As world stock markets plunged last month, investors were able to take heart from the latest words of Warren Buffett: he announced he was buying stocks. "A simple rule dictates my buying: be fearful when others are greedy, and be greedy when others are fearful," proclaimed the man known as the Sage of Omaha. "And most certainly, fear is now widespread, gripping even seasoned investors."
Nobody could disagree. But the extreme fear that was in evidence did not guarantee that stocks had hit a bottom. Mr Buffett acknowledged as much in his column for The New York Times and advised investors not even to try to time the market. His point was that stocks were cheap, so people should not take the risk of waiting.
"I haven't the faintest idea as to whether stocks will be higher or lower a month - or a year - from now," he wrote. "What is likely, however, is that the market will move higher, perhaps substantially so, well before either sentiment or the economy turns up. So if you wait for the robins, spring will be over."
But is there a way of telling when equities might have hit bottom? Certainly, when fear overcomes greed to this extent, attempts to assess a stock's value according to fundamental measures go out of the window. Instead, all people have to go on is history. Yet precedents as extreme as last month's "Black October" are few. Only four periods in the past 100 years (two of them during the 1930s) match the 46 per cent slide since the Standard & Poor's 500, the world's most widely tracked index, reached an all-time high in October last year.
In the weeks since its latest extreme sell-off, the S&P regained as much as 20 per cent at one point but continues to trade in a volatile range. Almost all other markets worldwide have displayed similar patterns, shedding even more than the US and rebounding more before lapsing into a range.
Nobody thinks that anything like normality has been restored; the weeks since then have included the biggest crash in hist-ory for the foreign ex-change market and further acute difficulties for the credit market. Historical pointers look ominous too: after the crash of 1929, anybody buying when the market started to stabilise would have lost more than three-quarters of their money over the next two years.
But some are beginning to set aside such fears. "For an unparalleled 20 years, global equities, especially US equities, have been overpriced," Jeremy Grantham, founder of the Grantham Mayo Otterloo fund management group in Boston and for years a voice that counselled caution, wrote late last month. "Now, finally, they are cheap and likely to get cheaper. Likely, I believe, to set up a once-in-a-lifetime investing opportunity (or maybe twice in a long career)."
He is only starting to buy, however, and has not deployed all the cash he has available. Buying now makes sense for those truly in it for the long term; the bottom may still be ahead.
W hen markets experience dislocations as extreme as Black October, as Mr Buffett implies, what is important is the mass psychology reflected in the market's moves.
Indeed, there are ways to assess investor psychology. The Chicago Board Options Exchange's Vix index, for instance, measures "anxiety" from how much investors in options are prepared to pay to protect against future falls in the S&P 500. This measure was not around during previous great bear markets but the spike in volatility in recent years took it to all-time highs. When sentiment becomes this extreme, it is fair to hope that stocks have become too cheap, creating bargains once confidence returns.
Looking instead at realised volatility, or the extent to which stocks vary on a monthly basis, the S&P last month was as volatile as it had been at any time since 1929, except October 1987. All previous spikes in volatility returned swiftly to calm.
Yet another measure comes from US Treasury bills, short-term loans to the US government that are arguably the safest investments in existence. When T-bill yields fall, investors are prepared to pay more for that safety. At the worst of the recent panic, T-bill yields reached 0.02 per cent, their lowest since 1940. They have barely recovered. Again, such an extreme desire for security might imply that the market cannot fall much further and is ready to swing in the other direction.
A further measure, comparing the yields available on the highest-rated - triple-A - bonds with bonds of only slightly lower quality once again shows the desire for safety leaping to extremes not seen since the 1930s.
A popular model - although academics suggest that it is flawed - compares the yields on stocks to the yields on bonds. The model does have a kernel of common sense. When bond yields are very low, as they are now, that could signal that it makes sense to pay more for stocks, in search of a higher return.
There is therefore a respectable case to be made that in the past few weeks fear has conquered greed to an extent that goes beyond the rational. The return of stability in due course creates the opportunity for a big rally. But this may merely be a "bear market rally". In the past two weeks, the S&P both staged its 20 per cent revival and then gave up half of it. That is not surprising - the bear market after 1929 was punctuated by several rallies of 20 per cent or more.
A market bottom cannot hold if stocks remain too expensive, so long-term measures of valuation have proved to be great indicators of bear market tops and bottoms. One suggested in the 1930s by Benjamin Graham, an investment theorist, looks at cyclical price/earnings ratios, where the price of a stock is compared with the average of the earnings per share it has produced over the previous 10 years. This controls for the effect of the profit cycle, which otherwise would mean that year-by-year multiples tended to be higher when profits were at the low point of the cycle and lower when profits were high.
This measure has been revived by Yale University's Robert Shiller, who gained fame for publishing Irrational Exuberance, in which he warned that stocks were badly overvalued, shortly before the internet bubble burst in 2000. Enhancing that fame in recent years have been his indices showing that another bubble, in US housing, was about to burst. Shiller p/e multiples reached their greatest extremes in 1929 and 2000, two market tops that were followed by bear markets. In October 1987 they peaked at a much lower level, giving an important clue that the market break of that year would prove to be less significant. Further, the lows have coincided with long-term market bottoms. There is also a tendency to revert to the mean over long periods.
