Monday, 13 May 2019

Uber, delusion, and ride-hailing's structural economic inefficiency

Uber's much-anticipated IPO occurred on Friday, and to the surprise of almost no one (sensible), the stock ended the session down 7% (despite the offer having already been priced at a discount to initial ambitions, after Lyft's weak post-IPO showing). In my view, the stock likely has much further to fall, and in many respects the Uber story is the perfect exemplar of this cycle's excesses.

The most extraordinary thing about Uber is that despite all the hype and putative 'wealth creation' it has generated to date, which has propelled its valuation to some US$80bn - since it's founding the company has done nothing but burn copious amounts of cash, and is yet to prove it even has a viable business model. The company is currently losing US$3-4bn a year on US$11bn in revenue, and revenue growth is already sharply slowing (already down to the teens YoY, and relatively flat sequentially).

In order to deflect attention from its unflattering financials and slowing growth, the company has tried to sell investors on the idea that an investment in Uber is an investment in the 'future of mobility'. The company envisages a world where private vehicle ownership ceases, and people simply use ride-hailing for their transportation needs - i.e. 'transportation as a service' (TaaS). However, this vision of the future ignores the inconvenient economic realities discussed below, which are already manifestly evident in the company's substantial operating losses, and worse, for the techno-optimists, the coming arrival of full vehicle autonomy (self-driving cars) is actually an existential threat to whatever residual of a viable business model does currently exist. Far from being the future of mobility, Uber is actually on the eve of being technologically disrupted. Uber investors are actually going short the true future of mobility. 

The bulls believe that Uber's current losses reflect an inefficient level of scale. Given the economic realities I discuss below, there is actually a good chance that the truth is the opposite - in order for Uber to survive, it will need to shrink to profitability. As I discuss, there are potentially profitable niche applications for Uber, but any attempts at the mass-market domination of private transportation are likely to only be met with (and indeed have been met with) mushrooming losses. Ultimately, Uber may prove to be worth only a handful of billions (>90% less than it's current valuation), as it profitably services a few niche markets, but at low margins, due to vigorous competition, up until full vehicle autonomy arrives, in which case there is a good chance the company will be put out of business.

So what are these economic realities I speak of, and why don't ride-hailing economics stack up? Let's first consider food delivery, which does make economic sense. If I want to order food from given location X, it is more economically efficient that a driver already in close proximity to the restaurant heads to the restaurant to pick up my food, and then delivers it to me. The total travel time involved might be, say, 5 minutes for the driver to head to the restaurant, and then 15 minutes from the restaurant to my house, whereas if I was to go and pick the food up myself, it would involve a 30 minute round trip.

The aggregate level of human transportation labour involved is therefore 20 minutes instead of 30 minutes. This creates a meaningful labour-saving gain from outsourcing. Restaurants also have economies of scale in ingredient procurement and cooking, and it is therefore economically rational for food preparation to be outsourced. Food delivery is therefore a rational economic service. Perhaps for this reason, there are currently profitable food-delivery companies - albeit that profit margins are very thin, as the service is commoditized and easy to replicate, and barriers to entry are low.

However, the economics of ride-hailing are different. If I want to go from A to B, I can drive myself, and it might take say 15 minutes. However, if I hail a car, it may take 5 minutes for the car to travel from the hailing point to my location, increasing the level of driver labour to 20 minutes. However, it gets much worse, because the real total is 35 minutes, because I still need to sit in the car for 15 minutes while I'm being driven, and thus contribute an additional 15 minutes of 'commuting labour' (particularly because reading/working in a car as a passenger is nausea-inducing, and both drivers and passengers can listen to the radio/music/podcasts). A sum total of 35 minutes of human time are now need to ferry me from A to B, instead of 15 minutes, which is an extremely wasteful undertaking, and loss of aggregate efficiency/productivity for the economy.

This is the core reason why private car ownership and operation has always been the primary means by which people have commuted by automobile, instead of the entire industry having become dominated by outsourced private taxis. On account of their structurally higher economic cost, the latter have had to confine themselves to niche applications, such as ferrying people to and from the airport; providing short distance commutes in CBD areas where parking is expensive and scarce (e.g. in NYC); and transporting intoxicated drivers home.

Labour is expensive, and few people are both willing and able to afford to pay someone $10-15 an hour to drive them around when they can easily do it themselves, at no additional personal time cost. Private chauffeurs are rare for the same reason private cooks and private butlers are. It's an economically inefficient/wasteful use of human labour, and by definition only the rich will ever be able to afford to hire such people, because it's a fallacy of composition to believe we can all have private 'help'. Uber's mass-market ride-hailing ambitions are therefore completely unrealistic.

Furthermore, it was reported in this morning's WSJ that there have recently been widespread protests by Uber drivers, bemoaning low wages. In the US, some drivers make as little as $5 an hour, and the majority no more than low-teens - at or even below the minimum wage in many US markets, despite foregoing many mandated employee benefits (as they operate as outsourced contractors). Already, Uber's drivers are earning uneconomically low wages (and thereby providing a significant labour cost subsidy), the yet the company is still absorbing very significant operating losses. And despite the company providing an additional 'company subsidy' of US$3-4bn a year (operating losses) to customers, on US$11bn in revenue, the company's top-line growth is already radically slowing, despite Uber having only tapped a tiny fraction of the putative total addressable market it is targeting.

These outcomes - along with the fact that Uber has had to absorb mountainous losses in order to grow over the years - are easy understood when these economic realities are considered: Because ride-hailing is structurally economically inefficient, the only way the company is able to generate mass-market demand for their services is by underpaying drivers and underpricing its services (the service is likely about 30% underpriced at present - that is to say prices need to be raised by about 50% - if one considers the hikes that would be required both to pay drivers sustainable wages and make a low-single-digit positive operating margin). They have compensated for the structural economic inefficiency by not pricing that inefficiency into the cost of their services.

The question is, what would demand look like if they raised prices 50%? The answer is, many of their current customers would probably stop using the service, or use it much less frequently. But if revenues started to decline, it would bust the growth narrative, and the company's valuation would collapse. This explains why the company has continued to lose copious amounts of money - there has been no other way for them to maintain the illusion of this being a viable, high-growth tech company, and thereby continue to inflate the valuation. Uber has tried to reign in losses some over the past few years, to get the company 'IPO ready', but have mostly succeeded in merely slowing top-line growth and stabilising losses at a still extraordinary-high level.

Is there still a place in the world for Uber? In the medium term, likely yes. Certainly, the service is much better than conventional taxis in the pre-Uber world, and despite their high cost, the world has needed taxes for niche applications for decades. The highest value-add is in highly built-up areas with limited parking availability - the ability to avoid both the time and cost of looking for parking is a meaningful economic gain from taxies/ride-hailing. So there is definitely room for an Uber type service.

However, NYC and Silicon Valley based investors forget that the majority of the world doesn't live in these areas, and Uber is radically overestimating the size of its economically addressable market. Most people live in suburbia, where it is easy to commute to work and where there is ample, free and convenient parking available. I'm reminded every time I go back to New Zealand to spend time with my parents that private vehicle ownership isn't going to cease any time soon. Commuting with one's own car is cheap, reliable, fast, and comfortable. Why wait around for an expensive Uber when you can just whip yourself down the road to the local store or restaurant in 5 minutes, and park for free outside on the road or in designated parking areas outside?

To what extent it will be Uber ultimately providing these niche services is also unclear. Uber certainly had an initial lead in technology, but traditional taxi companies have now had ample time to develop their own online apps with comparable services, and have also had to reduce prices. When Uber increases its prices to more economic levels, traditional taxi companies will be in a strong position to defend their businesses and fight back. Many other local and international ride-hailing competitors have also emerged and replicated Uber's customer value proposition. Uber originally had a technological lead, but has now lost it, and in capitalism, it doesn't matter how much you've innovated and improved the world if other people can easily copy you.

