Skip to main content
number cruncher

What are we looking for?

To see whether it is true that you should avoid stocks that are heavily shorted.

More about today's screen

This is very similar to a series of screens we ran in last week's Number Crunchers about the favourite stocks currently among short sellers ( Short sellers are betting that a stock will fall.

We'll ask for help from Morningstar CPMS. Jamie Hynes, senior consultant at CPMS, ran a screen that looks for the most heavily shorted stocks among the 680 largest Canadian stocks it follows. He determined the most heavily shorted stocks by calculating number of days required to cover short interest and he made this calculation by taking the latest short interest and dividing it by daily average volume over last 200 days.

The attached table ranks stocks by largest short interest.

More about CPMS

CPMS is an equity research and portfolio analysis firm owned by Morningstar Canada. It maintains a database of about 680 of the largest and more liquid Canadian stocks, plus another 2,100 U.S. stocks, and spends a lot of time adjusting for unusual accounting items in each company's quarterly results to make sure screens can perform correctly.

What did we find out?

Mr. Hynes ran a test on this screen back to September, 2003, when CPMS started gathering short data. He created an equal-weight portfolio of the 15 most-shorted stocks and re-selected the names every month. He found out that this strategy has produced an annual return of 5 per cent since 2003, versus 11.1 per cent for the S&P/TSX composite total return index.

"This includes a 60-per-cent loss in 2008," Mr. Hynes pointed out. "Since then, the screen has bounced back strong with a 66-per-cent return in 2009 and a 41-per-cent return in 2010, but the value of the portfolio is still at pre-2008 levels (a 60-per-cent loss requires a 166-per-cent gain just to break even).

Next week, Mr. Hynes will take this screen a step further and show how you can find heavily shorted stocks that can beat the benchmark.