pairs trading strategy and statistical arbitrage

The North American Journal of Economics and Finance

Highlights

•: Presents a simultaneous multiple pair-trading strategy through supervised encyclopaedism.
•: Diversifies portfolios via a multi-asset twin-trading basket and GARCH volatility.
•: Trailer truck-parametric tolerance limits are predictive functions that find trading signals.
•: Select depress/upper tolerance limits from the largest discrepancy return open.
•: Two annual testing periods' returns provide strong lucrativeness.

Abstract

Pair trading is a wide used market neutral strategy and also a statistical arbitrage method acting that allows investors to take a position in 2 assets with connatural trends in their historical information in order to gain low-risk of infection net. Combining both "diversification" and "pair trading", this study proposes a applied mathematics learning method to research the most promising pair among triple pair assets for for each one trading clock time. We incorporate estimated volatility into permissiveness limits equally a prognosticative function for finding buying and selling signals in order to capitalize on marketplace inefficiencies. One-step-ahead excitability prediction follows either the exponentially weighted oncoming ordinary (EWMA) method acting operating theatre the GARCH model. The study selects five artificial intelligence service (Three-toed sloth) stocks in the U.S. equities market to target profitableness through the proposed scheme with a rolling window training approach over two annual testing periods from April 2022 to March 2022. We recommend that conservative investors use $p$ -content at 95%, which is less adventurous and can generate positive excess profits. The idea slow this strategy is to help investments be more diversified and also more profitable.

Keywords

Exponentially weighted vibratory normal (EWMA) model

GARCH model

Multiple pair assets

Prognosticative function

Tolerance separation

Volatility presage

View full text edition