The investment universe consists of EM equities limiting our investment universe to big stocks only. Buy/hold long-only strategy is using ML algorithm that computes return prediction (eq. (15)) from 36 firm-level characteristics (available on Refintiv platforms; the willing reader can look them up in Appendix B – Characteristics definition) using the ENS model accounting for transaction costs (eq. (14)).
ASSET CLASS: stocks | REGION: Emerging Markets | FREQUENCY:
Monthly | MARKET: equities | KEYWORD: Machine Learning, Emerging Markets
I. STRATEGY IN A NUTSHELL
Invest in top-decile emerging market stocks using machine learning models, including trees and neural networks, with monthly rebalancing and hedging via a broad EM index
II. ECONOMIC RATIONALE
Machine learning captures underreaction in emerging markets, utilizes data efficiently, and adapts faster than linear models, generating superior risk-adjusted returns over longer holding periods.
III. SOURCE PAPER
Machine Learning and The Cross-Section of Emerging Market Stock Returns [Click to Open PDF]
Matthias X. Hanauer, Technische Universität München (TUM); Tobias Kalsbach, Robeco Asset Management
<Abstract>
This paper compares various machine learning models to predict the cross-section of emerging market stock returns. We document that allowing for non-linearities and interactions leads to economically and statistically superior out-of-sample returns compared to traditional linear models. Although we find that both linear and machine learning models show higher predictability for stocks associated with higher limits to arbitrage, we also show that this effect is less pronounced for non-linear models. Furthermore, significant net returns can be achieved when accounting for transaction costs, short-selling constraints, and limiting our investment universe to big stocks only
IV. BACKTEST PERFORMANCE
| Annualised Return | 2.8% |
| Volatility | 5.32% |
| Beta | N/A |
| Sharpe Ratio | 0.53 |
| Sortino Ratio | N/A |
| Maximum Drawdown | N/A |
| Win Rate | N/A |
