Testing efficient market hypothesis for the dollar–sterling gold standard exchange rate 1890–1906: MLE with double truncation

This paper tests an efficient market hypothesis for the dollar–sterling exchange rate under the classical gold standard in the period 1890–1906. The test is based on maximum likelihood estimation of doubly truncated time series. The results show that weak-form efficiency can not be rejected if truncation by the gold points is incorporated. The results, however, are sensitive to the gold points estimates used.