Back to Community
Pairs Trading Strategy incorporating regime switching based on volatility

I just stumbled upon this paper on SSRN and am finding it fascinating. Wondering if anyone else has read through it and tried to reproduce the results? Some of the Sharpe Ratio reported are quite impressive.


The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.

2 responses

I am also curious if anyone is able to replicate the results, either on Quantopian or offline, as I have been unsuccessful so far despite making an effort to replicate the exact setup as the paper.

One problem I noticed is that OLS estimation of hedge ratio based on previous day data is often incorrect for the current day. Good cointegration indicated by KPSS test for the previous trading day does not necessarily predict good cointegration for the current day.

IMHO the paper has omitted some important details such as the Bollinger band parameters and entry/exit multipliers for the two regimes. It also did not mention how gaps between trading days are handled in terms of rolling stddev calculation, which may be a significant problem because the algorithm uses only 20-minute bar closing prices, so the Bollinger band would likely need to span intraday data on multiple days.

It also did not mention how gaps between trading days are handled in terms of rolling stddev calculation, which may be a significant problem

This is such a proverbial "third rail" when doing intraday calculations. If anyone has any derivations for volatility estimators and filters for inhomogeneous time series, I would love to get my hands on them. With some assumptions about the underlying process, it must be totally doable, but I haven't worked it through yet.