Despite the bear market, analysts are predicting record earnings in the coming quarters, writes Proinsias O'Mahony

US ANALYSTS continue to be too optimistic in their earnings estimates, with new data showing second-quarter earnings forecasts were the most inaccurate since records began.

According to Bloomberg data, estimates matched actual results in fewer than 7 per cent of companies that reported a second-quarter profit. That's the lowest level since the firm started compiling the data in 1992. Year-over-year earnings declined by 23 per cent in the second quarter, or more than twice the 11 per cent decline forecast as recently as July.

Analysts have been consistently behind the curve in the present bear market. Earnings estimates from almost 1,800 Wall Street analysts were 8 per cent too high in the third quarter of 2007. The following quarter saw that figure swell to 33 per cent, when growth predictions of 11 per cent were followed by an actual 22 per cent decline.

At the start of the year, profits at financial companies were forecast to rise by 22 per cent. Now they're expected to decline by 48 per cent.

Despite attracting widespread opprobrium for their inaccuracy, analysts' forecasts continue to look wildly optimistic. Analysts are predicting financial stocks will return to profitability in the third quarter, with overall Standard & Poor's (S&P) 500 earnings hitting record levels in the fourth quarter.

Operating earnings of $24.61 per S&P unit are forecast for the fourth quarter, which would surpass the quarterly record of $24.06 set in the second quarter of last year. Next year is forecast to be even better, with operating earnings projected to be more than 23 per cent above the annual record set in 2006.

Price targets, too, suggest analysts have not recognised the emergence of a bear market. At the beginning of the year, a Bloomberg survey of equity strategists found that analysts were expecting the S&P 500 to gain 11 per cent this year.

With the exception of Deutsche Bank, which has retained its price target and is predicting that US markets will end the year 30 per cent above current levels, all strategists have reined in their original estimates. Despite that, they continue to take a positive viewpoint, estimating US markets will gain by almost 14 per cent between now and the end of December.

In Europe, too, analysts have been playing catch-up with reality. Earnings growth of 10 per cent was predicted at the beginning of the year. That has since been slashed to 1.2 per cent.

Analyst inaccuracy is not a new phenomenon. Contrarian fund manager David Dreman has studied the subject of earnings forecasts extensively. In his book Contrarian Investment Strategies, he concludes that "your odds are 10 times greater of being the big winner of the New York State Lottery than of pinpointing earnings five years ahead".

Stephen McClellan, former analyst and author of Full of Bull, a new book that takes a sword to Wall Street practices, is similarly sceptical. "Wall Street is not suited to be an investment manager, financial adviser, or stock selector," he writes. Riven by an inherent "conflict of interest", analysts are "eternally bullish".

British equity strategist and behavioural finance expert James Montier takes a kinder interpretation.

Complaining earlier this year that earnings forecasts were far too optimistic, he wrote that analysts "all acknowledge the sense of lowering forecasts in aggregate. However, when they discuss such a move with the companies they cover, the companies response is that it won't happen to them".

Montier labels it a "fallacy of composition problem" in which "all the analysts think 'their' stocks are immune from the cycle".

Others argue that the current environment is notoriously difficult to predict. Still, none of the above observations explains why forecasting accuracy has been steadily declining for years.

That deterioration, it seems, is caused by a law introduced in 2000 which demands that companies release any information that affects profits to the public, thereby prohibiting executives from giving "whisper numbers" or sensitive information to favoured analysts. Analyst accuracy peaked that year and has fallen in six of the seven years since.

Irrespective of the reason why analysts have been so far off-track, earnings forecasts are likely to come down this year.

"We continue to believe that the contagion effect of subprime and credit-related issues within financials has largely been underestimated," writes Brian Belski, sector strategist at Merrill Lynch.

"Therefore, we believe there remains a high likelihood of additional weakness to earnings estimates heading into the months ahead until analysts and companies alike become even more conservative with their assumptions."

Howard Silverblatt, equity analyst at Standard & Poor's, is more blunt.

Questioning projections of record earnings, he says analysts need to "get real".