Autonomous Driving Has A Little 'Dot-Com Bubble' To It

by: Vince Martin


Autonomous driving optimism has boosted a number of stocks in 2017.

But some of the current sentiment seems like a much (much!) smaller version of the dot-com bubble in a few interesting ways.

It's too early to short even if that's the case - but long-term investors buying in to the trend should be careful.

Optimism toward autonomous driving has boosted a number of stocks in 2017 - even if most of the biggest winners from the trend so far admittedly have other tailwinds. Nvidia (NVDA) has gained 34% YTD, and doubled since the election. Alphabet's (NASDAQ:GOOG) (GOOGL) Waymo has been valued at $70 billion. Even after a recent decline, Tesla (TSLA) is worth $54 billion - more than Ford (F) or General Motors (GM), at least on a market cap basis - and has added almost $25 billion in market value since early December.

Those are just a couple of examples. BlackBerry (BBRY) was called the "next Nvidia" - by a well-known short seller, no less. Mobileye (MBLY) got $15 billion from Intel (INTC). Even Avis Budget Group (CAR) - which in theory would be a victim of a true self-driving revolution - got a double-digit bump last month off a simple partnership with Waymo.

From here, the optimism looks overwrought. In fact, it reminds me of the 1998-2000 dot-com bubble - albeit in kind, and not in degree. (To repeat: not in degree, or anywhere close. I saw the dot-com bubble play out live at a tech-focused retail brokerage, and I've long argued that its sheer insanity has somehow become underestimated in hindsight.) But there are three aspects of the current sentiment toward autonomous driving that I find concerningly reminiscent of the Internet boom:

  1. The future trend is real, but...
  2. Not everyone will win, and...
  3. Timing matters.

And all three aspects suggest that investors might be getting too optimistic, too quickly.

Autonomous Driving Is The Future - But The Internet Was, Too

The problem with the dot-com bubble wasn't the 'dot-com' part. The core optimism toward 'Internet stocks' turned out to be correct at least in some sense. E-commerce revenue went from zero to $400 billion last year. The Internet has changed social life (Facebook (FB)), business, media (Netflix (NFLX)), and myriad other facets of modern civilization. Dot-com stocks weren't Pogs or Beanie Babies. There was some logic behind the broad thesis, and there was real potential, some of which turned out to be realized eventually.

For example, think of some of the classic busts of the era. and its silly sock puppet often make the list of "biggest busts." But just sold for $3.35 billion in the largest e-commerce acquisition ever. Selling pet food and pet supplies over the Internet is a real business model - simply didn't figure it out. Neither did the other four public companies with a similar model. Problems with spending (like's Super Bowl commercial) and unit economics decimated most dot-com flops once the tide turned. But in some, if not many, cases, the businesses models weren't wrong, just early.

Peapod still exists, albeit in a different corporate form. Beenz and Flooz look like Bitcoin predecessors if you squint hard enough. The reason the dot-com bubble busted wasn't because the dot-com era never arrived. It was because the projections were too grand, too broad, and too early.

Some of the discussion around autonomous driving sounds rather similar - particularly the 'too broad' and 'too early' parts. I have little doubt that at some point self-driving cars will be ubiquitous, or close. But one lesson of the dot-com bubble is that even if the trend plays out, it doesn't mean that every stock tangentially related to that trend will benefit. Yet some of the recent moves in stocks even in the vicinity of autonomous driving suggests that type of widespread optimism is starting to be priced in.

The move in Avis Budget Group stock toward the end of last month seemed a perfect example of this effect. Avis Budget signed a deal with Waymo to "service and store" 600 minivans for Waymo in Phoenix, where Waymo has a pilot program.

It's not a big deal, at all. It does nothing to address the near- to mid-term threat from Uber (UBER) and Lyft (LYFT) to the rental car model, let alone the long-term potential of self-driving cars. And yet Credit Suisse is citing it as evidence of rental car owners becoming fleet operators in a "driverless future."

