With the NASDAQ going above 5,000 for the first time since the year 2000 valuations in tech are once again on everyone’s mind, mine included. It has been a long period writing here on Continuations that I have thought valuations were too high and they have only gone higher since. All the arguments I provided back in 2012 are still true and in particular the line that “rates of return available on many other investments are at historic lows” — this is especially true for interest rates. Very low interest rates provide double fuel for stock prices. First, because investment dollars migrate from fixed income to equities. Second, and more importantly, because discount rates are low. If you have ever built a DCF model you know just how insanely sensitive the valuation is to the interest rate. And with technological deflation it is actually a reasonable expectation that interest rates will stay super low.
Yesterday, Mark Cuban wrote a click-bait titled rant on valuations which I read after it had popped up multiple times in my Twitter stream. While I didn’t think it was entirely coherent it did give me an interesting idea though for a weird interaction between public and private markets that I had not previously considered. During the original Internet Bubble, public market valuations relatively quickly sky rocketed in an all out frenzy. Private valuations didn’t have that much time to follow because a lot of companies went public quickly and then the bubble burst.
This time round things have been growing much more gradually over time and the supply of public companies has been small compared to the amount or private wealth creation as companies have been slow to go public or have avoided it altogether (see my related posts on still waiting for IPO 2.0). So now we have a different phenomenon: demand in the public markets outstrips supply which results in well higher prices. But then in turn it is those high public market multiples that inform private market valuations. And voila you have a case of MC Escher‘s famous picture of hands drawing each other (dear Internet: someone please put “public” and “private” on the hands and have them write 10x each; addendum: gorgeous illustration provided by Alec Hutson).
Because it is happening gradually and because the logic looks internally consistent (and add to that the low interest rates), this could continue to go on for quite some time. This strikes me as the classic case of Nassim Taleb‘s point about fat tailed distributions where it is the higher order moments (kurtosis) that really matter. So the process looks very smooth and gradual for quite some until there is a sudden and fairly violent swing