Oh Lordy. Another "paradigm shift".

It is a term that, like "new economic reality" or even "global consensus", tends to signal that any accompanying thesis deserves handling with industrial tongs. Remember the guff that justified the dot.com boom?

This time, it is Guy Allen, a Louis Dreyfus China grain boss, who is talking about "paradigm change".

Mr Allen has said the commodities market has taken one of the steps up it does every 30 years or so.  So that average prices until the mid-2030s are likely to be far higher than those in the previous period, beginning in the early 1970s.

The relief for farmers and long-only investors over the thesis is that it may be better for them if it does not completely hold water.

Step by step

Sure, the thesis appears to hold plenty of good news for prices.

Look at corn prices.  These took a step up just before World War I from vibrating around $0.45 a bushel to oscillating near $0.75 a bushel.

A spike towards the end of World War II raised this focal point to a $1.33 a bushel, and it rose to nearly double that in the early 1970s, when Russia stockpiling and inflation supported prices.

If you believe, as Mr Allen does, that history is a guide to the future - that every so often "a change in thought and a change in mindset" alters trading patterns - then extrapolating the graph would appear to point to corn oscillating somewhere around $3.50 a bushel.

Peak before trough 

The bad news is that this is lower than today's price. Indeed, the blow for bulls about Mr Allen's steps is that, unlike real ones, their highest points tend to be at the leading edge.

Take the step beginning in WWII. US corn prices hit $2.83 a bushel in 1948, according to Commodity Research Bureau data, never to return for a quarter of a century. Indeed, they fell back nearly to $1 a bushel in 1961.

A comparison for wheat with the 1970s sends a similarly disappointing signal.

The Chicago wheat price graph for 2005-09 bears a striking similarity to that of 1971-76, as Mr Allen shows. But if prices behave now as they did 1976-78, they can be expected to fall to about 35% of peak value, less than $4.50 a bushel, over the next couple of years.

'Very, very tight' 

In fact, it may be investors preferring to look at fundamentals whose hearts Mr Allen warmed most.

In separate comments, he pegged global grain at the equivalent of about 18% of use, signalling, as he said, a "very, very tight" market. It is roughly half the percentage reached in 1986, a grim year for prices.

And Louis Dreyfus isn't forecasting the market getting any looser until after 2012.

Fundamental link

That may be a better reason to invest.

Historical correlations are not always accurate indicators to the future. In equity markets, buoyant prices are testing Robert Schiller's observations on the link between long-term price-earnings ratios and company valuations.

But the need to eat, and the availability of food, is not a theoretical dynamic.