Chamath Palihapitiya: it could take three years for the market to “accurately” reprice late-stage cos
Former Facebook exec turned VC Chamath Palihapitiya has long been a controversial figure in the investing world. Both brilliant and combative, Palihapitiya came to be known most widely by ushering in the era of special purpose acquisition companies, or SPACs, beginning in the fall of 2019, when he helped Virgin Galactic become a publicly traded company through a SPAC he formed.
Palihapitiya went on to take five more companies public via SPACs before the boom ended abruptly last year, and while investors who followed him into some of his SPACs lost money — as did investors in many hundreds of others SPACs that materialized in 2020, 2021, and last year — Palihapitiya reportedly doubled the roughly $750 million he invested.
Many blame him for aggressively promoting his own interests — including during numerous CNBC appearances — at the expense of less sophisticated investors. Others continue to heed his investing advice, considering that Palihapitiya seems adept at identifying investing opportunities early. (This editor recalls his appearance in 2014 at a packed bitcoin conference in San Francisco where he argued that everyone should have 1% of their assets in Bitcoin. At the time, each Bitcoin was valued at $520.)
Both camps might be interested in a recent appearance by Palihapitiya at an investing conference in Miami where he said he thinks up to seven years of high interest rates would be good for the venture industry, that America’s deteriorating relationship with China is a boon for the country, and where he talked about generative AI and where he thinks the real money will be made. It’s worth reading if you have a few minutes. His comments have been condensed for length and clarity. You can check out the full interview here.
On the impact of higher interest rates:
There are a couple of dirty little secrets [in the venture industry]. One of them is that only 10% of all of the firms in our asset class actually generate real returns, which means 90% are basically floundering around, burning money. The other thing is that we have always consistently generated a high-single-digit DPI [a term used to measure the capital a fund has returned thus far to its investors] — 1.7x is like the 30-year average — yet we are the worst offender when it comes to showing [the institutional investors who fund VCs] paper markups or TVPI [which represents both realized profits and unrealized future profits]. So there is this dance that that this industry has been able to play because rates have been at zero [but] as investors, the asset class is challenged in order to generate real returns [because the companies we have funded] have, as a result of all this excess capital, been more poorly run than otherwise, so we need to course correct. We need these rates to be sustained for 5, 6, 7 years, frankly, hopefully, in order to really flush it through the system.
On what Palihapitiya thinks of now of crypto, SPACs, and other innovations that investors poured into between 2018 to more recently:
Early-stage venture, largely in healthcare and software and deep tech and recently energy transition — that’s been our bread and butter. But we sometimes go a little off-piste. In early 2011, I went off-piste and made a huge bet in Bitcoin when it was $80 a coin. It just seemed like just an unbelievably massive risk reward. We did the same thing in the mid 2000s; we did it in SaaS and in deep tech.[With] SPACs, we stumbled into this thing because we wanted to raise money for a bunch of our companies that were extremely capital intensive, and we demonstrated something that, in a moment, just caught a lot of wind. We did six of them. I think there were 650 of them just in 2021, so we [represented about] 1% of the market. I think we bought good companies; I think we sold well, quite honestly. But it’s one of these things where it was fueled by a moment in time of just enormous excess liquidity. And now I think we’re sort of back to basics. So for us as an institution, we’re kind of back to early-stage venture. . . . I’m the largest LP my fund, so when there’s a window, I go for it. And that was a moment where we tried to go forward.
On what he makes of the later-stage market right now:
At the end of last year, I looked at six or seven [convertible notes]. These were all extremely well-known companies that all of you would know on a first name basis, and they all came to me trying to raise converts. And I said, ‘Well, here’s the real market clearing price of these companies,’ and none of them took my money. And instead, they did a convert to basically deflect and kick the can down the road on valuation.
So we’re at that point in the market where all the boards of these private companies refuse to budge on valuation. And the reason is because it impacts meaningfully their DPI or their TVPIs that they’ve given to LPs. And so it’s a very difficult part of the private markets right now to invest in, because you will not be allowed to do true price discovery because nobody wants to take the real hits.
The best companies will do it. You saw Stripe [do it]. That’s probably the best technology company in Silicon Valley proper being built right now. Klarna did it. The through line there is Sequoia, which is an extremely disciplined and incredible organization; they’re able to enforce that discipline. But other companies, other venture funds, they don’t want to look at the TVPI decay. . . . [so] we’re going to continue to sort of over-index into early stage and do as many good deals as we can see, and let the chips fall where they may.
. . . . Because this is now just money bad. And when [more venture] folks leave the market, those companies now become more prone to get repriced accurately . . . [But] I honestly think that’s like three years away. I thought it was going to be three quarters away. At first, when we were thinking about, like, how much capital are we really going to be allocating over this next period, we cut it by two thirds, because we just didn’t see the opportunities in the late stage anymore.
How he thinks about America’s deteriorating relationship with China and the accompanying technology bans being advanced:
It’s an unbelievable boon for America. And it’s unbelievable boon for America’s technology sector. The thing is, when you look inside of China, they are extremely good at process engineering. They’re also extremely good at additive manufacturing. They’re extremely great at things like speciality chemicals, but all of those things when you think about the precursors come from American, European and Australian companies that now have a huge incentive to diversify that supply chain away from China, [and that] benefits American companies in a massive way.
China’s response is [so far] muted. So for example, we said, ‘We are going to slow down the flow of extremely advanced semiconductor manufacturing equipment into China’ and China’s response was, ‘We are not going to allow you guys to get the input components to certain silicon wafers that are used in PV cells.’ I mean, if you had to rank these things, no offense, but we can make solar cells. The equipment that you need to get to two-nanometer scale in chip design comes from the Dutch, the Germans, and the Americans.
On ChatGPT and the generative AI craze more broadly:
What ChatGPT shows you is just the amazing value in allowing computers to assist you in doing work. It’s like a calculator replacing the abacus replacing a pen and paper. [But a] friend of mine told me this [Warren Buffett quote] yesterday, which I love. He told the story about refrigeration and the story he tells is that the people and the person who invented refrigeration made some money. But most of the money was made by Coca Cola, which used refrigeration to build an empire. And I view these large language models as refrigeration. Will there be some money made in it? I think so. But the “Coca Cola” has yet to be built. And those are the companies that are really going to monetize it.
Here’s a basic thing about machine learning that’s worth knowing: if you take 1,000 of the same inputs and give it to Facebook and Microsoft and Google and Amazon, they’ll all come up with the same machine learning model. But if you have one extra thing, one little ingredient that all of those other companies don’t have, your output can be markedly different. It’s like giving two great chefs three ingredients; when you give a third chef one extra one, that person has the ability to do something very special. So right now we’re in the world where everybody is crawling the open web, [but] we’re going to move to a world where as everybody gets sophisticated enough, where when refrigeration is widely available, someone is going to say, ‘ You know what? This site? I’m not going to allow anybody else to access it; it’s only me only for my models.’ And those models will become better. We have to let that play out a little bit.