Steve Keen Has Receipts

When the Stock Market Crashes and Trashes AI, Remember Keen’s War Against The Neoclassical Economists

Steve Keen isn’t one of those elitist economists.

Keen wanders the internet and will challenge anyone — even someone with one follower — the moment they claim government debt is going to destroy the country…probably by the end of the week.

Should the conversation turn to how banking works, and Keen’s interlocutor casually mention fractional banking, Keen will give them two choices. A:) learn the truth about private credit or B:) remain a stupid idiot.

If the man can be saved he will educate them. Free of charge. At no small cost to his own time.

I say this with admiration. Steve Keen is a crank. He’s also right. Unlike him, I treat my stupid idiots with humanity and compassion. When I make his arguments, I open by telling the other person I’m not perfect — I might be wrong about this or that. I try to give us equal standing. I’ve even told Steve, on ‘X’, that he should try it. Be humble.

Of course, I never get any further than he does because other people form their own opinions of me, without my permission.

Steve Keen spends a great deal of time arguing that banks do not operate according to the traditional “loanable funds” model.

In the loanable funds story, savers deposit money into a bank, and the bank lends that same money to borrowers. Lending is constrained by prior saving.

Keen argues this gets the mechanism backwards. Banks respond to credit demand (when Ponzi schemes become popular). That’s a nice way of saying they’ll give money to anyone if the bank down the street gets away with it.

When they make loans, they create deposits. Reserves are obtained afterward, as needed. In his view, lending is driven less by a pool of existing savings and more by the willingness of banks and borrowers to expand credit.

As they’re doing, in my opinion, by directly and indirectly financing SpaceX with OpenAI and Anthropic to come. These days the question isn’t will we get an Enron? It’s how many! (BTW Steve, many of us also saw the 2008 crash coming.)

Anyway, Keen’s distinction matters because it changes how you interpret a financial crisis. If banks are limited by savings, crashes should occur when lending resources run short.

If credit is created independent of reserves, crashes occur when speculative borrowing reaches the point where income no longer services the debt. If you’re thinking this sounds intuitive, it is.

Keen’s frustration is understandable. If his description of banking is correct, then an enormous number of people are operating with a mental model that doesn’t describe reality. That’s not economics. That’s a shared delusion.

Why do people accept neoclassical economic theories about banking? Because this can be “solved” mathematically:

Keen’s formula might be: Is the banker lending to AI companies and has a plane ticket to Costa Rica and a go-bag in his office drawer? (I’m making a joke here. Keen also uses copious amounts of data and math).

Why is it so difficult to convince people that economics isn’t a supply-demand mechanism run by Vulcans?

For years, I’ve been running a similar Don Quixote quest. My windmills are audio equipment manufacturers.

Many manufacturers sell 32-bit float recorders by claiming they capture sounds that would otherwise be lost — that the extra dynamic range saves performances that 24-bit recording would clip or ruin. I’ve argued for years that this framing confuses the storage format with the physical limitations of microphones, preamps, and analog circuitry.

Once your analog electronics have hit their ceiling, more digital storage space cannot recover information that was never captured in the first place. The bits have nothing to save.

The subject matter isn’t economics but the pattern is the same.

Keen is describing a mechanism inside the banking system. I’m describing a mechanism inside an audio recording system. And in both cases, we made the same mistake: we assumed people were interested in the mechanism. They’re not.

We both wake up believing today will be different. (Even though my arguments can be empirically tested, it doesn’t make a difference). There’s no science people can’t ignore.

Investors want to make money. They don’t particularly care whether money originates from loanable funds or credit creation. Content creators want great audio. They don’t particularly care whether a recorder’s advertised dynamic range reflects a genuine improvement in captured sound.

Their primary concern is practical success.

The problem is that learning a new mechanism often requires abandoning an old one. That means admitting that a previously trusted explanation was incomplete or wrong. Psychologists call this uncomfortable middle stage the “valley of despair” — the period between losing confidence in an old understanding and gaining confidence in a new one.

Many people choose to avoid that painful journey altogether.

Economists call this rational ignorance. If a belief is working well enough, and revising it offers no obvious practical benefit, there’s little incentive to look closer.

Suppose it were proven beyond all doubt that banks create credit exactly as Keen describes. Most investors would keep doing what they’re already doing. Their goal isn’t to understand banking — it’s to earn returns. Likewise, if it were proven that a 32-bit float recorder captures no additional information under real-world conditions, many creators would keep buying it. Their goal isn’t to master signal theory — it’s to feel certain they aren’t missing something.

No one can begin to invest if they’re fearful.

No one can create content.

The antidote to fear is confidence. And the fastest path to confidence is imitation.

People watch successful investors and copy their moves. They watch successful creators and copy their gear. The mechanism matters less than the reassurance that someone who made it took the same path first.

This is why arguments about mechanisms are so hard to win. You’re not fighting evidence against evidence. You’re fighting evidence against an existing feeling of certainty. And certainty, even when it rests on a flawed model, delivers confidence.

Put another way, you can’t get someone to examine a mechanism if doing so first requires them to admit they were wrong — to go back to being fearful.

Consider two friends who looked at Tesla ten years ago. The first follows the crowd and buys shares. The second, thinking more like Keen, holds back. He believes the banks are flooding the market with cheap credit, inflating a bubble that will eventually burst.

The first friend makes money for ten years. He sells everything yesterday. The market crashes tomorrow and Tesla falls to a fraction of what it was. Both friends were right — but one made money and one didn’t.

Most people understand the appeal of following the crowd. Most people understand, somewhere in the back of their mind, that following the crowd ends in tears.

What separates people isn’t whether they grasp both of those things. It’s how far into the valley of despair they’re willing to go, and how far into the future they’re willing to look.

How much any person wants to understand about mechanisms — financial, technical, or otherwise — is impossible to predict. Plato never left instructions for getting someone to climb out of the cave and into the light. Neither Steve Keen nor I can force someone to think step into the sunlight and leave their confidence behind.

Steve Keen has chosen understanding over comfort. So have I.

Though my commitment has limits. I believe understanding audio recording makes me better at it. But I also know I’m ignoring mistakes I’m probably making elsewhere. My marriage works. I suspect a good therapist would find plenty wrong with it. I don’t want to know.

I’m sympathetic to those who want to believe their 32-bit-float recorder puts them ahead.

Some happiness lives in knowledge. But there’s a part of all our lives — I believe — where ignorance is bliss.