The wisdom of crowds

But when aggregating the individual ‘guesses’ of a large number of individuals, the margin of my error will be more or less offset by someone else’s. With a large enough crowd, the net result will be eerily close to the actual number of ping pong balls.

Where crowd wisdom becomes unreliable is when you begin to introduce anchor points.  Telling an individual ‘there are 5,000 jellybeans in this can’ before asking how many ping pong balls are in that car will bias the subject toward the 5,000 number – even though there is obviously no relationship between the two scenarios.

A number of conditions must be present to maximise the wisdom from a crowd and to minimise irrationality or biases influencing outcomes.  Among these, constituents should be diverse and independent.  Less communication between group members is also better. Importantly, individuals needn’t be particularly intelligent, with the group as a whole inevitably being more accurate than the smartest of its individual members.

Tying this back to market efficiency, a number of factors undermine the process of the crowd reaching an accurate consensus. First, analysts are often not independent.  If a brokerage firm isn’t seeking to do business with the company in other parts of its business, for example by organising financing that produces large fees, the individual analyst may feel some sense of obligation to the company. Perhaps the company gives good access to the analyst and the analyst feels this may be compromised if they produce an opinion wildly different from the consensus.

Second, we’re anchored before we begin the analytical process by the company’s share price.  Imagine analysing a company and having no clue what the prevailing share price is. Despite the wild share price swings from quarter to quarter or year to year, it’s rare to see analyst price targets that are materially different (say, +/- 50%) from the prevailing price. The human desire to avoid looking like a fool amongst your peers doesn’t end when you leave high school.

Third, we’re not so diverse.  A large global company may have 20 or 30 broker analysts researching it, but the backgrounds and life experiences of those analysts may be very similar. They belong to the same industry and may also socialise together. The environment is conducive to herding behaviour, which provides opportunities for investors who can exploit these biases by reaching their own intelligent conclusions.

We recently took a snapshot of the broker recommendations on Rosenbauer’s website:

[analyst ratings.PNG]

Rosenbauer share recommendations

Source: Rosenbauer

For context, the stock price recently touched €50, and has been tracking in the 55 to 60 euro range for most of this year. At the time of their publications, each analyst was bullish with a ‘buy’ recommendation, and each had targets around 15% or so above the prevailing market price. Very sensible. Very safe.

Two weeks later, Markus Remis retained his Buy rating but decreased his price target to €59.5, though there’s been no material change to the business. However, in August the share price fell 10-15% below their price targets from May. Ratcheting down his price target keeps it tight relative to the new share price. We wonder, will Christian and Stephan follow suit?

This might seem comical, but it reinforces some critical lessons. First, yes, markets can be rational and efficient, but that doesn’t mean they are always rational and efficient. Markets are comprised of humans, and humans are social animals. We are influenced consciously and subconsciously by others. We fear being wrong, but it is much more acceptable to be wrong with the crowd than wrong on your own.

Second, think independently. Seek out contrary opinions and try to destroy your ideas. If you can’t kill an investment case, it could be a great investment.

Lastly, University-educated, experienced, professional analysts aren’t necessarily going to achieve better outcomes than sensible, reasonably intelligent individuals who have a passion for seeking the truth, understanding how businesses work, and identifying great businesses to buy at attractive prices. As long as humans dominate investing, there will always be undervalued securities, and it’s our job to find them. Understanding human biases gives you a great head start.

Contributed by Peters MacGregor:  (VIEW LINK)