2
$\begingroup$

I am working on stock prices and stock returns and I'm supposed to do some forecast on these data. The stock prices series is not stationary and even if the stock returns series is, it is a white noise. I can't do any forecasting on the stock returns since it is a white noise. My professor suggested that I should adjust an AR(1) or MA(1) process, but I don't really understand what she means by adjusting these processes?

Thank you for your answers

$\endgroup$
0
$\begingroup$

What she probably means is to detect anomalous data points (pulses) or a mean shift or a trend change then adjust the original series for these effects. What she probably means is that if the series has a model where the errors have non-constant variance requiring either weighted least squares or a power transformation adjust accordingly.

$\endgroup$

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.