Stephen King: Not wanting to get it wrong, economists fail to get it right

"When in doubt, trust the consensus." Is this a wise recommendation? I suspect not. If you take the consensus numbers calculated by Consensus Economics - which are based on the forecasts put together by members of the economics community - there are some fairly shocking errors that are repeated year after year.

Did you know, for example, that US economic growth has been stronger than consensus in six out of the past seven years? Did you know that German economic growth has been weaker than consensus in five out of the past seven years? And did you know that upside surprises to Japanese growth in any one year are invariably followed by downward surprises in the following year?

The puzzling thing about all of this is that economists like to believe that people are, for the most part, rational - and that means they won't make the same mistake year after year. In many walks of life, this is true. If you have a rotten holiday, you're unlikely to head back to the same place the following year. For economists, though, it seems not to be true. We, collectively, appear to believe that the US simply cannot keep delivering the growth rates that it has achieved over the past few years (and, to be honest, I'm as guilty of this belief as anyone else). We also cannot believe that the German economy could do quite so badly on such a consistent basis.

Why is this? Perhaps the most obvious explanation is that most forecasts are made with some prior knowledge of what the consensus is. And, because no economist wants to feel too uncomfortable, there is very little desire to be too distant from consensus.

The consensus is bit like a campfire on a chilly night: you don't want to stray too far away because, far from the warming embers, it's very cold, very dark and sometimes very scary. The problem, though, is that while economists are happily sitting together, singing a few songs about the delights of inflation targeting and eating from their billycans, there are all sorts of unpleasant bogeymen hiding out in the bushes, waiting to grab these latter day soothsayers, dragging them off to totally unexpected worlds.

It's really not difficult to prove the point. Consensus Economics publishes not just the "average" forecast for economic growth - and a host of other variables - for the year or two ahead. It also publishes the highest and lowest forecasts from within the range of numbers that help form the consensus. In the US, economic growth has been outside this range in five out of the past seven years. For Germany, it's been three years out of seven. In Japan, it's four years out of seven. In other words, more often than not, there is not a single forecaster who is either sufficiently optimistic or sufficiently pessimistic about a country's economic prospects in the year ahead.

It makes sense, then, to think about the economic bogeymen that could come along to upset the apple cart, to think about some of the major risks that might afflict the consensus in the year ahead. Before doing so, though, it's worth restating what the consensus actually is.

For the coming year, think warm campfire. Collectively, economists believe that growth around the world will be decent, but not excessive. They believe that inflation will drift higher. They believe, on the whole, that central banks will be raising interest rates. And they believe that currencies will be reasonably stable: no repeat, then, of the dollar sell-off through 2003. As soon as I look at this view of the world, the alarm bells start to ring. Why? Because this story is very similar to the consensus view a year ago. To be fair, the consensus at the beginning of 2003 did very well on both growth and inflation, but got the interest rate and currency implications completely wrong. Against expectations, short-term interest rates fell in both the US and the eurozone. Against expectations, the euro experienced a stratospheric rise against a dollar that, for one reason and another, was no longer able to defy gravity.

So what are the key risks for the year ahead? I haven't got the space to go through all of my "top 10" in detail but you can see them listed in the table. It will be important to keep one eye on Asia, particularly Japan and China. These countries have been happily intervening in the foreign exchange markets, buying up US assets, in an attempt to prevent their currencies rising against the dollar. This kind of policy is fine as long is growth is weak and inflation is low, but rapid increases in foreign exchange reserves tend to boost domestic liquidity, so growth and inflation might not remain low forever. Should either growth or inflation surprise on the upside, central banks in Asia might become more relaxed about currency appreciation, thereby sending the dollar into freefall. And, with shades of 1987, I wouldn't want to be a holder of US equities in this situation.

Closer to home, policy makers will be grappling with the implications and consequences of inflation targeting. Economists have a lot of faith in policy makers but I reckon that policy makers might sometimes think that this faith is misplaced. In the US, the need to keep growth going is a political imperative, but most of the available "shots in the arm" have been used up. What happens if the labour market remains weak? A further sharp depreciation of the dollar, perhaps, together with a stalled recovery.

In the UK, the Bank of England has just started to embark on a policy of weaning consumers off debt. There are two obvious risks: either the weaning is harder work than expected - in which case, interest rates will have to rise a lot further than the consensus suggests - or, alternatively, consumers capitulate more quickly and violently than expected, in which case, the Bank is forced to cut rates to head off another threatened recession.

Oil prices are a potentially thorny issue. They're a lot higher than most of us expected a year ago. What happened to the supposed "peace dividend"? Do higher oil prices present a threat to inflation, as the consensus appears to be implying? Competitive markets suggest not. If so, the biggest risk must be to output: higher oil prices leave western companies and consumers worse off and, with pricing power a long-forgotten ability, higher oil prices must threaten the pace of economic recovery.

Then there's the central banker's puzzle. It's quite plausible to suggest that growth will surprise on the upside in countries where productivity growth is strong - the US, for example - but that inflation will surprise on the downside. After all, this has been a feature of many economies for a number of years now. What should central bankers do? Raise rates, because there is a cyclical recovery in economic activity? Or cut rates, because inflation threatens to turn into deflation? An academic debate, perhaps, but one that now appears to be raging in the Federal Reserve. Suddenly, we will be faced with central banks that no longer are able to offer the transparency and supposed certainty of the recent past.

Overall, I reckon the recovery in 2004 will not be as strong as the consensus would like. And regardless of the pace of recovery, I suspect that heightened levels of competition, combined with the impact of offshoring on labour costs, will push inflation down to levels not a million miles away from outright deflation. The US will pursue dollar depreciation still further, eventually prompting the ECB to cut interest rates. And the falling dollar itself threatens severe economic and financial market dislocation, particularly if Asian central banks lose their appetite for ever more dollar assets.

Whatever the outcome, though, you should already be thinking about how the consensus will be dragged off by the bogeymen who always threaten to damage the reputation of the economics profession.

Stephen King is managing director of economics at HSBC

stephen.king@hsbcib.com