Random Walks Near Dollar Par

Thursday's quiet market action and a dollar trading near par with the US dollar produced some good examples of how currency fluctuation affects stock prices. All we need to do is compare prices of stocks that with dual listings on both the TSX and one of the US markets. Today CAD traded between about 0.986 and 0.991 to USD during market hours, or about 0.9% to 1.4% below par. We should expect that TSX-traded stocks should have traded about that much below their US counterparts during the day.

I picked two companies for this comparison: Dragonwave (an Ottawa company) and Suncor. Suncor is the more liquid issue, trading many times more shares per day than Dragonwave. Since trades are discrete events (price is not a continuous function of time), pricing errors should be more common for Dragonwave. The two comparison charts for March 18, 2010 are shown on the right.

One thing that immediately pops out is how much more closely the stocks traded than the currency difference would indicate. The variation in all the stocks I monitored today were even greater, with some in near lock-step all day (0%), while others showed persistent differences of 2%. Interestingly, I saw better tracking in some low-priced shares, although I expected more variation since prices on the TSX on any stock over $1 are never fractional cents, while this is common on NASDAQ, which would add in variability due to rounding. There are of course many more reasons for the generally poor tracking of dual-listed stocks, though it is perhaps mostly due to order book fluctuations and accumulations on the bid and ask sides of the market which can cause episodic minor volume avalanches when a sale moves the market price.

It is interesting to speculate whether these pricing errors can be exploited by, say, buying when the price spread is high and selling a few minutes later when the gap is closed. Unfortunately this does not look like a profitable strategy. When you consider slippage risks and trading commissions, and the intensive effort involved, I have grave doubts whether this could ever be a dependable strategy. Still, it is an interesting topic to think about.