It’s the time of year when predictions are in order. Not by us, but by other people. We have spilled plenty of pixels on why forecasts are folly (see this, this, this, this and this); we won’t revisit that well-trod ground, at least not today. Instead, I wanted to discuss the rather annoying tendency of commentators to extrapolate market sentiment to well, infinity and beyond.

Two recent news items have reminded me that it’s time to discuss sentiment. The first was this weekend’s Barron’s Strategist Outlook; the second was the CFA Institute’s Global Market Sentiment Survey.

Let’s cut to the chase: Sentiment readings have very little correlation with markets, except when they reach extreme levels. I defy you to find anyone who has consistently made money off of the American Association of Individual Investors weekly sentiment readings; even on a monthly basis, they are so noisy as to be useless to traders.

In our office, we have been debating what it means when Barron’s survey of 10 top Wall Street strategists is devoid of bears. Collectively, they expect the Standard & Poor’s 500 Index to gain 10 percent next year. Their predictions range from 2,100 to 2,350 for the S&P (it’s now about 1,990), while they anticipate gross domestic product growth of 2.75 percent to 3.5 percent.

This could be taken as a contrary signal that we have become too bullish and are due for a major correction or worse. The problem is that strategists are bullish pretty much all the time, with very few bears showing up in most surveys.

There are many reasons for this: Markets tend to go up three out of four years, so most of the time it’s a good statistical bet to be a bull — even if that bet is based on the absence of any analysis. Then there is the career risk for being bearish: To quote Bob Kargenian of TABR Capital Management, “If you’re bullish and wrong, you usually have plenty of company. But if you’re bearish and wrong, it’s almost unforgivable.” Morgan Stanley’s former strategist Barton Biggs was reputed to have told a variation of that maxim: Bullish and wrong and clients are angry; bearish and wrong and they fire you.

These are variations of John Maynard Keynes’s admonition: “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.”

Consider this: Barron’s panel was bullish in their forecasts for 2014, 2013, 2012,  and 2011, too. Below is a table of the S&P’s performance and the forecasts:

 Year       S&P Return    Barron’s Outlook

 2014         7.5% (YTD)       10%

 2013        29.6%               10%

 2012        13.4%               12%

 2011         0%                   10%

As you can see, the forecasts are almost on autopilot and have only come close to being right once, in 2012 (this year is still to be determined).  Relying on the Barron’s Outlook for your investment strategy would be hit or miss at best. And if you shorted the market in any of the four years in anticipation of a decline, using the Outlook as a contrarian indicator, you would have lost money. 

Maybe we can glean something from the broader CFA Institute’s 2015 Global Market Sentiment Survey. Alas, it is based on an online survey conducted in October. Out of 119,817 CFA Institute members, 5,259 responded, giving a 4 percent response rate. This should put you on notice as to all sorts of sample set biases.

Regardless, the CFAs predict that the U.S. and China will be “the best investment opportunities.” So what else is new? Their forecasts for gold at $1,216 an ounce (now about $1,209) and crude oil at $91 a barrel (it was in the mid-$80s then and now is about $55) suggest a lot of recency effect at work — in other words, extrapolating the past into the future.

Can we learn anything from the CFA report? I doubt it. For context, we would need to see the past five or so years of this report, then try to determine if there is some sort of a pattern against which we could trade. 

The bottom line remains: The future is unknowable, sentiment does matter, but only at extremes.

Copyright 2018 Bloomberg. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.