Here’s my problem with the global financial crisis (GFC). In the autumn of 1996, I thought my friend Alex had lost his mind. Alex – not his real name – was and is an economist. He’s also one of the smartest people I know. And yet that autumn, he and his partner agreed to pay, for an 85-year-old weatherboard house on just 800 square metres of land in an untrendy inner-city suburb, the enormous sum of $400,000.

An exercise in failed prediction

My view then was that the Melbourne property market was approaching bubble territory, and that the bubble would probably pop sometime before 2000. I was worried that Alex had bid overconfidently and would end up with a home worth less than he had paid for it.

Alex the economist didn’t apply much economics to the decision. “This is what people are paying,” he shrugged, “and we want more space.”

Jump forward three years and prices had moved even higher. I saw rampant overconfidence and expected the housing price bubble would burst after the Sydney Olympics. My housing price predictions failed miserably again.

Jump forward another eight or nine years to the global financial crisis, and again I fully expected that Australian prices would go down and stay down as buyers lost confidence. Instead, they bid the market up yet again. Buying Alex’s home today would leave you little change from $2 million.

Is that crazy? I think so. But here’s the thing: you’d be mad to rely on me. I thought we were headed for bubble territory in 1996. On Australian housing, I have a solid 21-year record of bad price calls.

The prediction problem behind the GFC

So you see my problem with the GFC. We’re coming up to its tenth anniversary. It happened largely because borrowers and lenders did not realise that US housing prices had risen too high. Yet my experience with Alex’s house tells me one thing, and it’s this: I cannot reliably spot an asset price bubble. That’s despite my years of reading about the factors driving Australian and global housing prices.

And I’m not the only one who can’t pick a bubble. I run into people from time to time who say they can, but pretty much all of them lack convincing proof. Most seem to have selective memories about their own record; some have predicted disaster for years.

Economists, thankfully, are less likely to kid themselves about their bubble-picking abilities.

The profession does have a fairly compelling theory about how bubbles happen, courtesy especially of the US economist Hyman Minsky and his “financial instability hypothesis”. Minsky argued particularly that “periods of prolonged prosperity” tend to push people into overconfident judgments about how much they can make in the future and how much they can borrow to do it. Eventually that overconfidence gets out of control, the economy reaches a point where confidence suddenly evaporates, and we have a crisis.

How do you identify that point when confidence becomes dangerous overconfidence? It’s easy to see in hindsight and hard to spot at the time. It’s not really buried in the mechanisms of the economy, but in people’s heads. The great physicist Isaac Newton, having lost a fortune in the South Sea Bubble, declared that “I can calculate the movement of stars, but not the madness of men.” Minsky, a mathematician, never found a reliable formula for defining crazy overconfidence either.

Not for lack of effort...

So economics today has no good theory of when we have a bubble, or of when it will end, or of how to assess the size of the financial bets on the bubble, either.

And so we don’t have a strong recommendation on how to predict and cut short the next crisis. You can read about that in a feature elsewhere on this site. It’s not for lack of trying. Economists know this problem matters. They just can’t find a convincing way to solve it. Until they do, asset price bubbles seem destined to keep revealing the frailty of our predictive abilities. I fear the GFC will not be the last time they hurt us.