Paying attention to returns: Forecasting returns as nonlinear AR time-varying drifts using multihead attention and Singular Spectrum Analysis

Adriel Wáng

Chaos, Solitons & Fractals, 2025, vol. 199, issue P3

Abstract: Many believe that informational efficient markets make predicting returns impossible, but methods like Moving Averages (MA) and Exponential Smoothing show it is possible. This fact opens the door for applying signal processing techniques to the theory of financial mathematics. Here, we use Singular Spectrum Analysis (SSA) to predict price movements driven by Lévy Stable Stochastic Differential Equations. In a study of 4-hour cryptocurrency returns, we show that SSA improves the forecast accuracy of OHLC/4 data. Using Bitcoin (BTC), we show how a multihead attention model can predict prices up to 24 h (6 steps) ahead, such that attention and spectral filtering work well together.

Keywords: Levy Stable; Singular Spectrum Analysis; Signal processing (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0960077925008069
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:chsofr:v:199:y:2025:i:p3:s0960077925008069

DOI: 10.1016/j.chaos.2025.116793

Access Statistics for this article

Chaos, Solitons & Fractals is currently edited by Stefano Boccaletti and Stelios Bekiros

More articles in Chaos, Solitons & Fractals from Elsevier
Bibliographic data for series maintained by Thayer, Thomas R. ().