Pairs trading, or statistical arbitrage, is an effective market-neutral trading strategy. Usually fundamental or quantitative analysis is used in order to determine which pairs are suitable for trading. We have previously discussed several pairs selection methods based on quantitative measures such as stock cointegration, correlation, pair distances, etc.
Reference  introduced a new pairs selection method based on the Hurst exponent,
One of the critical steps in [Pairs] Trading is the pairs selection, but not too much attention has been given to the stock universe before pairs selection. In this paper, we have introduced a preselection procedure based on the stocks comovement measure through comovement functions based on comovement studies on physical particle systems. Therefore, portfolios with less volatile stocks have been selected, and it has been observed that, with this new modification, [Pairs] Trading is also profitable in periods of low volatility.
We find the paper interesting. Our comments are as follows,
- We’re of the same opinion that candidate selection is one of the most important steps in pairs trading. It is our understanding that the proposed selection method consists of 2 steps: i-selection of the underlying stocks based on comovements, ii-selection of tradable pairs based on the Hurst exponent.
- The pair selection method based on the Hurst exponent makes sense. Here the Hurst exponent of the weighted difference of the logarithm of prices is calculated, pairs are then selected and trading signals are generated directly using the difference. In contrast, other pairs selection methods make use of indirect measures such as cointegration or correlation.
- We think that it’s worth trying the pair selection method based on the difference of price, instead of the logarithm of price.
- The method for selecting the underlyings (step i) based on price comovements resulted in low-volatility stocks. This does not seem consistent with the empirical observation that pairs trading is considered an implicit short volatility trading strategy.
Regarding the last bullet point, the authors also noted,
However, on high volatility conditions, the strategy does not work as good. A plausible explanation of this phenomenon could be that, during periods with prolonged downward movements in the markets, volatility of the stocks is increased, and the model proposed in this paper is too slow to capture this faster change in the volatility of the preselected stocks.
Regarding the first bullet point, we believe that the pairs selection method can be improved by further performing, e.g., a robustness test in order to minimize divergence risks.
 J. P. Ramos-Requena, M. N. López-García, M. A. Sánchez-Granero, J. E. Trinidad-Segovia, A Cooperative Dynamic Approach to Pairs Trading, Complexity, vol. 2021, Article ID 7152846, 2021.
What's your question? Ask it in the discussion forum
Have an answer to the questions below? Post it here or in the forum
The momentum has been building for years, but this week made it official: the epicenter of the gold mining industry now lies firmly in New York.
Tighter lending puts focus on spread between high yield and Treasuries
Watch Financial Post reporter Stephanie Hughes count down the stories that made headlines