Predicting industry stock returns : empirical evidence from NYSE U.S. 100 Index components
1University of Oulu, Oulu Business School, Department of Finance, Finance
|Online Access:||PDF Full Text (PDF, 1.7 MB)|
|Persistent link:|| http://urn.fi/URN:NBN:fi:oulu-201511122129
|Publish Date:|| 2015-11-16
|Thesis type:||Master's thesis
This thesis paper test for stock return predictability in the largest and most comprehensive industry data set analyzed so far, using three common forecasting variables: the earnings-price (EP), dividend-price (DP) and book-price (BP) ratios and a new (proposed) predictor variable — the residual equity-price (REP) ratio. Considering the three common forecasting variables, each has price as the denominator, which explicitly represent the value an investor pays to own a share of stock. However, the numerators of these ratios do not reflect the explicit gain or loss to the investor for owning the share of stocks. That is, these three common forecasting variables do not reflect the most vital characteristic of a shareholder being a residual-owner. Thus, evidence of return predictability using these ratios could be interpreted as clueless and as a result of chance. The proposed predictor variable — the REP ratio, provide antidote to this distortion.
The data contain over 10,000 quarterly observations from 88 companies listed on the New York stock exchange. The companies (stocks) are pooled into a panel of industries according to the NYSE U.S. 100 Index classifications. I conducted pooled regressions — employing the methodology of Hjalmarsson (2010), as well as time-series regressions for individual companies.
The empirical results indicate that the three common financial ratios (EP, DP and BP ratios), as well as the REP ratio are fairly robust predictors of quarterly excess stock returns, in industry dataset. The null of no predictability is clearly rejected in the pooled regression for a number of industry panels, as well as in a number of individual company level time-series regressions. However, considering their out-of-sample performance — the REP ratio tremendously outperforms the historical average forecasts, as well as the forecasts based on the EP, DP and BP ratios. The evidences are generally inline with those found by Lewellen (2004), Campbell and Yogo (2006), and Cochrane (2008), with U.S. data (index data); and Hjalmarsson (2010) with international data.
© David Mensah, 2015. This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.