minsky and the double tap

are we having fear yet? part two

This was on my mind to write earlier this week, but Poppy’s text was so good I couldn’t delay it. I also had half a mind to write it up in a more rigorous format, but the kind of data I would need to substantiate this thesis doesn’t really exist. So it’s here, in the form of “structured speculation”, which (like “narrative probability” I will defend as a valid statistical technique).

I thought I’d update my post “Are we having fear yet?” from six months ago, in which I admitted that I was worrying a little bit about a crisis in private credit markets. In doing so, I find I am less worried about that. But I also found myself worrying about that in some sense, a non-crisis can be just as dangerous as a crisis. The idea here is that Minsky’s Financial Instability Hypothesis isn’t easily tamed. (If you’re not up on Minsky, just think of it as a more rigorously argued version of the joke “every financial crisis happens exactly when the last person who remembers the previous one has died or retired”, and you won’t go far wrong.

Let us fire up the “valve amplifier of history”, and look at the dot com crash of the 2000s. Famously, this did surprisingly little harm to the real economy. Although there was a lot of destruction of value, it was mostly unrealised equity gains[1], and the dot com stocks themselves were funded with VC and IPO money rather than debt. So there was no widespread cascade of bankruptcy, no credit crunch as banks were forced to call in loans, and we were all off to the races again within a year or so.

Except … that’s a slightly misleading and self serving history. The reason that the dot com bust had so few consequences is that there was a massive policy response to it. It shaped Federal Reserve policy for the decade to come, with one of the consequences being the real estate bubble which ended very very badly in 2007/8. One of the very first things I posted on the internet, was this piece, demonstrating that the future course was by no means unforeseeable from the perspective of 2002:

“…On the other hand, if you had been placing bets on a US double-dip recession so far, you’d have lost them, because Alan Greenspan and his merry gang at the Fed have a solution to this problem. Basically, the solution’s pretty simple and it involves screwing interest rates down to the floor until mortgage rates follow them down to Low Low Prices levels, and pointing out to the Great American Consumer that it’s “Bye-Bye, Magic Stock Market Bubble Money!” but “Hello, Magic Housing Market Bubble Money!”. Marvellous.

Cleverer readers at this point will be formulating an objection. The objection goes along the lines of:

“Yeah, yeah, laughing boy, but what happens when the housing bubble bursts then?”

Which is a damn good question to ask, particularly since the official policy of the Federal Reserve appears to be “hmmm yeh, never thought of that, I suppose we’d be kind of f**ked”.

The roots of the Great Financial Crisis were in a triple coincidence of three ways in which the Minsky cycle can operate:

1. An inappropriately large and sustained policy stimulus, fuelled by the perception that it is not risky to do so because of a large recent negative shock.

2. A widely shared perception of invulnerability and positive animal spirits, fuelled by the perception that the last widely predicted disaster turned out not to be so bad (possibly because of 1)

3. A deregulatory financial cycle (possibly driven by 2).

In retrospect, the COVID stimulus doesn’t look like such a great bit of policymaking as it seemed at the time, although it still has defenders (including, at various times, me). But, unless you’re trying to sell a flat in London, or a Quantitative Easing portfolio, you might consider that it hasn’t had much consequence in the shape of financial instability. So you might call this a “small, Type 1 Minsky shock”.

The fact that the COVID episode didn’t lead to mass bankruptcy is also, to my mind, a small Type 2 Minsky shock. I read a lot of the output of financial regulators for my day job, and my perception is that they are fighting a losing battle to keep a big asterisk on the track record of loan losses, and to remind bankers that they owe much less to their own underwriting skill than they do to a huge injection of fiscal bankruptcy-proofing and to the fact that inflation has significantly eroded the principal value of debt. As with the Silicon Valley Bank crisis (another small Type 2), I am by no means arguing against bailouts . But it is really important not to be coy when you use taxpayers’ money to bail out the financial system, because they are little blighters for forgetting that you did so.

One reason to always be very up front about bailouts is that doing so gives you political clout with respect to regulating the industry. (Ask UBS if you don’t believe me). So far, we have not had any material Minsky Type 3 shocks since the global crisis, but it looks pretty obvious that one is on its way; the Fed appears to be leading the way.

All of which is to say – I am not currently particularly worried by the private-credit-data-center-AI financing nexus. It all seems to be financed by people who can afford to lose the money, and in a way in which the linkages to the banking system (and therefore to the supply of credit to the real economy) are not so bad as to create the potential for a 2008-style freeze. Like Chinese real estate, the very fact that it would happen in a post-Lehman world makes a Lehman episode less likely.

But … just saying that, it’s very much rhyming with the dot com era, isn’t it? The scenario to worry about might not be that we mishandle the private-credit-data-center-AI bubble. It might be that we handle it in a way that looks really well executed, with a big public stimulus, and that we learn from this that the current global architecture works really well, that the financial system is making good decisions again and that it’s no big risk to continue to cast a critical eye over the more burdensome parts of the regulatory system.

It’s always the double tap that gets you. (Envoi: perhaps this means that future historians will look back and say that the best thing that the US Treasury did in the late 2020s was to provide the global economy with the invaluable service of keeping us worried).

---

[1] One might think of this in terms of the difference between the amount that the victims of Bernard Madoff actually lost (money they gave him, less money they got back) and the amount that they thought they had lost, based on their fraudulent statements.