- Replicate the pairs trading results in the paper Illuminating the Profitability of Pairs Trading: A Test of the Relative Pricing Efficiency of Markets for Water Utility Stocks.
- Use the historical data in the accompanying file csv.
- Use the BIC criteria for the number of lags to use in all ADF tests. Consider a max number of lags equal to 10.
- Submit your work as either a Jupyter Notebook with Python code or an R Markdown notebook with R code.
Update: March 22, 2019
- I decided that you could take another month trying to get the exact replication of the paper’s trading strategy implementation. I don’t want you to spin your wheels on a lost cause. So here is what I want you to do instead.
- Estimate equation (6) in the paper for each of your pairs. Then form the residuals as outlined (see equations (7) and (8) in the paper).
- Esimate σ from the estimated residuals.
- Calculate the upper and lower bounds for the trading rules as:
UpperBound = +1.0 ∗ δiσˆ
LowerBound = −1.0 ∗ δiσˆ
- for i = {1,2,3} and δ = {0.25,0.5,0.75} as in the paper.
- Make three time series graphs with the residuals plotted against the upper and lower bounds for each δi.
- Fill in the following table with the number of buy and sell signals for each trading rule:
δi Buy Signals Sell Signals
- Explain what a buy signal and a sell signal means in the context of pairs trading.
- For example, identify the first buy and the first sell signals for the δ1 = 0.25 trading rule and interpret them both qualitatively and quantitatively. How many days is each position held for? What the return on the trade?
1
