We propose DeePM (Deep Portfolio Manager), a structured deep-learning macro portfolio manager trained end-to-end to maximize a robust, risk-adjusted utility. DeePM addresses three fundamental challenges in financial learning: (1) it resolves the asynchronous “ragged filtration” problem via a Directed Delay (Causal Sieve) mechanism that prioritizes causal impulse-response learning over information freshness; (2) it combats low signal-to-noise ratios via a Macroeconomic Graph Prior, regularizing cross-asset dependence according to economic first principles; and (3) it optimizes a distributionally robust objective where a smooth worst-window penalty serves as a differentiable proxy for Entropic Value-at-Risk (EVaR) - a window-robust utility encouraging strong performance in the most adverse historical subperiods.

A large-scale benchmark of modern deep learning architectures for financial time-series forecasting and position sizing, with a focus on risk-adjusted performance.

We propose a novel time-series trend-following forecaster that is able to quickly adapt to new market conditions, referred to as regimes. We leverage recent developments from the deep learning community and use few-shot learning. We propose the Cross Attentive Time-Series Trend Network - X-Trend - which takes positions attending over a context set of financial time-series regimes. X-Trend transfers trends from similar patterns in the context set to make predictions and take positions for a new distinct target regime. X-Trend can also take zero-shot positions on novel unseen financial assets. It both forecasts next-day prices and outputs a trading signal. Furthermore, the cross-attention mechanism allows us to interpret the relationship between forecasts and patterns in the context set.

We introduce the Momentum Transformer, an attention-based deep learning architecture which outperforms benchmark momentum and mean-reversion trading strategies. Unlike state-of-the-art Long Short-Term Memory (LSTM) architectures, which are sequential in nature, the attention mechanism provides our architecture with a direct connection to all previous time-steps. Our architecture enables us to learn longer-term dependencies, improves performance when considering returns net of transaction costs and naturally adapts to new market regimes, such as during the SARS-CoV-2 crisis. The Momentum Transformer is inherently interpretable, providing us with greater insights into our deep learning momentum trading strategy, including how it blends different classical strategies and the past time-steps which are of the greatest significance to the model.

To improve the response of momentum strategies to regime change, we introduce a novel approach, where we insert an online change-point detection (CPD) module into a Deep Momentum Network (DMN) pipeline, which uses an LSTM deep-learning architecture to simultaneously learn both trend estimation and position sizing. Our model is able to optimise the way in which it balances 1) a slow momentum strategy which exploits persisting trends, but does not overreact to localised price moves, and 2) a fast mean-reversion strategy regime by quickly flipping its position, then swapping it back again to exploit localised price moves.