PsyFi Search

Monday, 20 March 2017

Idiots in Investing Echo Chambers

Investing in the Dead

Some people like to wander round graveyards. I get the same sort of ghoulish pleasure out of frequenting investment discussion boards. They're full of Pollyannas, forever only able to see the good in the stocks they invest in.

Sadly they're almost always wrong. But it's kind of fun watching them keeping the reanimated corpses of zombie stocks moving about through the power of sheer stupidity.

The Glad Game

Pollyanna was the eponymous child heroine of Eleanor Porter's 1913 novel whose main characteristic was a determination to find the good in any situation, no matter how grim.  Unsurprisingly, then, the Pollyanna Principle refers to the behavioural bias in which people see only the positive side of any particular situation, a trait otherwise known as exhibiting positivity bias.

Pollyannaism may well have been a valuable evolutionary trait, enabling our ancestors to strive for continued existence in an unfriendly world devoid of anything my children would regard as life's essentials: good cellphone signals, unlimited broadband and an infinite supply of sugar. However, in these more enlightened times it can be a positively negative force. Especially for determinedly positive investors in the smaller type of "exciting" stock.

Bigger, Better: Older, Wiser?

So, technically, Pollyannaism is a tendency for people to remember nice stuff more than nasty stuff. It's closely associated with optimism bias, for which there's decent biological evidence suggesting that it's hardwired into our brains (for this read Unrealistic Optimism and the Impoverished Investor or anything by Tali Sharot). The original research for the Pollyanna Principle was produced by Margaret Matlin and David Strang in 1978 who showed that people judge pleasant stimuli to be larger in size than unpleasant or neutral stimuli. 

In related research it's been shown that we think more desirable objects are closer than less desirable ones and the older we are the more we prefer positive or neutral images (see Eric Allard and Derek Isaacowitz (2008) and Andrew Reed, Larry Chan and Joseph Mikels (2014)). This latter finding, the age-related positivity effect is possibly because the older you are the more you realise that life is too short to focus on the downside of life. 

So how does this read across to bulletin boards?

Talking Down

Ed Croft, who's the CEO of the UK based social investing website Stockopedia has done some interesting research on the performance of stocks that are the favourites of UK bulletin boards. This showed that over a couple of years the 100 most discussed stocks on the UK's favourite discussion boards yielded a magnificent negative 22% compared to the negative 3.6% you'd have got investing in an index tracking the UK's biggest stocks.

(In the interests of transparency ... my only connection with Stockopedia is the large check I give them for access to their data analytic tools each year).

There are a whole bunch of reasons why this underperformance occurs, some of which I've discussed in the past. Confirmation bias is probably top of the list, where everyone simply affirms everyone else's beliefs but Pollyannaism is likely a close second.

Echo Chambers

Usefully, Shiliang Tang, Qingyun Liu, Megan McQueen, Scott Counts, Apurv Jain, Haitao Zheng and Ben Y. Zhao  have published a recent paper on this very subject - Echo Chambers in Investment Decision Boards. And very interesting it is, too:
Discussion post sentiment has negligible correlation with future stock market returns, but does have a positive correlation with trading volumes and volatility. Our trading simulations show that across different timeframes, this misinformation leads 50-70% of users to underperform the market average.
If that is even vaguely accurate it explains Ed Croft's research - between 50% and 70% of discussion board users underperform the market. They go on to analyse how this actually works and identify that there are a few, key people who determine the overall sentiment and that:
"... many members actively resist negative sentiment, thus minimizing viewpoint diversity"
Well, who'd have thought?


In detail they identify four critical problems that underlie the underperformance:

#1: Poor User Generated Information: most bulletin board posters have trading strategies that don't work even though, presumably, they think they do. Board sentiment about stocks is uncorrelated with the stock's future performance. Or, in other words, opinion on bulletin boards has no predictive value.

#2: Failing to Incorporate New Information: board sentiment about stocks is uncorrelated with the stock's past price movements. Basically, if stuff happens it's ignored, new information is not assimilated.

#3: Resistance to Viewpoint Diversity: most board sentiment reflects only a small number of actual users and users tend to resist negative viewpoints. So actually only a few people are responsible for the overall direction of opinion - and then there's an active intolerance of alternative views.

#4: Mistimed Activity Adds Noise: the volume of posts about a stock is predictive of future stock trading volume and volatility. But, as we can see from the previous points, this is uninformed opinion so it's just adding noise (behaviour I previously blogged about in Idiot Noise Traders). There's no real signal in this because no new information in being incorporated in the trading decision, other than the willingness of people to post positively about the stock.

So basically we have people who have no information and no ability to analyse information and who refuse to countenance alternative ideas, convincing each other to trade. No wonder the most popular boards see significant underperformance. It's a wonder these people can afford internet access let alone have the spare capital to trade stocks.

The Fix

Most problems exposed by behavioural finance research are quite hard to fix. It's one of the reasons it's nowhere near as popular a topic as it ought to be, it just doesn't give easy answers. For those you'll need to go read some other investing blog. You'll be in good company, it'll be more popular.

The truth is that investing should be done as much as possible based on objective data and as little as possible based on subjective opinion, especially the Pollyanna-ish opinions of others. Even then some investments will go wrong and some investments that have gone right will eventually go wrong. Investing is hard, otherwise everyone would have a yacht.

But in this case the answer is very easy. I'll leave the last word to Ed Croft, who after all runs a website that includes discussion boards. Ed's conclusion:
"The simplest secret to making money reading bulletin boards is simply don't!"

Pollyanna Principle and positivity bias added to the Big List of Behavioral Biases

Related Posts


1 comment: