Building of F-Score-Like Models on the Example of the Polish Stock Market

Abstract

Theoretical background: Value investing is one of the most popular investing approaches. In their frame, there could be a high B/M investing strategy identified. F-Score, developed by Piotroski, is a scoring model applied to the sample of high B/M entities. Its purpose is to select companies with strong financial foundations and buy their shares for the investment portfolio to generate positive market-adjusted returns in the following periods. The effectiveness of this model was mostly empirically confirmed, especially regarding developed markets. Purpose of the article: The main aim of the paper was to build F-Score-like models based on the data from the Polish stock market. The main hypothesis concerned the higher effectiveness of such models than F-Score, as the specificity of a given market should result in a better fit to the data. Research methods: Building of the models based on the discriminant analysis and formation of the investment portfolios based on the indications of these models as well as F-Score. Finally, backtesting of the portfolios built to assess their effectiveness. The sample covered most of the Polish-listed companies. The period taken into account was 2012–2022. Main findings: Models built (X-Score and Y-Score) were less efficient than F-Score. Moreover, they led to generating negative rates of return (both raw and market-adjusted). On the other hand, using of F-Score for the analyzed period seems to be purposeful due to the 1.35% mean annual market-adjusted return generated. Apart from the scoring models analyzed, the research partially confirmed the advisability of using a high B/M investing strategy. Generally, the results obtained are in line with the findings of most of other authors –regarding the F-Score effectiveness. However, an approach based on Mohanram’s idea – using the differences between absolute values of a given variable and median from the sample – proved to be inadequate in the Polish stock market.