A popular research area has been investigating the statistical relationship between financialratios and stock returns since ratios are perceived as useful in forecasting future rates of returns(Barnes, 1987). Literature on stock predictability has evolved over the past few decades. Initialevidence that market returns are predictable was questioned by later studies that found suchpredictions did not hold in subsamples. Nonetheless, once methodological corrections have beenmade, some financial ratios, particularly dividend yield, earnings per share, and book to marketvalue of equity have been found to consistently forecast market returns for long periods (Lewellen,2004).1A challenge on those studies is the selection of financial ratios to test since ratios tend tocontain overlapping information