I have a time series of log-returns of a stock. I want to determine whether the time series is just white noise or if there are some other pattern. How to I use the definition of white noise to make a conclusion?

  • 1
    $\begingroup$ Can you add a graph of what your time series currently looks like, and its ACF? $\endgroup$ – Jon Nov 1 '17 at 19:56
  • $\begingroup$ anyway, financial time series implies "some other pattern" ,all the time. $\endgroup$ – tagoma Nov 1 '17 at 20:11

You want to look at an autocorrelation function (ACF) plot. If no lags are significantly correlated, then you basically have white noise or a MA(q) process aka moving average.

You can use this guide here to compare what your ACF plot looks like to determine if your time-series is "white noise" or not. Guide to ACF/PACF Plots

Also, feel free to browse through previous time-series related posts like:

  1. How to interpret ACF and PACF plots

  2. Help interpreting ACF- and PACF-plots

  3. Interpreting ACF and PACF Plot


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.