Some thoughts on the new Internet bubble

Going back to my comments from yesterday about the utility of macro, I want to talk a bit about credit excesses and valuation manias. The overall gist here is that manias are endemic to our system because psychology plays a big part in social systems. And while it is debatable how well macro policy can “lean against”, it is clear that individual actors can see a mania unfold and act to prevent the wealth transfer and wealth destruction that unfolds in its wake.

For example, in the mid-2000s, it was clear to me that we were already in the midst of a housing bubble that was well advanced. I could not for the life of me understand why people would pay 40 or 50% more to buy a house than rent an equivalent one. I ran the numbers over and over again in the DC area and every time, it showed extreme overvaluation.

Now, at the time, I would say to people, “look, why should I buy this or that house, when I could rent it for 30 or 40% less.” The response I got usually boiled down to the concept that owning a home is worth paying more. OK, but how much more and why? No one could answer that question satisfactorily. But they weren’t bothered by that at all because that’s how manias work in distorting perspective and rationality.

A friend of mine who believed me would ask me, “how do I make money from this?” And while I had puts on the likes of Wamu and Countrywide Financial, I don’t think that’s necessarily the right way to look at it. A better way to look at it is to understand that it is hard to recover from large losses of capital. The goal is not necessarily to profit from the mania’s parabolic rise or its demise. Rather, the goal is to not lose capital, to avoid unnecessary losses by chasing yield or chasing return.

Recently, a lot of people have been talking about a stock market bubble and a bond bubble. I want to say a few words about these so-called bubbles.

About two weeks ago, I wrote the following:

“Large and slower-growing established players like Google, Apple, Cisco Systems and Microsoft, some of which were bubble stocks in the tech bubble, are not seeing anywhere near the explosive surge in shares that other less established players have seen. To the degree there is a bubble in technology developing, it is not hitting the likes of Oracle, Ebay, HP or Qualcomm. Yes, these companies have relatively high multiples compared to the overall market. But their growth is also higher. They are not bubble stocks by any means.

“On that score, a lot of this makes sense. A recent McKinsey study found that high-growth companies give investors returns that are five times greater than even medium-growth companies. So, right there you can see why we have this bifurcation. Some of these high growth companies are going to zero. But the ones that survive are going to be massive winners for their investors. And given the dynamism in the technology sector, there are a lot of options to choose from, just as there were during the Internet bubble days. And with all the liquidity chasing high returns for the few that will survive the inevitable shakeout, a ‘bubble’ is all but inevitable. I would argue, we are already seeing one.”

My point here is that we are in a land grab in the technology sector as the industry moves away from the PC to a more mobile and cloud-centric universe. The dynamics of that kind of paradigm shift make manias likely, even inevitable because an investment in companies coming to prominence in this space right now is like buying a call option with huge amounts of implied volatility. That optionality is worth a lot of money – so much that it creates the kind of price movements that naturally lead to bubbles or manias.

When I wrote about bubbles in early May I wrote that “the traditional economic simplification of price movements as independent of prior price movements is wrong. Market prices are not independent of prior price movements simply because markets are a social function in which actors base their actions in part on the actions of one another. We have momentum investors for a reason. And while price actions seem random when market observations occur close to the probability mean, the lack of independence in price movements becomes ever more important the further prices veer from the mean.”

In a market with a large number of new entrants with high optionality and high implied volatility, you are going to get price action that creates a psychology of herding, momentum, and eventually mania. What you choose to do as this occurs is the real question.

Look at Uber, the mobile ridesharing/taxi company, for example. I just took my first Uber ride this morning. I got picked up by a licensed taxi operator who charged me basically what he would have charged via a normal metered fare. The transaction was painless and automatic. I like the concept of Uber as a disintermediator. But what about valuation?

Uber raised a pre-IPO round of financing that valued the company at $18.2 billion. Given a reported $200 million revenue run rate, that’s a 90x revenue multiple, a number that compares very unfavorably to a reasonable 15-20x earnings multiple for a moderate growth company. In order to justify this kind of multiple, we need to see some heroic assumptions about market share, revenue growth, margins, barriers to entry and regulation.

NYU Professor Aswath Damodaran did a valuation on Uber that I found compelling and what he found was a value of just under $6 billion for Uber, even using very optimistic assumptions.

Uber valuation

The interesting bit here is that the majority of the valuation from Uber comes from the terminal value calculation in the discounted cash flow model that Damodaran uses. This implicitly means that Ubers real value is not from earnings over the next ten years but rather from the sustainable franchise it builds which will be a cash machine after that period and into the future. I have done hundreds of DCFs and all I can say is that any company whose value is all in the Terminal value had better have some serious barriers to entry to sustain its market position.

Christopher Mims at the Wall Street Journal says

Ride-sharing company Uber is great at execution and long on vision, but one thing investors who just valued the company at $18.2 billion might be overlooking is the fact that the “moat” around Uber’s services is incredibly shallow. This means competitors — and they will be legion — might have little trouble taking business away from the company and limiting its ultimate size.


The savviest Uber drivers I’ve encountered keep two phones or even three, one for each of the ride sharing services they participate in. One San Francisco driver drove a cab for a private taxi company. He also had one phone for Uber, one for its primary competitor, Lyft, one for cab-specific Uber clone Flywheel, and a fourth for receiving dispatch orders from the cab company itself.

His point: there are no barriers to entry here. Think regulatory protection, natural monopoly, high capital expenditures for entry, network effects or brand. Of all of these barriers Uber might have network effects and brand. But is this what is going to lead it to huge market share, growth, margins and staying power? Look, let’s call a spade a spade. Uber is a bubble stock. And it hasn’t even gone public. And the dynamics which have led us here are the optionality and volatility in a changing technology world that I spoke of earlier. It’s fantastic for innovation but it doesn’t mean the valuation isn’t crazy.

I believe we are in a bubble. Not all areas of the market are bubbles. The overall market is not a bubble. But excess has clearly increased in a way that will cause a massive loss for many investors. With Uber, at least there is no debt behind the mania. The real problem is when credit markets become frothy and then crash down to earth, with debtors defaulting en masse. We are not there yet, not even close. The macro outlook in the US, for example is very bright. But the signs of excess are building.

Links to follow later today

Comments are closed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More