Pairs trading (sometimes also known as statistical arbitrage) is a very popular trading strategy between traders, and it has become a favorite strategy for investigation by financial academics. The most well-known variant is stock’s pairs trading where the trader buys and simultaneously sells two stocks (that follow each other) when they diverge from the normal synchronized moves.
The equity universe is broad and therefore, it is time-consuming to look for pairs that are correlated or cointegrated (aka. they move together). ETFs provide a much smaller universe, and they have one great advantage (compared to stocks) – the ETF is a portfolio of stocks and is, therefore, more resilient to unexpected news related to a single stock (this often breaks a promising convergence in stock’s pairs). Pairs trading on country ETFs is, therefore, an easy and promising version of a pairs trading strategy.
As prices in a pair of ETFs were closely cointegrated in the past, there is a high probability those two securities share common sources of fundamental return correlations. A temporary shock could move one ETF out of the common price band. This presents a statistical arbitrage opportunity. The universe of pairs is continuously updated which ensures that pairs which no longer move in synchronicity are removed from trading and only pairs with a high probability of convergence remain.
Confidence in anomaly's validity
Backtest period from source paper
Notes to Confidence in Anomaly's Validity
Period of Rebalancing
Notes to Indicative Performance
per annum, annualized (geometrically) daily return 0,075% (data from table 3B – one day waiting period, 5 top pairs selected)
Notes to Period of Rebalancing
Number of Traded Instruments
Notes to Estimated Volatility
estimated annualized daily volatility (data from table 3B – one day waiting period, 5 top pairs selected)
Notes to Number of Traded Instruments
Moderately complex strategy
Notes to Maximum drawdown
Notes to Complexity Evaluation
Simple trading strategy
The investment universe consists of 22 international ETFs. A normalized cumulative total return index is created for each ETF (dividends included) and the starting price during the formation period is set to $1 (price normalization). The selection of pairs is made after a 120 day formation period. Pair’s distance for all ETF pairs is calculated as the sum of squared deviations between two normalized price series. The top 5 pairs with the smallest distance are used in the subsequent 20 day trading period. The strategy is monitored daily, and trade is opened when the divergence between the pairs exceeds 0.5x the historical standard deviation. Investors go long on the undervalued ETF and short on the overvalued ETF. The trade is exited if a pair converges or after 20 days (if the pair does not converge within the next 20 business days). Pairs are weighted equally and the portfolio is rebalanced on a daily basis.
Thomakos, Wang, Schizas: Pairs Trading on International ETFs
Pairs trading is a popular market-neutral trading strategy among finance practitioners that has been recently evaluated for U.S. stocks (Gatev, Goetzmann, and Rouwenhorst 2006). In this paper, we examine the pairs trading performance as well as the source of the profitability using international exchange traded funds (ETFs). Our results suggest that there are returns from pairs trading on international ETFs. Those returns could partially explained by economic factors.
Strategy's implementation in QuantConnect's framework (chart+statistics+code)
Miika Sipila: Algorithmic Pairs Trading: Empirical Investigation of Exchange Traded Funds
The objective of this thesis is to study whether the algorithmic pairs trading with Exchange Traded Funds (ETFs) generates abnormal return. Particularly, I firstly study whether the trading strategy used in this thesis generates higher return than the benchmark index MSCI World and secondly even higher return than stocks. The dataset includes over 66,000 possible pairs of ETFs worldwide from 2004 to 2012. In addition, I use the empirical results from the relevant papers in comparison. To test the hypothesis, I first apply cointegration tests to identify ETFs to be used in pairs trading strategies. Subsequently, I select ETF pairs to compose a pairs trading portfolio based on profitability and finally compare the results to the benchmark index and the empirical results of the relevant papers. The empirical results of this thesis show that pairs trading with ETFs generate significant abnormal return with low volatility from the eight year trading period compared to the benchmark index as well as stocks traded with pairs trading strategy. The cumulate net profit is 105.43% and an annual abnormal return of 27.29% and with volatility of 10.57%. Furthermore, the results confirmed market neutrality with no significant correlation with MSCI World index.
Leung, Li: Optimal Mean Reversion Trading with Transaction Costs and Stop-Loss Exit
Motivated by the industry practice of pairs trading, we study the optimal timing strategies for trading a mean-reverting price spread. An optimal double stopping problem is formulated to analyze the timing to start and subsequently liquidate the position subject to transaction costs. Modeling the price spread by an Ornstein-Uhlenbeck process, we apply a probabilistic methodology and rigorously derive the optimal price intervals for market entry and exit. As an extension, we incorporate a stop-loss constraint to limit the maximum loss. We show that the entry region is characterized by a bounded price interval that lies strictly above the stop-loss level. As for the exit timing, a higher stop-loss level always implies a lower optimal take-profit level. Both analytical and numerical results are provided to illustrate the dependence of timing strategies on model parameters such as transaction cost and stop-loss
Doering: Does Pairs Trading with ETFs Work?
Several studies have demonstrated that pairs trading with single stocks achieved significant excess returns over long periods, with diminishing returns in recent years. However, practitioners extended pairs trading to ETFs. We discuss potential benefits of ETF over stock pairs and test whether ETF pairs trading is profitable, using NYSE Arca trading data over a time span from 2001 to June 2016, and compare our results to stock pairs trading. We find that ETF pairs achieve average excess returns of up to 27 bps per month, but are usually less profitable than stock pairs. Additionally, our results suggest that though ETF pairs trading involves lower arbitrage risks compared to stock pairs, this advantage is more than outweighed by a higher market efficiency effect.