You are here:

Supercycle doubts

Created:
9 May 2008
Written by:
Chris Dillow

If you want to annoy an economist - as if you would - a good way is to talk about a "supercycle". Talk of cycles is bad enough - complex, irregular behaviour cannot be reduced to a neat predictable periodicity. Adding the prefix "super" just compounds this with pretentious big-think.

Advertising

And such fancy talk might be unnecessary. Much of the rise in commodity prices in recent years has a more prosaic explanation - the weak dollar. This has caused central banks in those middle eastern and Asian economies which peg their currency to the greenback to print money to support the dollar. And some of this cash leaks into higher demand for commodities.

Since January 1995 the correlation between annual changes in the dollar's trade-weighted index and moves in commodity prices in the following 12 months has been 0.56. This alone explains almost one-third of the variation in commodity price inflation. These are impressive numbers considering the volatility of the data and the fact that so few people talk about this relationship.

However, although the dollar's fall in the year to May last year successfully predicted that commodity prices would rise in the 12 months since then, the actual rise we've had has been two standard deviations greater than that predicted by the post-1995 relationship: 53.9 per cent against a predicted 15.5 per cent*.

Is this discrepancy evidence of a supercycle? Possibly - but if it is, it only began in the last few weeks. It could also be a sign of (temporary?) supply disruptions, a speculative bubble, or just noise. It's too soon to say.

However, we could soon have a better test of our competing hypotheses. The dollar has strengthened recently. If this continues - a big if, of course - I'd expect commodity prices to weaken next year.

If they don't, then the supercycle hypothesis will gain credence. But until then, I'll be sceptical.

* The equation is: 12M change in S&P/GSCI index = 10.88 - 1.78 x annual change in $, lagged 12M. The standard error is 17.9 percentage points. The intercept implies that commodity prices rise by almost 11 per cent a year if the dollar is flat. Whether this is evidence of a "supercycle" or just a normal Ricardian trend is a matter of taste.

MORE DILLOW WISDOM...

Read more of Chris Dillow's unique comment articles here.

Or see his blog at http://stumblingandmumbling.typepad.com.

  • Back to top

GLOBAL BANKING CRISIS

Get the latest news here!

Promotional Feature

The world's oldest pets

We all want our pets to live long and happy lives. One of the easiest ways to help your four-legged friend get over life's many hurdles is to get them covered with pet insurance.

by Sainsbury's

Promotional Feature

Turkey Resort Round-up – Beyond Istanbul

Holidays to Turkey tend to focus on the country's magnificent coastline. Dotted with coves and lined with sandy beaches, Turkey is a sun-worshipper's paradise.

by TUI