Combining Changepoint Detection with Momentum and Fast Reversion
Summary
The strategy combines online changepoint detection with a Deep Momentum Network built around an LSTM. The network learns trend estimates and position sizes, while the detection module supplies a suspected turning point and a measure of its severity. The system can then balance slower time-series momentum, which aims to stay with persistent trends, against faster mean reversion, which can reverse exposure around localized moves. This is designed to address momentum losses near abrupt regime changes.
Backtests covering 1995–2020 report a one-third improvement in Sharpe ratio after adding changepoint detection, with an approximately two-thirds boost during 2015–2020. The reported benefit is strongest during nonstationary periods. These figures are backtest results, not evidence of live trading performance; the excerpt gives no details on markets, transaction costs, validation design, or robustness checks. The approach therefore suggests a way to adapt momentum exposure, while leaving important implementation and generalization questions open.
Key ideas
- Online changepoint detection is added to an LSTM-based momentum and position-sizing model.
- A severity score helps the model vary its response to detected changes in market conditions.
- The strategy blends persistent-trend exposure with faster mean-reversion behavior.
- Backtests report higher Sharpe ratios after adding the detection module, especially in recent tested years.
- The excerpt omits market, cost, and robustness details needed to assess live applicability.
Tags
Full text
# Slow Momentum with Fast Reversion: A Trading Strategy Using Deep Learning and Changepoint Detection # Slow Momentum with Fast Reversion: A Trading Strategy Using Deep Learning and Changepoint Detection Momentum strategies are an important part of alternative investments and are at the heart of commodity trading advisors (CTAs). These strategies have, however, been found to have difficulties adjusting to rapid changes in market conditions, such as during the 2020 market crash. In particular, immediately after momentum turning points, where a trend reverses from an uptrend (downtrend) to a downtrend (uptrend), time-series momentum (TSMOM) strategies are prone to making bad bets. To improve the response to regime change, we introduce a novel approach, where we insert an online changepoint detection (CPD) module into a Deep Momentum Network (DMN) [1904.04912] pipeline, which uses an LSTM deep-learning architecture to simultaneously learn both trend estimation and position sizing. Furthermore, 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. Our CPD module outputs a changepoint location and severity score, allowing our model to learn to respond to varying degrees of disequilibrium, or smaller and more localised changepoints, in a data driven manner. Back-testing our model over the period 1995-2020, the addition of the CPD module leads to an improvement in Sharpe ratio of one-third. The module is especially beneficial in periods of significant nonstationarity, and in particular, over the most recent years tested (2015-2020) the performance boost is approximately two-thirds. This is interesting as traditional momentum strategies have been underperforming in this period.
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.