The bust that Swan didn't see

The government has completely missed the target on iron ore price predictions. Had it considered the irrationality of producer behaviour, it might have known what was coming.

As recently as May 2011, the Australian Budget Statement forecast only a gradual decline in iron ore prices. "The terms of trade are forecast to increase 19¼ per cent in 2010-11, underpinned by strong increases in the prices of Australia’s key non-rural commodity exports, before declining gradually over 2011-12 and 2012-13 as increasing global commodity supply starts to match growth in demand,” it said.

As is not news to anyone today, the decline in Australia’s terms of trade in the first three months of financial year 2012-13 has been anything but gradual: it has fallen five times as much in three months as Treasury expected to happen over the entire year.

Figure 1: Australia’s terms of trade since 1960


As Mark Twain allegedly once remarked, "prediction is pretty difficult, especially when it involves the future”, and it’s not surprising that the budget forecast hasn’t come to be. But what is surprising is that those making this forecast didn’t even consider one of the classic models of commodity prices – the Cobweb Cycle.

This model was first developed to explain "the Hog Cycle” – the volatile cyclical pattern of prices for pigs in the US, which would surge one year, only to plunge the next, and return to boom levels again about every four years.

Figure 2: US live hog prices 1990-2000


The cycle appeared to result from out-of-phase changes in the number of pigs being slaughtered each year (production), the price of pigs (price), and the number being bred for sale (inventory) moving out of phase with each other. Excess production on year would lead to falling prices and less breeding of pigs, which resulted in less stock available, and a rising price, followed by more production and then a falling price once more.

Figure 3: The out of phase dynamics behind the cycle


Mordecai Ezekiel, the first economist to attempt to model this pattern, argued that it could be generated by a stylised "supply and demand” graph, with price starting out of equilibrium. If there was a very low supply to the market one year, then demand – and price – would be very high. This would encourage producers to supply much more output the next year, which would result in a glut that depressed prices.

Production based on those depressed prices would be low the next year, leading to high prices; rinse and repeat. The time path of prices, when drawn on the supply and demand diagram, looked like a spider’s web – hence "The Cobweb Cycle”.

The basic thesis is that supply didn’t "start to match growth in demand”, but overshoots it. Then demand reacts to the oversupply, causing price to plunge – and producers react to the fall in price, cutting supply, which then leads to price rising once more. Price thus cycles up and down, and the best prediction of future price is that, if it’s high now, it will be lower in the future.

To twist a genuine Twainism, though history doesn’t rhyme, the iron ore price sure looks like a hog cycle to me. Price peaked before the GFC, then plunged because of it; China’s massive stimulus caused demand for iron ore price to skyrocket; pushing prices to a high of $US177 a year ago; then every woman and her dog dug holes in Western Australia and elsewhere – and the price has plunged to $128 last month and below $90 last week. This utterly predictable behaviour implies that the high terms of trade we had experienced courtesy of China’s stimulus program couldn’t be expected to last.

Figure 5: The Iron Ore Price


So why did the government’s economic advisors NOT consider this model when advising Wayne Swan about where iron ore prices – and our commodity-dominated terms of trade – might head in the future? Possibly because, like so many other sensible models in economics, the "Cobweb Model” fell afoul of the "Rational Expectations Revolution” in economics.

The seemingly sensible proposition that "people learn from their mistakes” was used to argue that the Cobweb Model wasn’t reliable, because it assumed that people don’t learn from their mistakes. Rational hog-farmers, it was asserted, could work out what the equilibrium level of supply should be, given current information about supply and demand, because they would have a model of the market in their heads. They would base their expectations on that model, rather than just reacting blindly to price movements, and their expectations would be accurate.

As American economist John Muth, a father of rationalal economic modelling, put it: "I should like to suggest that expectations, since they are informed predictions of future events, are essentially the same as the predictions of the relevant economic theory."

When rational expectations were introduced into the model, there was no "Hog Cycle”, just a series of random movements of the price around the equilibrium level.

That, of course, is hogwash – as is the very notion of "rational expectations”. Expectations are not "informed predictions of future events”, because you can’t have "information about the future”. Instead, as John Maynard Keynes argued in the last Depression, the future is subject not to risk but uncertainty, and "it would be foolish, in forming our expectations, to attach great weight to matters which are very uncertain.”

Keynes continued that: "It is reasonable, therefore, to be guided to a considerable degree by the facts about which we feel somewhat confident, even though they may be less decisively relevant to the issue than other facts about which our knowledge is vague and scanty.

"For this reason the facts of the existing situation enter, in a sense disproportionately, into the formation of our long-term expectations; our usual practice being to take the existing situation and to project it into the future..."

This, of course, is precisely what Messieurs et Mesdames Forrest and Rinehart et al did: they saw sky high iron ore prices, euphorically believed they would continue, and geared up their expansion plans – in Forrest’s case, literally – in a race to have their hole in the ground ready before all the others. This is the other reason why hog-like cycles exist in commodity prices: every producer plans to be the first to get their product to market, and in that race the devil – or the banks – take the hindmost.

Now of course there are holes not just in the ground in Western Australia, but in the Australian budget as well, as tax revenues that were expected from mining super-profits are now unlikely to eventuate. Treasurer Wayne Swan will surely get sick of having to spin that this sudden slump in iron ore prices couldn’t have been predicted. But that too is hogwash – any little piggy could have told him so.

Steve Keen is a professor of economics and finance at the University of Western Sydney and author of Debunking Economics and the blog Debtwatch.