares into common shares OR take the payment stipulated by the LP (which typically ranges from 1-2x the invested capital) and the LP is compounded by the dilutive nature of the PIK accruing dividends.

So long story short, as VC/Growth investor my math ISN'T what the implied "valuation" of my round is but more so: Is the worst case liquidity scenario going to cover my LP (in which case I get at least all my money back and could even generate 10%+ IRR) AND is there some chance that the best case liquidity event blows everyone out of the water and we all get filthy rich? If the answer to the worst case scenario question is a "confident yes" and the answer to the best case scenario is "there's a chance", it's probably a deal worth doing.

The Term Sheet is 10x more important than a round's implied valuation.

Anonymous said...

In response to the above comment about liquidation preferences, the liquidation preferences of Uber stock are public. I've done the calculations with them, they aren't meaningful.

In my opinion, Uber is easily a 100B company. Easily. You can calculate what the value of the company is based on existing revenue streams but if you had done the same thing to Google 1 year in, where would you be? Nowhere. Uber is in its infancy. It has > 16 billion in cash. It has a ton of smart people working for it. It has a business model and business that is excellent.

It has many areas to grow. It has technology and change on its side to put winds in it sails. To bet against it is to bet against self-driving cars, automation, increased use of the internet, ordering things to be delivered wherever you are, electric cars and everything else.

It is the leader in its space, it has excellent execution, it has a smart management team and a ton of money. Any analysis of existing business is looking at a trailing number, their business in the past. It is not looking enough at the future.

The question is simply this. In 20 years, are you going to summon a car with your phone that will drive to you. Answer: yes. In 20 years, are trucks going to be driving on their own on the freeway? Answer: yes. In 20 years, is Uber could to be the company providing technology to car companies for basically everything? Answer: yes

How about integration with all other transport. Airlines, trains, etc. How about shipping? You want to connect the boat and automated loading docks to the self driving trucks and make a complete delivery service. All automated.

And there are all the other areas that we can't even imagine at this point. But give a group of technology people a core business they make money from and let them run with it and what do you get? You get a trillion dollar tech company in 2040.