Market Ethos: When is a sell really a buy?

Market Ethos: When is a sell really a buy?

by Craig Basinger, Chris Kerlow, Derek Benedet, Shane Obata, Connected Wealth, RichardsonGMP

Many investors agree, it is best to be a contrarian when it comes to making money. Betting against the herd is often a more profitable endeavour than simply following the crowd. The reasoning is pretty straight forward, if everyone likes or dislikes something that is probably already baked into the price of the asset. But being a contrarian is also often lonely. And when it comes to investing, in a world filled with uncertainty, many choose to stick with the crowd or consensus instead of taking that lonelier contrarian road.

Pretend you were considering an investment in one of two companies. For the sake of this exercise, they are pretty much identical: similar quality management teams, similar valuations, operate in the same industry, etc. The biggest difference is one company has eight buys while the other has two buys, five holds and a sell. Which would you choose?



If you were a contrarian, you may be equally or more interested in the 2nd company. But most would choose the one with all buy ratings. This reduces the risk of regret, a powerful behavioural bias. If the share price declines and you followed the herd into the company with all buy ratings, you can justify your decision with “well everyone said to buy”. However, if you purchased the 2nd company and the share price declined, well you own that since the analysts clearly didn’t like that one. Regret avoidance is a strong emotion.

The tricky part is, it is more often the companies with fewer buy ratings that tend to perform better. We tracked the performance of all S&P 100 member companies over the past ten years, broken down by performance into quintiles. Quintiles represent the top 20%, next 20%, next 20%, etc. so the universe is ranked based on your criteria. In this case, percentage of buy recommendations. This captures a full market cycle back to 2007. The annualized return for the quintile of companies with the least number of buy recommendations (unloved) was +8.2% compared to the quintile with the most buy recommendations at +5.6%.

This trend was even more evident in Canada. Among TSX member companies, those in the top quintile with the most buy ratings actually declined over the past ten years. The least loved quintile enjoyed an annualized return of 3.4% compared to -3.8% for the most loved. Over ten years that really added up (chart 2).

While these are impressive variances in performance that would support not blindly piling into the companies with the most buy ratings, it is far from a hard and fast rule. Looking at each month’s performance the unloved only outperformed the loved quintile just under 60% of the time. Or in other words, 40% of months saw the companies with the most buy recommendations actually outperforming. As a result we have been researching tweaks to this research that would have a more reliable batting average or success rate.

Instead of simply bucketing the companies into quintiles based on their percentage of buy recommendations, we looked for companies that were unloved but started to become less unloved. These are companies that had fewer than 30% of total analyst ratings giving them a buy for at least three months. We call this an unloved company and measured the subsequent performance once the percentage of buys rose over 30% as our signal.

Pages ( 1 of 2 ): 1 2Next »