Investor beware: don’t buy into the irrational exuberance of market analysts
Studies of share performance indicate most investors would be better off tossing a coin
Taking stock: investors should establish if a banking relationship exists between an analyst and the company being assessed. photograph: spencer platt/getty images
There’s an old Wall Street joke that you don’t need analysts in a bull market and you don’t want them in a bear market – a sentiment many Irish investors, still raw from the decimation of bank share prices, will share.
But what does the research say? Do analysts move share prices? Should earnings forecasts and price targets be taken seriously? Or do analysts even care if their forecasts are wrong?
Analyst reports can contain important points but should you take stock-price forecasts seriously? No is the answer. An analysis of 100,000 12-month stock-price forecasts made between 1997 and 2002 found prices were either at or above the target price on just 26 per cent of occasions.
Prices hit the analyst target at some stage during the 12 month-period on just 35 per cent of occasions. On average, actual prices were 35 per cent below the target price.
Not all analysts believe in such forecasts. GMO analyst James Montier scoffs at “pointless forecasts” and argues analysts should “return to doing as their name suggests: analysing, rather than trying to guess the unknowable”.
Analysts make directional market calls all the time but how often do they get it right? About 47 per cent of the time, according to the CXO Advisory website, which graded 6,459 public forecasts made between 1998 and 2012 – slightly worse than chance.
One of the more famous market timers, Marc Faber, often credited with predicting the 1987 market crash, exactly matches the overall accuracy rate of 47 per cent. Another high-profile forecaster, Goldman Sachs’ Abby Cohen, comes in at just 35 per cent. Worst of all, at just 22 per cent, is Robert Prechter, once a forecaster of great influence – George Soros once reported being “stunned” by Prechter’s decision to turn bearish prior to Black Monday in 1987, describing it as “the crack that started the avalanche”.
In fact, the most searched-for gurus on CXO’s site were typically only average to well-below-average in terms of accuracy. They are well-known due to “systematic” media exposure, not accuracy. Such gurus “may have motives other than accuracy in publishing forecasts”, CXO adds – they may simply be prone to “extreme forecasts to attract attention”.
It’s who you know, not what you know, that matters. According to one US study, analyst stock recommendations outperformed when the analyst went to the same university as a senior officer of the company being tracked. Investors could profit by creating a simple strategy of “going long on the buy recommendations of analysts with school ties and going short on the buy recommendations of analysts without ties”.
This pattern is even more obvious in the UK, the study found, diplomatically concluding that “school ties facilitate information-sharing”.
Ask the analysts
Earlier this year, US academics carried out a revealing survey of 365 market analysts. Less than a quarter said accurate and timely earnings forecasts helped determine pay packets. Just 35 per cent said making profitable stock recommendations was important in terms of pay.
One in six had been pressured by management to increase earnings forecasts; one in four had been pressured to lower forecasts (to allow the company beat estimates).
Almost 40 per cent believed issuing a downcast earnings forecast would result in losing access to management or being “frozen out” on conference calls.
And access remains crucial, despite regulations prohibiting the passing on of confidential company information. More than two-thirds said private phone calls with executives remained the most important part of their job. “It’s a way for them to broadcast,” one analyst said. “We’re sort of like a megaphone for them.”
Most analysts, one admitted, are “worried more about what management thinks of them than they are about whether they’re doing a good job for investors”.
Conflicts of interest
Many will say an analyst is unlikely to be objective if he is reviewing a company with whom his firm has an investment banking relationship. But is this fair?
Well, it appears so. A study entitled Do Independent Analysts Provide Superior Stock Recommendations? examined forecasts over the 1994-2003 period. (“Independent” in this case means not having an investment banking relationship between one year before and two years after a recommendation.)
Independent analysts were less likely to issue “buy” and “strong buy” recommendations and more likely to issue “sell” recommendations than “conflicted” analysts. “Strong buy” and “buy” recommendations of conflicted analyst significantly underperformed the market while their independent brethren matched the market.
Conflicted analysts’ “hold” recommendations significantly underperformed the market, while independents’ “hold” recommendations outperformed. Conflicted analysts were also slow to downgrade stocks.
The results are unequivocal, and strongly indicate investors should establish if a banking relationship exists between an analyst and the company being assessed.
“Analysts have been persistently over-optimistic for the past 25 years,” a 2010 McKinsey study found, projecting earnings growth of 10 to 12 per cent per year, compared to actual growth of just 6 per cent – almost 100 per cent too high. Earnings growth surpassed forecasts in only two instances during this period.
Why? James Montier has long argued analysts will acknowledge the need to lower forecasts in aggregate. However, when they discuss concerns with the individual companies they cover, they believe company assurances that it “won’t happen to them”.
“All the analysts think ‘their’ stocks are immune from the influence of the cycle,” says Montier.
As Thomson Reuters analyst Greg Garrison recently told the Wall Street Journal, analysts “always tend to be more optimistic when they’re looking far ahead”. Just as Montier suggests, they think any negative news will not affect their individual companies, only for reality to dawn as earnings announcements draw nearer.
Estimates get cut nearer the earnings date, and may be below the actual number by the time of the earnings report. The company then beats lowered estimates, allowing analysts to feel pleasantly surprised.
The same process happens year in, year out. Analysts’ so-called “optimism cycle” has even been suggested as a reason for the “Sell in May” phenomenon, whereby markets enjoy strong returns as they look towards the new year only for summer doldrums to set in as expectations get lowered.
Do “buy” and “sell” ratings move share prices? No, according to a recent paper that examined more than 44,000 changes in recommendations between 1997 and 2003.
The results were surprising. Previous studies had estimated analyst changes can move share prices by up to 4 per cent. However, the authors of this study said big share price moves were usually caused by significant news events – for example, company earnings – with the recommendations “piggy-backing” on unrelated news (almost 80 per cent of recommendation changes followed such news events).
The study filtered out the effects of these news events by examining share prices 20 minutes before and after the analyst recommendation changes.
Analysts are not the “market movers and shakers the world has come to think of them as being”. Their recommendations “don’t add much value and investors know it”, the paper states.