Significantly, the cyclical p/e now suggests that stocks are slightly cheaper than their long-term average, for the first time since 1991. That in turn implies that those who can afford to wait a number of years for a pay-off should start to feed money into the stock market.
A variation on this theme is to look at very long-running trends. There are various ways of doing this, but Mr Grantham's simple model holds that over time, stocks rise by inflation plus 2 per cent. This produces an upward trend line, to which stocks will eventually revert. On this basis, stocks have recently snapped back to fair value, having been wildly overpriced. As with the cyclical p/e model, this suggests that the stock market had been overvalued since 1991.
Again this is encouraging, but the news has to come with the caveat that markets are prone to overshoot and become too cheap after prolonged periods when they have been too expensive. Hence, both these measures are consistent with stocks falling much lower before they find a bottom, even though they are currently fairly priced.
A further measure that has signalled market tops and bottoms over the years is known as q. This refers to the ratio of the value of a company to its net worth - the value of the equity on its balance sheet. Again, over time, the market value of a company will tend to converge on this measure. Again, after being overvalued for more than a decade, stocks now appear to be at about fair value according to q.
T he final, critical element in the equation concerns the future. How bad will this recession be? Usually, market bottoms are related to recessions, with markets generally beginning to recover when the economy is on the floor and some months before a broader upturn.
For those who believe this recession will end in the early months of next year, therefore, there appears to be an argument to buy stocks now, as they already implicitly discount a long and severe recession. Tim Bond, of Barclays Capital, suggests that equities might present the "buying opportunity of a generation". According to Barclays' model of implied economic forecasts from credit spreads, the market is discounting as much as a 15 per cent dec-line in real gross domestic product for the US next year. By way of comparison, the average annualised GDP contraction during the Depression was 14.1 per cent and the worst single year, 1932, saw a contraction of 23 per cent.
If the final scenario is any better than this - with the political response to this recession very different as lessons of the 1930s have been learnt - stocks could be ready to rise. But Albert Edwards, a strongly bearish equity analyst at Société Générale in London, suggests that this is a mistake. "Previously, equity rallies in the depth of recessions were driven not by predictive power but by expanding multiples as interest rates and bond yields collapsed. Bad economic news was good news for valuations."
This time, he suggests, the world is dealing with a historic overhang of debt. Interest rates have already been slashed without turning round either the market or the economy. Rather than correlating with bonds, therefore, equities are more likely to become correlated with economic growth. That in turn means stocks might not rally ahead of an economic recovery.
A confident attempt to call a bottom to the market requires, in any event, a belief that the economy will be rebounding soon. Yet there is a flip side to that. "It is typical for great bubbles to overrun badly. But usually we don't invest our money on estimated likely overruns, but instead filter our money in slowly and hope to get lucky," says Mr Grantham.
"After all, if stocks are attractive and you don't buy and they run away, you don't just look like an idiot, you are an idiot."
Down time
The mother of all bear markets started in 1929, lasted at least until 1932 and coincided with the onset of the Great Depression. The Standard & Poor's 500 index fell 86 per cent.
Although many market scholars would count it as a separate incident, a further drop in equities began in 1937. The S&P endured a new fall of 54.5 per cent in the space of a year, though ending above its high from 1932. It took 25 years to get back to the 1929 peak.
The precedent is alarming: at this point in the 1929-32 bear market, stocks had another 75 per cent to fall before hitting bottom.
Worst of days
If another example is needed, there is always the Black Monday crash of 1987. But this could be a red herring. It was the worst single day in stock market history, but the plunge came when the market had grown very frothy, at a time when the economy was reasonably strong, and there was a swift rebound.
The economic backdrop this time is far more menacing and the market has now fallen far more from top to bottom. Comparisons with 1987 are therefore wishful thinking.
Oily slide
The other era that most closely compares is the bear market that began in 1973, coinciding with the oil price spike and the beginning of the great "stagflationary" cycle of the 1970s. The S&P fell 47.8 per cent.
At its worst so far, the 2007-08 sell-off displayed a virtually identical loss of 46.9 per cent - after reaching these depths much more swiftly.
A few months ago, with fears of inflation resurfacing as oil prices surged to unprecedented levels (just as they did, for different reasons, in 1973), this was a very popular parallel. But now, with central banks cutting rates to fight fears of deflation, the comparison looks stretched.
False bottom?
After the "internet bust" of 2000, the S&P fell 49.1 per cent. This is fresh in the mind but its lessons are questionable. It started with stocks more overvalued than ever before - and the sell-off stopped before a return to normal valuations.
But was the bottom during the WorldCom scandal of 2002 a false one? On one theory, the rebound came only because of excessive cuts in interest rates that stoked the credit bubble. That implies that this latest sell-off can end only once it has tested whether 2002 was a false bottom - meaning stocks need to fall significantly from last month's low. If they breach that, parallels with the 1930s look all the more disquieting.
