Here is some interesting data I have compiled this morning concerning the efficacy of stop losses in the Canadian market.

High to Low range within each month of six high volume stocks.  Sample period Jan 1/09 to Mar. 31/10 (15 months)
Move:      5%     10%    15%
BNS    15/15    10/15    6/15       (Months stock fell more than target %)

RY    15/15    7/15    4/15

SU    15/15    14/15    9/15

IMN    15/15    12/15    11/15

TRP    12/15    4/15    0/15

BCE    11/15    1/15    1/15

In other words, if you had 5% stop losses on these six stocks, you would have been sold out of four of them in each and every one of the fifteen months in the sample, and you would have been sold out of all of them in eleven of the fifteen months; in total, you would have been sold out 83 out of 90 months (six stocks x 15 months).

With a 10% stop loss, you would have been sold out 48 of 90 times, and with a 15% stop loss, 31/90 times.

This is just measuring intra-month movement.  In effect, a stop loss really measures movement from purchase to the date on which the stop price is reached.  Clearly every stock would have been sold out long before the end of the fifteen month sample period, and if repurchased, would have been sold again.  And again – almost every month for almost every stock.  Bear in mind that during this period the TSX rose from 8,988 to 12,038 or 33.9%.

By using even a 10% stop loss, you would have certainly been sold out of Inmet (up 201%) Suncor (up 39%), and Bank of Nova Scotia (up 53%).

In my view, the stop loss is a trading tool, not an investing tool.  It is a device to control short term market risk, not long term investment risk.  If a stock is purchased on the basis of conviction as a long term investment, the stock loss becomes a way of locking in losses and missing out gains.  The fact is that the market has volatility, and the investor has to live with it.