Python for Finance Cookbook

Absence of autocorrelation in returns

Investigate this third fact about the absence of autocorrelation in returns.

1. Define the parameters for creating the autocorrelation plots:

N_LAGS = 50SIGNIFICANCE_LEVEL = 0.05

2. Run the following code to create the autocorrelation function (ACF) plot of log returns:

acf = smt.graphics.plot_acf(df.log_rtn,                             lags=N_LAGS,                             alpha=SIGNIFICANCE_LEVEL)

Executing the preceding code results in the following plot:

Only a few values lie outside the confidence interval (we do not look at lag 0) and can be considered statistically significant. We can assume that we have verified that there is no autocorrelation ...