When we explain the case for currency hedging to our clients, we often find ourselves running through all of the commonly cited arguments for leaving currency exposures unhedged, and explaining the flaws in each of them. Recently, a new argument for doing nothing about currency has emerged: isn’t it pointless to hedge currency exposures that arise from investing in multinational companies if these companies themselves have exposures to many different currencies?
Exxon Mobil is a good example of such a multinational company. Exxon Mobil is listed in the US, so if you measure the currency exposures in your international equity portfolio by reference to the country each company is listed in, an investment in Exxon Mobil would be counted as exposure to the US dollar. However, Exxon Mobil is a multinational company that does business all around the world, and derives more than half of its sales from outside of the US. An investment in Exxon Mobil carries with it an underlying exposure to a basket of currencies, which might even include your base currency. Isn’t it pointless to try to hedge the currency exposure you get from investing in Exxon Mobil just by hedging from the US dollar back to your base currency?
The answer is yes, it would be, if your entire international equity portfolio was invested in Exxon Mobil. However there are two other factors to consider.
Firstly, if your international equity portfolio is spread across a broadly diversified basket of companies, there will be some degree of averaging out across the portfolio. On an overall basis, the currency exposures that you calculate by reference to the country each company is listed in will be broadly similar to, albeit not exactly the same as, the aggregate of the underlying currency exposures of all of the companies in your portfolio.
Secondly, even though the underlying foreign currency exposure in your international equity portfolio will be less than the full value of the portfolio (because some of the underlying currency exposure from the companies in that portfolio will be denominated in your base currency), the converse applies to your domestic equity portfolio. Some of the companies held in your domestic equity portfolio will be multinationals, so you will have some additional underlying foreign currency exposure deriving from these holdings. In fact, for most pension plans, the additional foreign currency exposure deriving from the domestic equity holdings will exceed the additional domestic currency exposure deriving from the international equity holdings. This means that the overall underlying foreign currency exposure at the total plan level will be greater than the plan’s allocation to international equities. If anything, this strengthens rather than weakens the case for hedging at least part of your nominal foreign currency exposure back to your base currency.
Wouldn’t it be a better idea then to measure the underlying currency exposures across the entire equity portfolio, including both the domestic and international equity portions, and use these figures instead of the nominal currency exposures as a starting point for currency hedging? It would be if it was practical to do so, but unfortunately it isn’t. Companies don’t report their underlying currency exposures in a consistent, user-friendly format, and in any case these exposures are changing constantly as their business mixes and hedging policies change. That’s why we have to use the nominal currency exposure of our international equity portfolios, calculated by reference to which country each company is listed in, as a proxy for our underlying currency exposures as a starting point for currency hedging.
It is possible to test whether hedging currency exposures arising from investment in multinational companies makes sense by reference to empirical data. Firstly, let’s assume that a UK pension plan has a benchmark asset mix of 40% UK equities, 40% global equities (unhedged), and 20% UK fixed income. Over the period from 1 January 1992 to 30 June 2001 (the longest period for which we have return data for multinational companies), the annualised return for this benchmark would have been 11.16% pa, and the risk level (defined here as the annualised standard deviation of the quarterly returns) would have been 11.14%. Table 1 shows how the risk level would have been affected by varying degrees of hedging of the foreign currency exposures back to sterling.
The key point that Table 1 illustrates is that currency hedging reduces risk. In this example, fully hedging the foreign currency exposure would have reduced the risk level by 0.57%. This is equivalent to the degree of risk reduction that you would have achieved if you had left the currency exposure unhedged, and had instead reduced the equity exposure from 80 to 75% and increased the bond exposure from 20 to 25%. Alternatively, you could have increased the equity exposure from 80 to 85%, and hedged the foreign currency exposure, with no change in risk level. If you assume that equities will outperform bonds by 4% pa over the long run, this would generate a 0.2% pa increase in long term expected return at the total plan level.
Now let’s assume that the international equity component of this benchmark was instead composed solely of multinational companies (defined here as the 500 companies with the highest ratios of foreign sales to total sales out of the world’s 1,500 largest companies).
Table 2 shows the risk level figures corresponding to those in Table 1 under this scenario.
The figures in Table 2 show that, even if the international equity portfolio was invested solely in multinational companies, hedging part or all of the currency exposure would still have reduced the risk level. The risk reduction would not have been as great, but it still would have been worth pursuing – particularly when one considers that it is reasonable to expect that currency hedging will have no impact on longer term returns.
In summary, it would not be practical to hedge currency exposures based on the true underlying currency exposures associated with each company that you invest in. However, the nominal currency exposure calculated by reference to the country each company is listed in is a reasonable proxy for underlying currency exposures, and hedging on this basis should reduce risk even if your international equity portfolio consists entirely of multinational companies.
Bill Muysken is head of manager research – global at William M Mercer Investment Consulting in London

Topics