Keyboard worrier

Wednesday, May 16, 2007

The Kondratieff Cycle


Some investment analysts are "chartists" - they try to predict the future short-term movement of the markets, using patterns they think they've seen in the past. There are longer-term patterns too: we commonly talk of a "business cycle" of say 8 or 10 years.

Could there be really long cycles? Nicolai Kondratieff (or Kondratiev) (see Wikipedia article) thought so. His wave takes around 50 years and predicts decades of booms and depressions. His theory still excites professional investors today - see this article about Marc Faber, and Shane Oliver at AMP.

Of course, the question is how exactly to fit the pattern to our present situation. There is a nice graphic presentation here, showing past data and extrapolating to 2010. But look at other sites, too, like this one from 1998 - here the analysis suggests we have already hit the bottom.

Maybe the answer is that such patterns do exist, but the turning points are impossible to predict, so chartists stretch the waves. For example, you'll see in the second chart above (about technology, related to Kondratieff), that the first 3 cycles are set at 50 years, and the fourth at 40.
Sometimes an unexpected dramatic event starts the change, e.g the murder of Archduke Franz Ferdinand in 1914. And British history would have been very different if Guy Fawkes' 1605 plot to blow up King and Parliament hadn't been leaked. So you can't get the timing perfect.

But you can prepare. The two books recently reviewed on this blog explain why we should worry about the state of the US economy (and the UK has similar problems, maybe on a different scale). You don't need to know when the fire will start, as long as you've planned how to leave the building in an emergency.

No comments: