Shares digital

The performance potential of a stock appears to be inversely proportional to the esteem that analysts hold for it. Peter Webb proposes the inverse share law, which could reveal some bargains in the dregs of market opinion and some expensive froth at the top

Contrarian investing is all about picking out of favour stocks that appear undervalued. The market has been shown to consistently overvalue high-flying stocks and undervalue less-favoured stocks. Why is this possible when analysts are paid large amounts to analyse those very same stocks?

Overestimating future performance could be one reason. A review based on analyst estimates from the Value Line Investment survey, took 918 stocks for the 1987-89 period and found analysts on average overestimated earnings by approximately 9%. IBES, the largest earnings-forecasting service also found overestimation. Worse still they found that, despite revisions taken during the year, analysts’ recommendations still seemed to be far too optimistic.

The Journal of Finance also published an early analysis of long-term estimates by John Cragg and Burton Malkiel, author of A Random Walk Down Wall Street.

Their work – The Consensus and Accuracy of Some Predictions of the Growth of Corporate Earnings – looked at the projections made by groups of analysts at five respected firms, covering 185 stocks.

Random future

The researchers found that most analysts’ estimates were primarily based on linear extrapolation of current trends. In other words they just took past data and projected it into the future. This generated low correlations between actual and predicted earnings. Curiously, they found that analysts would have substantially improved their accuracy if, instead of extrapolating past growth rates, they had simply inserted the long-term company average growth rate of 4% annually!

Another study found that in fact corporate earnings actually seemed to follow a random walk. There appeared to be little correlation between past and future rates. Recent trends appeared to provide no insights useful for forecasts. A number of similar studies reached the same conclusion: changes in company earnings appear to fluctuate in an essentially ‘random’ fashion.

Distorting pressures

As well as general troubles with forecasting, a variety of other pressures act directly on analysts. Look at the list of what determines analysts’ bonuses and you might be surprised by what doesn’t go into calculating them. Accuracy of profit estimates is unlikely to be given major weight. Analysts also run a substantial political risk by downgrading a company. If an analyst recommends a sell, that company may welcome no further contact. And that analyst well knows that a sell order on an entire industry may close many doors – even end the career of a specialist in that industry.

For a case in point look at what happened when an analyst issued a sell recommendation on an Atlantic City casino owned by Donald Trump in the late 1980s. Trump was infuriated by the recommendation and insisted the analyst be fired for his lack of knowledge. Shortly after, the analyst decided to ‘spend more time with his family’. However, it turned out he was correct and soon the Casino went broke.

Sell orders don’t just affect the analyst personally. They can damage the entire firm. When an analyst at Prudential wrote some negative reports about Citicorp in 1992, Citicorp retaliated by refusing to give Prudential any business in underwriting bond issues. The same happened a year later when the same analyst criticised Banc One for its complex derivatives holdings. Banc One ceased its bond trading with Prudential. It is forgotten in the mists of time that Banc One eventually shed millions in derivative losses.

Errors and prices

Of course, the most important question we should ask is what do analyst forecasting errors do to stock prices?

One study measured just how accurate investors expect analyst forecasts to be. To do this it looked at any positive response to earnings news on the day it was accounted. The study grouped stocks into five sets or quintiles ranked by common value methods such as price to earnings.

The top 20% were considered the most favoured stocks, the bottom 20% were the least favoured. Overall, the bottom, ‘cheap’, 20% of stocks responded positively to earnings surprises. Per quarter, they averaged 1.5% above market and over the full year they beat the market by 4.2% a year. This means that the combined effect of all surprises, positive and negative, seemed to work in favour of unpopular ‘cheap’ stocks. Middle quintile stocks performed just below neutral and the top 20% ‘expensive’ stocks underperformed, returning 1% a quarter, or 3.5%, a year, less.

The conclusion of all this evidence is that unfavoured stocks, which have probably already discounted bad news, respond well to positive surprises. Expensive stocks, however, have already discounted good and growing earnings and therefore gain little. Perhaps one of the causes is that analysts appear to be over-optimistic for a number of reasons and discount too much into future earnings. In fact, accurate earnings forecasts are incredibly difficult to achieve and simply taking a much simpler guess is often more effective.

The best way to take advantage of this high rate of error is by simply investing in out-of-favour stocks. Doing so will limit your downside but give you plenty of upside potential.