Random walk hypotheses and profitability of momentum based trading rules.

摘要

The main conclusion of this thesis is that for all assets examined here momentum based trading rules yield superior risk adjusted returns compared to buy-and-hold strategies under both weekly and monthly time periods. Furthermore, for weekly data from the CRSP NYSE-AMEX equal-weighted index, the CRSP NASDAQ value-weighted index, the CRSP NASDAQ equal-weighted index, small cap stocks, and certain sectors, technical trading rules outperform the buy-and-hold strategies before adjusting for risk and after adjusting for transaction costs. The consequence of these conclusions is that technical trading rules are useful.; The first chapter of this thesis largely serves to update and refine the results of Lo and MacKinlay's 1988 article on testing Random Walk Hypotheses. Chapter 2 and Chapter 3 address the main results summarized above. The profitability of momentum based trading rules, specifically the filter rule, applied to market indexes, decile portfolios, and sector-sorted portfolios are examined with the help of several risk adjusted measures. Furthermore, the connection between returns to the filter rule and lag one autocorrelation is established. Chapter 4 then serves to complete our assessment of the profitability of technical trading rules, which we do by examining the performance of the MACD indicator and the Moving Average strategy.