Machine Learning Versus Fundamental Investment Analysis: A Meta-Analysis

The performance of machine learning algorithms in financial asset pricing is assessed through a meta-analysis in which the results of previous research is combined. The meta-analysis consisted of a research sample of 63 research papers on the application of machine learning algorithms in stock pricing, option pricing, and bond pricing. The 63 research papers aided in accepting the three sub-hypotheses, that machine learning algorithms are associated with higher pricing performance than traditional financial asset pricing tools, and the main hypothesis, that machine learning algorithms outperform fundamental investment analysis.