London, 18 May 2011: Reuters

News that UK inflation hit 4.5 percent in April and is more than a percentage point over target for the 16th month running has prompted another exchange of letters between the governor of the Bank of England and Britain's finance minister.

The letters have become an empty ritual. But one assertion stands out and deserves to be challenged.

In his letter, Governor Mervyn King wrote: "Inflation is high at present because it is being pushed up by the rise in the standard rate of VAT (sales tax), higher energy prices and import prices. But unless continually repeated, the impetus from these factors should gradually diminish and, as it does, inflation is likely to moderate."

"The extent and timing of that decline are, however, uncertain. Over the past two weeks alone we have seen how quickly commodity prices can change. And there are both upside and downside risks to the inflation outlook".

King's view is that food, fuel and import prices cannot keep climbing continually and when the rate of increase slows, levels off or goes into reverse, recorded inflation will fall, hopefully back to target.

The Monetary Policy Committee's newest member, Goldman Sachs economist Ben Broadbent, made a similar assumption when he told the House of Commons Treasury Select Committee Tuesday it is likely oil prices will not rise as much next year as in the past year.

The idea that food and fuel prices must soon level off and lead to headline inflation converging back down to the more steady core rate is popular with prominent central bankers. Former Fed Vice-Chairman Don Kohn used this argument in February 2008. Current Fed Vice-Chairman Janet Yellen employed it in April 2011. New York Times columnist and Nobel Economics Laureate Paul Krugman made it in his blog last week.

But none of them has given a convincing explanation why they think oil and other prices are set to level off soon. Why now and not six months ago or in two years time? They are not entitled to make this assumption unless they can justify it. So far explanations have been thin or nonexistent.

DODGY FUTURES FORECASTS

Both the Bank of England and the Federal Reserve appear to take their assumptions about oil prices directly from the futures curve. In its May Inflation Report, the Bank wrote: "Energy prices are assumed to evolve broadly in line with the paths implied by futures markets over the forecast period".

Kohn and Yellen also cited futures prices for their predictions that commodity prices will soon level out. In her remarks Yellen argued: "The current configuration of quotes on futures contracts -- which can serve as a reasonable benchmark in gauging the outlook for commodity prices -- suggests that these prices will roughly stabilise near current levels or even decline in some cases".

This looks suspiciously like treating the shape of the futures curve as a forecast for spot prices over the months ahead, which is certainly not how the curve should be interpreted.

There is a nasty suspicion central banks have looked at the WTI curve, seen it only slopes up a bit over the next couple of years, and assumed the market forecasts prices are set to level off. There is an even more nasty suspicion the Fed has looked at the Brent curve, seen that it is in backwardation, and decided prices are likely to fall.

Unlike a bond yield curve, the structure of oil futures contracts will ALWAYS show prices are expected to flatten out close to current levels. The forward curve is not a reliable forecast. It is simply the price at which commodities for deferred delivery can be fixed today. Expected spot prices are only one component.

And even those expectations have been notoriously unreliable. As one trader commented on the Financial Times Alphaville blog: "The forward curve is what we bet against".

King implicitly acknowledged the problem with his reference to how much commodity prices had changed in just the last two weeks. While the central bankers are relying on the forward curve to give them some comfort prices are about to level off, the world's most prominent oil analysts are predicting they will rise further in H2 2011 and into 2012.

Hedge fund managers are also convinced prices will rise further. Funds and other money managers had a record net long position in WTI-linked crude futures and options prior to the first price crash on May 5. Even afterwards they still had long positions equal to 323 million barrels of oil (almost 4 days of global consumption), down from 363 million barrels the previous week.

These market participants are supposed to be some of the best informed and they have real money at stake. King, Broadbent and the others do not have to agree with them.

But if they are going to hang so much of their analysis on oil and food prices flattening off, they need to explain why they disagree.

HERBERT STEIN'S LAW

Broadbent's comments, and those of the others, look more like a hunch based on Stein's law. Herbert Stein was chief economic adviser to U.S. President Richard Nixon. His famous law states "If something can't go on forever, it will stop".

It is one of the simplest but most important truths in markets. Unfortunately, it does not say precisely when unsustainable price increases will stop and how far they will go in the meantime before snapping back.

It is certainly true that food and fuel prices cannot continue rising at recent rates indefinitely. Sooner or later escalating commodity prices must draw a response from central banks in the form of higher interest rates, push commodity importing economies into recession, or some combination of both.

Certainly, the rate of increase must slow, since a constant rate would imply bigger and bigger dollar jumps in prices. But that does not mean commodity prices must stabilise in the current quarter, or the next one, or even the one after that. Nor does it imply whether oil prices will stabilise at $90, $110, $130 or even $180 (all are figures that have been mentioned recently).

In previous Inflation Reports dated February 2011 and November 2010, the Bank of England also thought energy prices would flatten based on the path implied by futures curves, expectations that proved incorrect.

It is tempting for policymakers charged with producing predictions for something as complicated as the UK economy to take a short-cut and borrow a forecast for one component, energy prices, by incorporating the futures curve.

In which case the Bank must highlight how much uncertainty there is around this commodity price forecast and how much uncertainty it introduces into the inflation projections.

It is not obvious oil prices are any easier to predict than the UK economy. Certainly it is not obvious King and his colleagues at the Bank and the Fed should be relying on prices flattening out soon.

It is also far from clear why the Bank and the Fed think that oil and other commodity prices will level off or fall in the next 12-24 months given the global expansion will be maturing and demand will be rising steadily -- unless King and Yellen are aware of some new oilfields the rest of the world does not know about.

Broadbent hinted at this uncertainty when he admitted: "There are things the MPC should worry about, particularly with regards to commodity prices. It treats these as one-off prices ... and that has been the most important single forecasting error. In an environment where the most important parts of the world economy are consuming commodities, you might do things differently".

Indeed you might.

Ends --

By John Kemp, Reuters market analyst – for Commodities Now.

The views expressed here are his own.