IMF gives no reason to think the worst is over

By Tony Jackson  Published: November 2 2008 16:24

The European Commission's latest dire survey of business confidence - described by one analyst as a "war bulletin" - is clear evidence that the long-expected hammering of corporate earnings has arrived.

It bears out one of the more startling statistics of recent weeks: that Volvo's European truck sales in the third quarter fell from last year's 42,000 to just 115.

At long last, sector analysts are beginning to catch on. According to Merrill Lynch, the ratio of earnings downgrades to upgrades, in both Europe and the US, was last month the highest in 21 years surveyed.

Even the top-down strategists, who have been warning of this for many months, are getting gloomier. Citigroup now thinks global earnings could fall 50 per cent from top to bottom. And that, as I shall discuss in a moment, could be an underestimate.

Bizarrely, though, the market still seems to find profit warnings surprising. Last week, GKN, the UK engineer, which has two-thirds of its sales to the auto industry, said this year's profits would be down 20 per cent. The shares fell 10 per cent on an otherwise quiet day. The news of collapsing world car sales, it appears, had somehow passed investors by.

That makes yet more pressing the question of how far corporate earnings have to fall. But first, let me touch on an arresting thesis set out in the latest World Economic Outlook from the International Monetary Fund.

In a study covering 17 developed economies over three decades, the IMF came up with three findings of particular relevance.

First, recessions preceded by a financial crisis tend to be deeper and longer than others.

Second, they tend to be worse again if the crisis is in banking, rather than in securities markets or foreign exchange.

And third, the countries hardest hit are those with so-called arm's length financial systems, such as the US or UK. If banks are free to innovate, they tend to build up more pro-cyclical leverage.

In practical terms, the IMF found that recessions linked to banking crises lasted twice as long on average as those not linked to any financial crisis, and the cumulative loss of output was about four times as great.

The present recession, plainly, ticks all the boxes - a fact that makes it unlike any in recent memory.

That of 1990-91, for instance, was only partly financial and not much to do with banking, being preceded by the Japanese stock market collapse and a junk bond crash in the US.

There was a severe Scandinavian banking crisis after that recession had got under way, but that was primarily local. What we have today is an increasingly global banking crisis, which is bad news all round.

With that, let us turn to the Citigroup thesis on the coming earnings recession. Citi envisages a 50 per cent fall globally, of which 10 per cent has happened already. And partly with the IMF findings in mind, it says the process could last up to three years.

Crucially, though, Citi assumes that corporate return on equity will only fall to the 8 per cent trough seen in the past two earnings recessions of 1992-93 and 2001-02.

The fall seems steeper this time merely because of the starting point - the return on equity at the peak having been a record-breaking 16 per cent.

But if the IMF thesis is right, there is every chance this recession will be more severe than the other two. There are also several factors at work that were less powerful before.

One has to do with pensions. The true scale of deficits by this time is anyone's guess, but one estimate last week suggested US corporations might have to put an aggregate $50bn into their funds in each of the next two years. In the old days the effect, such as it was, would have been hidden, not displayed on the face of the accounts.

Similarly, corporations of all kinds now have to mark all kinds of assets to market. Earnings will therefore be dented by the writedown of paper profits recorded in the bubble years.

What about valuation? Citi puts the global price-earnings ratio now at

10.4 times. And since one-fifth of the projected fall in earnings has already happened, that puts the trough p/e at 17 - bang on the long-run average.

Therefore, Citi says, the 50 per cent earnings fall it envisages is in the price. That is not dissimilar to the finding I wrote about last week, that the present p/e in the US is in line with the 10-year trailing average.

But, as I also said last week, that does not by any means imply the market has finished falling.

Equities are going through a welcome spot of calm at the moment. I do not expect it to last.