Avis CEO Larry De Shon is talking up his company's ability to offer "fleet-management-as-a-service," adding the "-aaS" buzzword to an opportunity that is at least a decade away from being a material contributor to revenue in the most bullish case. It's reminiscent of the gains every stock seemed to make in 1998 and 1999 every time they rolled out a website. And it highlights my second point (somewhat).

Not Everyone Will Win

There's a key difference between the Internet and self-driving cars, at least in retrospect. For the most part, particularly for consumer-facing sites, the Internet has been a 'winner-take-all' environment. Whether it's (AMZN) in books, or eBay (EBAY) in auctions, or Google in search, in many major verticals, there's basically one dominant player.

At the moment, that may not be the case in autonomous driving, assuming it plays out to the extent that many project. Nvidia and Intel/Mobileye may both have major automotive chip businesses with different manufacturers. Uber and Waymo and Tesla all could have their own fleets and/or target different aspects of the transportation space. It seems unlikely that there will be one dominant manufacturer making cars that use one primary software platform, which runs on one single company's chips. (It is possible, of course.)

But that, too, raises questions about some of the valuations in the space. Waymo is worth $70 billion now, according to Morgan Stanley. Uber reportedly is valued at $60 billion-plus. Nvidia has added $45 billion in market value since mid-November, with some of that coming from automotive enthusiasm. (Obviously, there are opportunities in automotive beyond Level 5 autonomous, but it seems likely that the self-driving market is creating some of the optimism toward NVDA shares.)

It's a huge amount of value creation for an opportunity that in some cases is a decade out - at least. (More on that in a moment.) I'd add, too, that one of the key differences between autonomous driving and other major secular trends is that autonomous driving should shrink the automotive market. A 'driverless future', as Credit Suisse termed it, would require far fewer cars.

There's little reason for 10 people to own 10 cars if cars can be ordered on demand quickly, efficiently, and cheaply. That in and of itself creates an odd disconnect between the valuations assigned to Ford and GM stock - which imply that their earnings may have peaked for good - and those assigned companies that supposedly will benefit from autonomous driving.

Fewer cars, by definition, should shrink the market for suppliers like Nvidia and Intel, or fleet managers like Uber, Waymo, and (apparently) Avis. Isn't that way so many auto parts suppliers trade at single-digit forward P/E multiples? Yet the market is acting in some cases as if every single supplier, and every single manager, has a huge opportunity that should be reflected in the current share price.

And some of the valuations seem to assume that the market opportunity grows substantially: Morgan Stanley's $70 billion estimate of Waymo's valuation was based on the company generating $1.25 per mile in revenue. That's $12,500 for a car traveling 10,000 miles a year - a figure that seems ridiculously high, particularly if it's Fiat Chrysler (FCAU) making the cars, as appears to be the case.

If the transportation market in fact shrinks - and, again, that appears to be the logical long-term conclusion - then it certainly looks like bullish projections are getting overwrought. And that makes it harder and harder to justify some of the recent moves, given that any profits from a 'true' autonomous market need to be discounted back over 10 years - at least.

Technology vs. People

Where I see the most interesting parallel to the dot-com boom is in the focus on technological capabilities while bulls ignore the problem of consumer adoption. Again, one key issue in the dot-com bubble was that investors and analysts massively overestimated how quickly consumers would move online. At the peak of the bubble (in February 2000), Forester Research projected $184 billion in e-commerce sales by 2004. The actual figure was $69 billion.

Uber and Waymo and whoever else is in the 'Taxi 2.0' business may be thrilled about autonomous driving. But I'd expect the ability of autonomous driving to penetrate the broader transportation industry to be much slower than tech evangelists believe at the moment. Because that's always the case. Technology isn't adopted when it's ready. People change slower than technology does - and the faster technology changes, the greater the gap becomes.

E-commerce capabilities were in full force by the early part of the 2000s, at the latest. But it was more than a decade before its impact was fully felt in the brick-and-mortar world. That's not because e-commerce providers somehow got better over that time (though that did help). The major reason is because people take time to change their habits.

The problem seems particularly acute in autonomous driving. A massive amount of customers have little to no interest in having their cars drive themselves. There are millions of drivers who will refuse to cede control of their vehicle to a 'robot'. There are many more who simply enjoy driving.