Uber also has a very limited 'network' effect, because drivers can and do jump to whatever platform offers them the best terms - indeed most Uber drivers use all available platforms, and they accept rides from the platform offering them the highest rates - and customers can do the same (most customers have multiple ride-hailing apps on the phones, and can easily choose the cheapest). This means that even if Uber survives, it will likely always remain an extremely low margin business.

Uber's vision of being the globally dominant brand for ride-hailing has also already been quashed. In many markets they were outcompeted by more nimble and aggressive local competitors, which has forced Uber to merge and/or leave many markets (they no longer exist in Singapore/South East Asia, for instance, having been forced to merge with local rival Grab, which is the surviving brand, in exchange for a minority stake; the same occurred in China with Didi). Given this pronounced forced retreat in the company's global ambitions - essentially the early adoption of the 'shrink to profitability' strategy mentioned earlier - it is remarkable the company's private market valuation doubled again during this period - a true testament to the extent of the current VC-funded tech bubble.

Uber can probably survive, but it will likely need to significantly shrink the size and breadth of its operations down to a few profitable core niches/markets (a strategy the company has already commenced), and it may ultimately end up being worth only a handful of billions of dollars, as the company operates a low margin business on a much smaller addressable market, where it also needs to contend with many other vigorous competitors, including not only other ride-sharing apps, but also newly-energized traditional taxi companies.

All this is - however - before considering the potentially devastating effect the arrival of full vehicle autonomy will have on Uber's business. The delusional Masayoshi Son has invested aggressively in Uber and sees it as part of his vision of participating in the coming technological singularity, but the truth is that Uber probably won't even survive the transition to autonomous cars, let alone the singularity (the latter of which is also another delusion, reflecting widespread technological hubris and "Moore's Curse" - the inappropriate use of the analogy of rising transistor density to typify coming exponential progress in other areas).

Many many well-resourced organisations are working on developing full vehicle autonomy, and it is most unlikely Uber will be the first past the post. And whoever develops the technology first will be able to offer a much more affordable fleet of fully-autonomous ride-hailing vehicles, bankrupting human-based ride-sharing platforms. Furthermore, even if Uber did happen to be first, longer term there will be many providers of autonomous fleets operating the same commoditized service, where the only entry barrier is capital (how many cars do you want to fund into a fleet in a given location). It's likely to result in an extremely capital intensive, highly competitive and low profitability industry.

Furthermore, if your own private car can autonomously drive you to and from your destination, and after dropping you by the door, can go and park itself, there is little value-add provided by ride-hailing at all - you can just use your own car, even if you're intoxicated (the latter feature will likely significantly reduce demand for taxis/ride-hailing, as this is a key use case). Why wait for someone else's car, when you can leave immediately in your own, the model of which you can select to match your own preferences with respect to comfort, style, size, safety, performance, and cost. Personal cars have been used for a very long time as an expression of personal identity, as well as to show off wealth and status, and there is a good chance that continues, and indeed makes a resurgence once cars are fully autonomous (much has been made of Millennials embrace of ride-hailing and eschewing private vehicle ownership, but a lot of this reflects the high cost of private vehicle ownership for cash-strapped Millennials, and the cost and hassle of obtaining drivers' licenses).

How this all shakes out remains to be seen, but it is far from clear Uber will survive this transition at all (although I tend to believe full autonomy is much further off than the technology bulls believe - probably at least a decade). And depending on the pace of development, Uber may not even succeed in making a profit before this point arrives, which means that it may not succeed in ever making a profit, period.

Uber is an extraordinary testament to the level of delusion and sheer irrationality that currently exists in the world of privately-funded tech. At this point, the extent of the irrational exuberance is very much approaching the level of the dot.com mania. The post-IPO sell-offs in Uber and Lyft will no doubt be ascribed to the 'short sightedness' of public markets, but the truth is that even in their current state of infatuation with high-growth tech companies, public markets are still more rational than private tech markets, and that is why the stocks are falling. A reckoning is coming in the private tech market, and the rush of tech IPOs this year suggests the tipping point may well be nigh.


