Investment Factor Timing: Harvesting The Low-Risk Anomaly Using Artiﬁcial Neural Networks
35 Pages Posted: 27 Apr 2020 Last revised: 5 May 2020
Date Written: April 18, 2020
We perform investment factor timing based on risk forecasts in reference to the low-risk anomaly. Among various risk measures, we ﬁnd downside deviation most suited for this task. We apply Long Short Term Memory Artiﬁcial Neural Networks (LSTM ANNs) to model the relationship between macro-economic as well as ﬁnancial market data and the downside deviation of factors. The LSTM ANNs allow for complex, non-linear long-term dependencies. We use LSTM-based forecasts to select high- and low-risk factors to set up an investment strategy. The strategy succeeds in diﬀerentiating positive from negative yielding factors and an accordingly constructed investment strategy outperforms every factor individually as well as a GARCH benchmark model.
Keywords: Factor Investing, Neural Networks, LSTM, Low-Risk Anomaly
JEL Classification: C58, C63, E37, G11, G17
Suggested Citation: Suggested Citation