Benedict Evans at Andreessen Horrowitz published a slide deck yesterday outlining his analysis showing that there is no ‘Bubble’ in tech. I read through the whole thing several times, and I could not escape the feeling that I did not agree with his reasoning.
First, let me point out that this deck is an amazing piece of work. By my count, Evans combined 13 separate data sources. The fact that he was then able to pull new insights out of that is awe-inspiring for data nerds like myself. No Python script could replicate this. It is a work of True Craft.
So it took me a while to understand why I did not agree with it. Evans sent me a PDF copy of the deck, and I went through and tried to relabel the headlines on each slide to see if I could craft a different narrative with the same facts. That was fun, but then I realized I had a more fundamental issue.
Evans is answering a lot of questions with his slides, but the main question is Are we in a Tech Bubble? A refrain that he and his colleagues must hear every day. As I discussed in a post a few weeks ago, I certainly do not have an answer to that question. Some things seem very bubbly, some things not so much. But before we answer that question, we need to make some things clear. What do we mean by Bubble?
A good definition I hear in the public markets goes something like:
“A Bubble is an investment in which no reasonable argument can be made that the fundamentals of the asset could ever pay back the investment.”
So by that definition, Amazon with its nearly infinite PE multiple is not a Bubble, because I can make a reasonable argument that they could flip a switch on any of their businesses and become immensely profitable instantly. (In fact, I have made that argument in regards to AWS. And disclosure I own 100 shares of AMZN.) At the same time, the 17th century Dutch Tulip Mania was a Bubble because no one could reasonably argue that any tulip bulb would live long enough to pay back its price at the peak.
So today, the question “Are we in a Tech Bubble?” becomes “Would a reasonable person invest in a private tech company today?”, and perhaps more to the point “Would a reasonable person invest in a venture capital fund today?”.
And this is where I start to diverge with Evans. I am not saying he is wrong, only that on the subject of external validation of valuation I am not convinced by his data. Much of his work sits inside the closed loop of venture investing. I think his analysis of the division of returns among various funding stages is very fascinating. Great work, but it does not answer the broader question. Venture valuation methods are very different than public market valuations. Maybe I am biased because of my work history, but I am a big believer in the long-term efficiency of public markets. So to prove to me that we are not in a Bubble, I need to see some connection between private and public valuations.
Now Evans does touch on this in his deck, looking at Tech investing in relation to the US GDP and to public market metrics. In particular, he compares PE multiples of the S&P IT Index to 1990’s levels, and indeed we are well below those. I think there are two things wrong with this approach. First, a quibble. That particular index is market-cap weighted, which means that Apple is a huge component, and Apple’s valuation is just outright confusing to everyone. I do not want to go too far down this rabbit hole, but put simply, tech stock multiples are higher than it appears in the data. Second, and more important, all of Evans’ data shows that current conditions are much lower than the late 1990’s peak, BUT current levels are in second place to that numbers. We are nowhere near the extremes of the last Bubble, but that does not mean we are not in a Bubble. It could just be that we are in a Bubble with a different shape.
When it comes right down to it, I actually agree with Evans’ basic point. Conditions are different now, companies are more profitable, venture investments are much less risky. But I wanted to write this after I saw a lot of people on Twitter and the blogosphere collectively breathe a sigh of relief on reading Evans’ analysis. As much as current conditions look good, that does not mean we should lower our guard. Evans proves that we do not need to become Prophets of Gloom today, but I think it is important to remember that conditions change very quickly. And they will change.