Uber looks to be a fairly obvious short to me.


LT3000

11 comments:

  1. This covers it.

    Firstly, Uber does present a network effect as a barrier to entry — albeit an indirect one. This network effect is a cross-side network effect between riders and drivers. However, the network effect has limitations. It appears to have limits in scale — exhausting at the national level in practice, and potentially the metropolitan level in theory. Additionally, Uber’s network effect is limited in breadth — based purely around rider/driver density and corresponding pickup times, utilization rate, and price. These limitations, along with the habitual nature of purchases in the ride-sharing market have placed the onus on being the first-to-market as a way of catalyzing the network effect and its associated competitive advantage. This emphasis on being first-to-market appears to explain the discrepancy between Uber’s national dominance but international retrenchment.

    https://medium.com/@heosphoros22/uber-and-the-network-effect-b6a4395c58ab

    ReplyDelete
  2. Agree with the barrier to entry point but I think the inefficiency argument for Uber isnt quite right. Personally Uber can save significant time avoiding parking (time and cost), furthermore it increases utilisation of idle vehicles.

    ReplyDelete
  3. take a look to LYFT a small palyer in a niche market

    ReplyDelete
  4. I believe you've omitted one efficiency taxis/ride-hailing provide - cars tend to depreciate not only with miles covered, but also with just time (both for rational and irrational reasons). So usage of taxis brings down depreciation costs by increasing car utilization.
    But that effect is probably minor.

    ReplyDelete
    Replies
    1. That is a fair point, but from the standpoint of economic efficiency, a car depreciates with use, and as noted in the article, ride-hailing actually increases the cumulative miles traveled.

      Granted that an unused car will still decline in value as it ages. However, bear in mind that Uber doesn't take care of all use cases for many customers (e.g. long distance road trips etc). So a meaningful proportion of their customers will still own a (now less utilised) car, which is an offsetting factor that acts to reduce the overall utilisation of the fleet.

      I'm not arguing that taxis/ride-sharing have no value-creating use cases. They do - that's why we have long had taxis. I'm arguing that that does not extrapolate across to the vast majority of private vehicle transportation.

      Uber is just a glorified taxi company. Definitely a better one to what existed before it's creation, but at base it is just a taxi company. They had an initial technological edge, but all taxi companies will develop their own app and/or partner with ride-hailing companies, and in conjunction with the lack of lock-in network effect, Uber's service won't be anything special. Everyone will compete on price and margins will be very low.

      Delete
    2. PS also consider that the depreciation effect is very small relative to the cost of labour. An average car costs 50% of an average annual wage, and lasts for 15-20 years. The all in cost of running a car is about 15c a km (including fuel costs, R&M and depreciated value). The depreciation component is small relative to the excess labour costs involved.

      Delete
  5. To those in the comments referring to "the idle vehicle argument", that became redundant when finance companies started offering car loans specifically for Uber drivers. If anything Uber has contributed to MORE cars being on the road, not better utilising existing ones.

    ReplyDelete
  6. I think what you're missing is that the industry likely ultimately settles as a duopoly (in the US). Sure, new entrants may enter but for them to compete, they will have to burn massive amounts of cash until they get to scale. So the question then becomes what does steady-state industry profitability look like and how is that split btwn the #1 and #2 player. I'm not sure what the answer is but I do believe that driver incentives will scale down meaningfully either way. For example, see Yandex Taxi in Russia, which is a monopoly and already profitable.

    ReplyDelete
  7. Self-driving vehicle can also be part of their long term strategy. There is no conflict in that perspective

    ReplyDelete
  8. What are you thoughts on bike / scooter sharing models like Lime, Bird, Ofo, etc?

    ReplyDelete