Many Canadian plan sponsors use a two-step policy to determine the hedging policy; first the strategic mix is determined, with how to handle the currency exposure a secondary consideration. A two-step policy is better than none at all, but it does not produce the optimal hedging policy. The result of the two-step process tends to be a lower allocation to international assets and a lower amount of hedging.

A more systematic approach is available. Use a single-step process whereby the asset mix and the hedging policy are determined simultaneously by treating hedged and unhedged assets as separate asset classes.

Consider a typical Canadian plan sponsor, with a strategic asset allocation of 40% Canadian bonds, 30% Canadian equity and a 30% allocation in international equities, of which half is U.S. equity and half is EAFE.

Table 1 shows the sources of international investment risk for a Canadian investor. The unhedged equity return is comprised of the asset's return in a particular foreign country plus the currency return. Since there is considerable difference between U.S. dollar and EAFE currency behaviour, the currency exposure of each should be examined separately.

Table 1 shows that over the 22-year period 1978-2000, the risk of U.S. unhedged assets was 14.5%, actually slightly less than the risk of hedged U.S. assets. This is due to the low correlation of USD currency returns with unhedged USD assets, as well as the relatively low volatility of the Canadian/U.S. dollar exchange rate. The risk of holding EAFE unhedged assets over this period was 16.6%, 220 basis points more than on hedged EAFE assets.

Optimal hedge ratio analysis is a type of cost-benefit analysis that uses a mean variance framework to determine how much currency risk you are willing to bear. It assumes that the expected return from currency is zero, therefore the main benefit from the optimal hedging analysis is risk reduction. The cost side of the analysis is recognizing that hedging is not free.

Determining the hedge ratio is an optimization problem. Inputs into the model include the investor's risk tolerance, an estimate of hedging costs, the proportion of international assets in the portfolio and a good estimate of the historical covariance of all asset returns in the portfolio, including currency returns.

When an analysis is completed for the representative Canadian plan in our example, the optimal hedge ratio for the U.S. equity is zero, i.e., it is optimal not to hedge the U.S. dollar exposure.

The result was driven by the small negative correlation of U.S. dollar currency returns with Canadian stocks and bonds, the low volatility of the Canadian/U.S. dollar exchange rate relative to the return volatility of domestic and international assets and the small correlation of USD currency returns with unhedged USD assets.

The model produces an optimal hedge ratio for the EAFE assets of about 30%, due to the small negative correlation of EAFE currency returns with Canadian assets, high volatility of the Canadian/EAFE exchange rate relative to the return volatility of domestic and international assets and the large positive correlation of EAFE currency returns with unhedged EAFE assets.

The recommendation for our example is therefore to leave the U.S. assets unhedged and adopt a 30% hedge ratio for the EAFE assets. These ratios should be reviewed on a regular basis as your assumptions about the stability of the environment change.