Financial ratios and stock returns reappraised through a topological data analysis lens

Firm financials are well-established predictors of stock returns, being the basis for both traditional econometric, and growing Machine Learning, asset pricing literature. Employing topological data analysis ball mapper (TDABM), we revisit association between seven most commonly studied financial ratios returns. Upon outlining methodology to finance literature, this paper offers three key contributions study pricing. Firstly, characteristic space is visualised showcase non-monotonic relationships in multiple dimensions that were as yet unseen. Secondly, means through which neural networks random forest regressions fit returns also visualised, showing where Learning contributing understanding. Finally, an initial application TDABM segmentation cross-section posited, with significant abnormal identified. Collectively these expositions signpost value researchers practitioners alike. The scope benefit limited only by availability information analyst.