Pairs trading based on wavelet decomposition

In the current research,wavelet analysis is used to analyze the time series of prices in a pair of assets into general and detailed time series, and the property of collocation between different and corresponding levels of analysis of two series is checked in order to find collinear pairs at different levels of analysis. And then its profitability is examined. In this research, the profitability of the pair trading system based on wavelet analysis was investigated on 14 indices of the Tehran Stock Exchange betwee 2013-2022. The results show that for the second level of detail in the wavelet analysis, the results are quite impressive and the number of trading positions is more than doubled, the daily return is increased to four times and the Sharpe ratio is also increased to about two times. The system formed based on the first level of detail also has a better profitable performance than the normal aggregation, and the performance of the third level of detail is within the limits of aggregation. In addition, the average duration of the transaction also shows significant decrease in the first and second levels. Profitability performance at the level of general series is generally weaker than the aggregate.