The Power of Long-Term Profitability and Compounding Effects: Empirical Evidence from the US American Market

Sammanfattning: For the US American market, we examine whether a ten-year non-rebalancing investment strategy that picks stocks largely based on historical return on equity and the degree to which a company reinvests earnings can beat the market. Delisted companies are sold and the proceeds reinvested along dividends at the market return, which we approximate by the Russell 3000 TR index. For ten-year periods starting between 1998 and 2005, we observe outperformance over the market in all periods - all but one being statistically significant at the 95% level or higher. The strategy performs especially well when the market performs poorly. As we exclude financial stocks from our portfolios, we argue that part of the outperformance in the periods finishing during or just after the financial crisis can be attributed to this decision. Assuming the existence of a small company size effect, part of the performance of our portfolios could be owed to their predominant picking of small capitalisation companies. Yet, we also observe alpha after applying factor analysis models which show the highest loadings on the size (SMB) factor and to a lesser extent on the value (HML) factor. Loadings on the profitability (RMW), investment (CMA) and momentum (MOM) factors are low or ambiguous.