Multi-Scale Time-Series Momentum Portfolios with Multi-Task Learning

Summary

DeepUnifiedMom is a portfolio construction framework that combines time-series momentum signals across multiple time horizons. It uses multi-task learning with a multi-gate mixture-of-experts architecture to form unified portfolios, aiming to capture momentum behavior that can be missed when strategies focus on a narrower set of time frames.

The document reports backtests across equity indexes, fixed income, foreign exchange, and commodities, and says the framework outperformed benchmark models after transaction costs. It provides no specific performance figures, benchmark definitions, sample periods, or details about risk controls in the supplied description. Those omissions make it difficult to assess robustness, exposure to common momentum risks, or live trading feasibility; the reported results should be read as backtest claims rather than evidence of future returns.

Key ideas

• DeepUnifiedMom combines time-series momentum signals from multiple horizons in one portfolio framework.
• The model uses multi-task learning and a multi-gate mixture-of-experts architecture.
• The reported backtests span equity indexes, fixed income, foreign exchange, and commodities.
• The description states that results exceeded benchmark models after transaction costs but gives no performance details.
• Backtest claims alone do not establish robustness or expected live trading performance.

Tags

# 2406.08742


# DeepUnifiedMom: Unified Time-series Momentum Portfolio Construction via Multi-Task Learning with Multi-Gate Mixture of Experts









This paper introduces DeepUnifiedMom, a deep learning framework that enhances portfolio management through a multi-task learning approach and a multi-gate mixture of experts. The essence of DeepUnifiedMom lies in its ability to create unified momentum portfolios that incorporate the dynamics of time series momentum across a spectrum of time frames, a feature often missing in traditional momentum strategies. Our comprehensive backtesting, encompassing diverse asset classes such as equity indexes, fixed income, foreign exchange, and commodities, demonstrates that DeepUnifiedMom consistently outperforms benchmark models, even after factoring in transaction costs. This superior performance underscores DeepUnifiedMom's capability to capture the full spectrum of momentum opportunities within financial markets. The findings highlight DeepUnifiedMom as an effective tool for practitioners looking to exploit the entire range of momentum opportunities. It offers a compelling solution for improving risk-adjusted returns and is a valuable strategy for navigating the complexities of portfolio management.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.