Statistical arbitrage is a quantitative finance strategy used by traders to profit from pricing inefficiencies between related financial instruments
The ADF Test is used to determine if a time series is stationary. Stationarity is essential when analyzing data for trading strategies as it suggests mean-reverting behavior. The ADF_Test.ipynb notebook walks through the process.
p-value ≤ 0.05: Strong evidence against the null hypothesis. Reject the null hypothesis. The data is likely stationary.
- 0.05 < p-value ≤ 0.1: Weak evidence against the null hypothesis. Depending on the context and other factors, you might still reject the null hypothesis.
- p-value > 0.1: Not enough evidence against the null hypothesis. Fail to reject the null hypothesis. The data is likely non-stationary.
Cointegration occurs when two or more time series exhibit a stable, long-term relationship. The cointegration_test.ipynb notebook uses statistical tests to identify cointegrated pairs, which can inform pair trading strategies.
Backtesting involves simulating a trading strategy with historical data to evaluate its past performance. This process helps in understanding how a strategy would have behaved in different market conditions.
| Metric | Value |
|---|---|
| Risk-Free Rate | 0.0% |
| Time in Market | 50.0% |
| Cumulative Return | 49.93% |
| CAGR﹪ | 3.13% |
| Sharpe | 1.01 |
| Prob. Sharpe Ratio | 99.95% |
| Smart Sharpe | 0.98 |
| Sortino | 1.73 |
| Smart Sortino | 1.67 |
| Sortino/√2 | 1.22 |
| Smart Sortino/√2 | 1.18 |
| Omega | 1.29 |
| Max Drawdown | -4.41% |
| Longest DD Days | 399 |
| Volatility (ann.) | 4.51% |
| Calmar | 0.71 |
| Skew | 2.53 |
| Kurtosis | 32.05 |
-
Backtesting Strategies:
- Modify parameters in the backtesting notebook to try different trading strategies.
- Run the notebook to observe results for each modification.
-
Cointegration Testing:
- Load currency pairs from the provided CSV files and run the cointegration tests to identify profitable pairs.
Below is a sample performance report generated from one of the backtesting notebooks: