The main purpose of this Master Thesis is the comparison of the different investment strategies based on their annual risk-adjusted returns for the US companies during the study period 2002-2015. Specifically, we intend to determine which investment strategy beats the market, which is determined by the returns, including the dividends paid, of the SP500 index, and which investment strategy works the best by comparing the annual compounded returns and the risk-adjusted Sharpe ratio (Brealey, et al. (2014)). Consistent with Piotroski (2000), we find that the value investing strategy is the best investment strategy because it yields substantial excess returns above the risk-free rate with the moderate risk, and that the market is inefficient. The Composite Index strategy based on the aggregation of F-Score (Piotroski (2000)) and E-Score (Bebchuk et al. (2009)) is the second-best strategy with substantial compounded annual returns and a very good risk profile. This investment strategy aggregates two profitable strategies: value investing strategy (Piotroski (2000)) and E-Index strategy (Bebchuk et al. (2009)). The E-Index strategy (Bebchuk, et al. (2009)) has clearly lost the edge over the market after 2007. We find that the worst and the only loss-generating strategy is the G-Index strategy (Gompers et al. (2003)) with a very high risk. The results of the G-Index and E-Index strategies are consistent with the learning explanation (Gompers et al. (2003), Bebchuk et al. (2013). The Composite Index for the aggregation of F-Score (Piotroski (2000)) and G-Score (Gompers et al. (2003)) is a result of the merge between F-Score (Piotroski (2000)) that yielded substantial returns and G-Index (Gompers, Ishii, Metrick (2003)) that yielded substantial losses. If we look at only the raw performance, we can’t recommend it as an investment strategy. However, this result confirms our hypothesis that Composite Index can act as a “fund-of-funds” of the investment strategies. All our results are robust for all firm sizes and industry classifications. However, our results suffer from certain limitations in our Master Thesis: a sample selection bias, limited data quality in the Datastream, no consideration of the following market mechanisms: actual trading, real-time stock spreads, transaction costs, income taxes, reference-day risk (Dimitrov et al. (2007)).
|Educations||MSc in Accounting, Strategy and Control, (Graduate Programme) Final Thesis|
|Number of pages||112|