There will be significant pushback from libertarians and government/corporate skeptics (both on the left and right) who are fearful of having every car being tracked at all times. There are significant, and legitimate, concerns about IT security: if Home Depot (HD) can't protect its consumers' credit cards, how sure are we that Uber can protect the code that is driving vehicles at 65 miles per hour?

And drivers likely will have a difficult time trusting robots, no matter what the data shows (or how many lives technologists correctly argue could be saved). As the old saw goes, everyone believes they're above-average at two things: driving and sex. One single fatality from a car operating under Tesla's Autopilot drew national attention - even though the subsequent investigation showed the driver had at least seven seconds to respond. Does anyone believe that the American public - or the rest of the world - is ready to happily (if metaphorically) hand over their keys and sleep in the backseat of a robot car?

In terms of so-called Level 4 autonomous driving, these issues are somewhat less pressing. Drivers can turn those features off, or at the least still drive in the way in which they are accustomed. But in terms of the grand bull case for many stocks with opportunities in autonomous driving, it's Level 5 - purely self-driving - where the major opportunity supposedly lies. That's the 'driverless future', where the huge disruption would occur.

And it's there that there will be an enormous conflict between what the technology offers (or claims to offer) and what consumers are willing to accept. Anecdotally, I'd argue that the majority of drivers I know, if offered a Level 5 vehicle, would say, "No thanks." More than a few would say, "You can take my steering wheel out of my cold, dead hands." It will take years, if not decades, to change those attitudes.

That's years, if not decades, waiting for these grand profits, which means they're discounted even further in the future. And, as such, I'd argue that pricing in benefits from a true 'driverless future' into any stock at the moment is close to foolish.

What To Do?

Tesla CEO Elon Musk has said that his company could reach Level 5 driving by 2019. That very well could be optimistic. Other manufacturers have more muted expectations, and there's a long-running debate as to whether Tesla truly can reach Level 5 without LIDAR.

But even that's largely a technological debate - which is only part of the argument. Even if we assume that Level 5 was available today, I'd argue that it still would take years for autonomous driving to garner any real share of traffic - particularly from individual drivers. Many, if not most, of those drivers are going to have a real problem in ceding control of the car, even as technologists and experts argue that self-driving cars will vastly improve safety. At the end of the day, people move slower than technology does.

As such from an investment standpoint, I think it's foolish to assign any present value to any stock based on the 'true' autonomous opportunity. Obviously, the fact that Nvidia's automotive revenue grew 52% in 2016 and another 24% in Q1 matters. Waymo has some value. But looking at a 'driverless future' that is technologically probably a decade out - and from an adoption standpoint, possibly a quarter-century out - I'd consider the broader opportunities optionality at best.

In the near term, I'm not sure that's really actionable advice. Shorting, for example, NVDA because investors may be overoptimistic about self-driving cars only works if a) that thesis is right and b) investors realize they're too optimistic - which might take years. (It also ignores datacenter, which I think is actually the far more interesting, and valuable, opportunity for Nvidia - but that's a different article.) A lot of investors were right about the dot-com bubble in 1998 and 1999 and many of them lost huge amounts of money regardless.

And, again, there's nothing close to the dot-com bubble from a valuation standpoint. If Waymo is worth $50 billion less than investors believe at the moment, the value of GOOGL stock drops less than 10%. Tesla looks overvalued to my eye (yes, still), but still has an opportunity to support its valuation even if self-driving isn't adopted as soon as some think. Nvidia has other opportunities; most of BlackBerry's near-term automotive revenue will come from infotainment.

But I'd suggest caution in buying any stock based solely on autonomous driving enthusiasm. Any profits from that type of market are years off - and as a result need to be discounted substantially. It's not clear who will win - but it is almost certain that not everyone will win, and certainly not to the extent that their respective bulls suggest. Autonomous driving technology may arrive within a matter of years - but that doesn't mean self-driving cars will dominate within the same time frame. And, as history shows, it doesn't mean that massive upside is on the way for every stock with exposure to that technology.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.