Introduction In a recent article Prakash, Reside and Smyser (1993) provided a procedure to estimate the global beta.1 This was necessitated due to the fact that the beta, the measure of relative systematic risk, in Sharpe’s (1963) market model is usually estimated in a single market context (e.g. New York Stock Exchange). The rates of return on an individual and market are respectively computed from the price and index movements in the NYSE. However, many stocks are traded in multiple markets with their own price and index movements. Hence using the data from a single market does not fully utilize all the information. Prakash, Reside and Smyser provide a simple way to overcome this. However, in their derivation they made an assumption that the rates of return on the stock and market are measured from their respective mean rates on return on the stock and the market. This assumption, though mathematically correct, may not be intuitively correct because the global rates of return, though unknown, should be universally the same but may in reality be observed differently in different markets. This chapter corrects this anomaly.
