the perspective of all countries, i, and in each country, for all foreign holdings, that is for all j not equal to i. As we see above, the demand from country i, dt> is a sum of three vectors. We consider the first and third components first because they are straightforward. The first vector is just 100 percent weight in the domestic bill of the z'th country, so this does not affect the universal hedging issue. The third vector is all zeros for the home country-that is, country 1-and for other countries is all zeros except two elements: the demand for the home country bill-element n + 1-is (1/X), the demand for the domestic bill of the zth country, element n + i, is -f 1/X). The second vector has three components, a scale factor, (1/xX); a matrix, JiJ'jy1 J'■> which it turns out is the identity matrix minus a constant matrix; and a vector (s - W - j). Recall that the vector s has proportion of market capitalization weights in the first n elements and zeros thereafter. The vector W has zeros in its first n elements and the proportion of wealth in each country thereafter. When wealth proportion equals market capitalization proportion, then the difference, s - W, has equal but opposite values in elements i and n + i. Premultipli-cation by /(/'/)_1 /'> because of its structure, preserves these values. Thus, consideration of only the contribution of s - W would create 100 percent hedging. It is the contributions from other components that lead to less than 100 percent hedging. First consider the vector /. From its definition it turns out that the first n elements are zero. For elements n + j that correspond to bills other than the domestic bill, the value is simply the product of the proportion of wealth in country ; times {-11%). The domestic bill is minus the sum of these values so that the sum of the elements is zero and thus premultiplication by JiJ'JY1 /', because of its structure, preserves these values. Now putting these results together, consider the demands for hedging from the home country. These hedging demands arise only in the second vector, and here the hedging demands are a constant proportion, 1 - (1/X), of the wealth in each country. Finally, consider the demands for hedging in any country i which is not the home country. The third vector affects only the demands for the bill of the home country and the domestic bill. Since the domestic bill does not affect foreign asset hedging, it suffices in considering hedging from the perspective of country i to consider only the demand for the home country bill. All other hedges will remain at the 1 - (1/X) rate seen in the home country. The contribution to home country hedging demand in the third vector is (1/X). The contribution to hedging demand from the vector,;', is minus (1/X) times the sum of weights from the countries other than the home country. Thus, the total demand for hedging of the home country is (1/X) times (1 - Wealth outside the home country), which of course is just (1/X) times the wealth in the home country. Thus, once again the hedging demand is 1 - (1/X) times the wealth in the country. Fischer Black's universal hedging result obtains, and the fraction hedged is the constant, 1 - (1/X). Clearly, the greater X, the risk aversion, is, the larger the fraction of currency risk that is hedged. In practice, these equilibrium equations provide us estimates of risk premiums for various global assets. Let us now examine the risk premiums for a number of assets in an example of a universal hedging equilibrium. We take as assets the largest developed global equity and government bond markets, as well as the