Online Portfolio Selection Using Moving Average Reversion

Summary

This article introduces moving average reversion (MAR), a multi-period alternative to the single-period mean-reversion assumption used by some online portfolio strategies. It proposes On-Line Moving Average Reversion (OLMAR), which applies online learning techniques to construct portfolios based on MAR. The motivation is that temporary highs and lows in stock prices may revert over multiple periods, while one-period reversions do not reliably appear in every dataset.

The authors report empirical comparisons on real datasets: OLMAR performed better than existing mean-reversion algorithms, especially where those methods struggled, and ran quickly. These claims are based on the article’s reported experiments, but the supplied description omits dataset identities, benchmark details, transaction costs, and numerical results. The strategy’s performance should therefore be understood as empirical evidence for the tested data, not a guarantee that multi-period reversion will persist or that results will transfer to other markets.

Key ideas

• MAR extends mean-reversion signals across multiple periods instead of relying on a single-period assumption.
• OLMAR applies online learning techniques to use moving average reversion in portfolio selection.
• The method targets settings where single-period mean reversion is unreliable.
• The authors report stronger empirical performance than existing mean-reversion algorithms, particularly on datasets where those methods failed.
• The supplied description reports fast computation but omits dataset and trading-cost details.

Tags

# On-Line Portfolio Selection with Moving Average Reversion


# On-Line Portfolio Selection with Moving Average Reversion









On-line portfolio selection has attracted increasing interests in machine learning and AI communities recently. Empirical evidences show that stock's high and low prices are temporary and stock price relatives are likely to follow the mean reversion phenomenon. While the existing mean reversion strategies are shown to achieve good empirical performance on many real datasets, they often make the single-period mean reversion assumption, which is not always satisfied in some real datasets, leading to poor performance when the assumption does not hold. To overcome the limitation, this article proposes a multiple-period mean reversion, or so-called Moving Average Reversion (MAR), and a new on-line portfolio selection strategy named "On-Line Moving Average Reversion" (OLMAR), which exploits MAR by applying powerful online learning techniques. From our empirical results, we found that OLMAR can overcome the drawback of existing mean reversion algorithms and achieve significantly better results, especially on the datasets where the existing mean reversion algorithms failed. In addition to superior trading performance, OLMAR also runs extremely fast, further supporting its practical applicability to a wide range of applications.

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.