Wall Street’s cautious bulls are likely to be wrong – again

There’s little reason to have faith in market strategists’ 2016 predictions

Wall Street strategists expect stocks to climb next year, although many expect returns to be muted, while precious few are pounding the table for equities. Is there value in such forecasts, or should investors ignore the crystal ball gazers?

Well, forecasters certainly got 2015 wrong.

"Wall Street's top strategists are uniformly upbeat about the outlook for stocks in the year ahead," Barron's reported 12 months ago, with the average strategist predicting a 10 per cent gain. Instead, stocks treaded water. Even at last May's all-time high, the S&P 500 was a mere 3.5 per cent above where it began the year. Autumn's plunge was followed by a decent rebound, but stocks ultimately went nowhere in 2015.

This year, the mood among strategists is generally one of subdued bullishness.


"Stay positive about the third-longest bull market on record, but curb your enthusiasm," reads the Barron's headline. Every strategist surveyed by the magazine predicted gains, with the mean prediction being an advance of around 7 per cent for the index. A separate Reuters poll of more than 300 analysts and fund managers came to an almost identical price target.

In many cases, strategists’ 2016 price targets are either at or below the same levels predicted 12 months ago, with some high-profile firms predicting below average returns. Goldman Sachs, for example, envisages another flat year for stocks, warning that valuations look challenging; Morgan Stanley says stocks are “likely headed for a choppy year of low returns”; Credit Suisse is concerned on multiple fronts, warning of “increasing headwinds” for equities; Merrill Lynch says to stick with stocks, but to expect “tepid” returns.

Always wrong

Despite this caution, no major firm is predicting that stocks will slip in 2016. Reassured? Don’t be.

Wall Street never forecasts anything other than gains. Earlier this year, Motley Fool finance columnist Morgan Housel analysed the forecasts of the chief strategists at 22 of Wall Street's biggest banks and investment firms. Between 2000 and 2014, he found, the average strategist did not forecast a single down year.

During the same period, investors suffered two major market crashes – the bursting of the dotcom bubble in 2000, which eventually led to a 78 per cent fall in the tech-laden Nasdaq, and the global financial crisis of 2008-2009, when the S&P 500 plummeted by 57 per cent.

Strategists’ forecasts, it turns out, are even more inaccurate than sceptical investors might have expected. Incredibly, they have been off by an average of 14.7 percentage points per year.

Nor are the figures wildly distorted by the global financial crisis: even if you exclude 2008, the annual error rate is a huge 12 percentage points, notes Housel. The average forecast came close to the market reality in just one of the 15 years.

Wall Street strategists are often thought of as perma-bulls, but the stats show they are as likely to underestimate bull markets as they are to overestimate bear markets. Indeed, while strategists were too optimistic in 2015, stocks actually outperformed average forecasts in eight of the previous 15 years.

This habitual inaccuracy is nothing new. Way back in 1932, US economist Alfred Cowles undertook a detailed study of market forecasts over the previous five years. The most successful forecasters, he found, did "little, if any, better than what might be expected to result from pure chance", while the poorest records "are worse than what could reasonably be attributed to chance". In 1931, Cowles noted, there were 16 bullish forecasts to every three bearish; stocks went on to fall by 54 per cent that year.

Ineptitude on forecasting is not confined to stock markets, as anyone familiar with the work of psychologist Philip Tetlock will know. Tetlock's 2005 book Expert Political Judgment documented his study of 82,361 forecasts made by 284 experts over the 1984-2004 period. His conclusion: the average expert "is roughly as accurate as a dart-throwing chimpanzee".

However, this doesn't mean successful forecasting is impossible, says Tetlock, who describes himself as an "optimistic sceptic". His new book, Superforecasting, relates the findings of the US government- funded Good Judgment Project, which tracked the real-time predictions of thousands of amateur forecasters over the last five years. Participants were trained in how to think in terms of probabilities, and encouraged to seek out information that might disprove their ideas. A small number of forecasters consistently outperformed year after year.

In fact, Tetlock found the best forecasters were even able to come up with probability estimates that were 30 per cent better than those of intelligence officers with access to classified information.

The elite forecasters were smart – on average, they scored above 80 per cent of people in IQ tests – but not geniuses. Modest and self-critical, they were “ordinary people, from former ballroom dancers to retired computer programmers”. Curious, data-driven and objective, they were invariably more interested in why they were right or wrong, rather than whether they were right or wrong. In contrast, most experts are “too quick to make up their minds and too slow to change them”.

Investors can learn from the superforecasters, says Tetlock. Beliefs should be seen as “hypotheses to be tested, not treasures to be protected”, while people should cultivate probability estimation skills.

On the latter point, even sophisticated thinkers make elementary mistakes. He refers to a column by New York Times Pulitzer Prize-winning journalist David Leonhardt, who said an online prediction market was "wrong" because it said there was a 75 per cent chance the US supreme court would rule Obamacare to be unconstitutional. However, the forecast was not "wrong" – it simply estimated there was a 25 per cent chance that the law would be upheld.


When estimating the probability of something happening, investors should always start with the base rate – that is, the average historical experience. For example, investors often buy into excitement around initial public offerings (IPOs), believing they may have found the “next big thing”. Few IPOs are big winners, however. In fact, research shows the average IPO is significantly overvalued, with most companies going on to badly underperform.

Market strategists can be similarly neglectful of base rates. In its 2016 market outlook, UBS says it expects "further volatility", which gives the impression 2015 was a year characterised by excessive volatility. In fact, the S&P 500 was never up more than 3.5 per cent during 2015 and never down more than 9.3 per cent. The spread between the year's high and low points was just 12.8 per cent, notes Michael Batnick of Ritholtz Asset Management – roughly half the average range seen over the last 45 years.

This same neglect is evident in strategist price targets. As mentioned earlier, the average strategist never predicts stocks will decline, even though annual declines are common – the S&P 500 has fallen in 24 of the last 89 years.

Financial firms are also wary of being labelled perma-bulls, however, so they are invariably cautious in their price targets, usually predicting gains of around 7 to 10 per cent.

Investors might assume this is in keeping with the historical base rate, given that stocks have historically averaged gains of over 9 per cent.

In reality, such returns are a historical rarity – over the last 85 years, equity returns have been in the 0-10 per cent range on just 14 occasions. In contrast, stocks have historically either risen or fallen by more than 20 per cent in more than 40 per cent of all years, but precious few strategists ever predict annual gains or losses of that magnitude.

The message from Philip Tetlock is that accurate forecasting is possible, but only if you follow an empirical, data-driven approach. In contrast, market strategists are ignoring historical base rates and sticking to the same forecast they churn out every year.

Those forecasts have been wrong in 14 of the last 15 years; 2016 is unlikely to be any different.