Data-Driven Investigation into Anomaly Trading Strategies: Evidence with Econometrics

Abstract

This chapter used empirical data from five developed markets and five emerging markets to perform an examination of anomalies using common financial economic approaches along with more innovative econometric models. Of the methodologies used to test for anomalies, the data-driven panel and quantile regressions were empirically found to be better suited over the traditionally common approaches to describe the non-linear, switching behavior of the anomalies. In the developed markets, the statistically significant small firms (size) had the highest average returns. In the developing markets, the lower price-to-earnings (P/E) ratios (value) had the highest average returns. In addition, the research found (1) a small country effect, (2) sales had a negative relationship with returns, and (3) a lower (higher) book-to-market (B/M) was associated with higher returns in the developed (developing) markets, indicating investors received a higher premium for growth (value) equities. The semi-strong form of the efficient market hypothesis was also found to be violated. The anomalies’ behavior varied between sorted portfolios, industries, and developed to emerging markets; though it was found to be consistent through time (not disrupted by bear or bull markets).

Citation

French, J. (2019), "Data-Driven Investigation into Anomaly Trading Strategies: Evidence with Econometrics", Choi, J.J. and Ozkan, B. (Ed.) Disruptive Innovation in Business and Finance in the Digital World (International Finance Review, Vol. 20), Emerald Publishing Limited, Leeds, pp. 221-245. https://doi.org/10.1108/S1569-376720190000020021