All-In Summit: Bill Gurley & Brad Gerstner on markets, downturns & investment cycles

The session is introduced as the “BG Squared” panel, with Brad and Bill as the featured guests and Sacks joked about as a guest of theirs. After some banter, a host puts the main question to Bill: markets are again sounding an alarm, and Bill has lived through several downturns, so how does this one compare with the great recession, the .com bust and 1987? The “predicted 5 of the last three recessions” line is a joke, answered with “a broken clock is still right twice a day.”

The speaker answering the question recalls a meeting with Howard Marks, described as a famous bond investor who writes a lot. After asking structural questions about venture capital, Marks called it a “really shitty industry” because cyclical collapse is built in: funds have 10 to 15 year lives, so barriers to entry are low while barriers to exit are very high. The speaker took from this that the cycle is not a smooth sine curve but a sawtooth, in which risk-on builds slowly and reflexively and risk-off arrives abruptly.

The speaker places the latest risk-on phase as running from 2009 until about five months earlier, followed by a risk-off phase that has lasted five months. The practical problem is that companies must adjust quickly to a new reality: cap tables may be stuck with too much liquidation preference, valuations have shifted and the cost of capital is very different. On the way up, people took more risk and made investments they would not have made when capital cost more.

A joke about a company naming a stadium for five years as a crypto company illustrates the point: a commitment that looked fine in the boom becomes larger than the market once prices fall. A listener assumes this refers to FTX, and the speaker uses it only as an example of a disproportionate commitment.

Asked directly about comparisons, the speaker says 2009 was not that bad, since things turned around fairly quickly, and that having another 2009 would be fine. The earlier .com bust was far more abrupt, and in the speaker’s account liquidity did not really return until roughly the middle of the 2000s, apart from exceptions such as the PayPal example another speaker raises. A fellow speaker calls it a long walk in the desert in which many great companies were started but many founders gave up.

The speaker adds that early-stage investors and founders who have not yet scaled may be little affected, and may even benefit, because access to talent gets easier and people become more pragmatic and rational. The catch is that the window on the far side is usually long. A further difference is 2020: a mini pullback in March was followed by such a forceful Fed response that markets took off again, and, as the hosts have discussed, that tool is no longer in the toolbox.

Brad frames his section around facts that could help founders judge the next six months, such as whether they can raise capital or whether markets bounce straight back. His starting claim is the “iron law of investing” that interest rates matter most: a 1% change in rates leads to a 15 or 20% change in a multiple, so multiples collapsed because expectations for inflation and rates moved sharply. He is responding to talk of a repeat of 1999 and 2000.

He plots rates from 2019 to 2022, noting they were taken to zero and that the Fed now describes a neutral rate of 2 to 3%, which has people hyperventilating. In his comparison, the cost of capital in 2000 was “crazy”, with rates rising from just above 5 to 6.5. The open question is whether rates return to 2.5 or 3, or whether the Fed is behind the curve and must go to 4 or 5 to kill inflation.