Is it possible to apply the idea of GARCH to time series models that include exogenous variables?

For example, say I estimate a cash flow forecast model. Does it make sense to model the residuals by use of a GARCH model? Or formulated alternatively: Does it make sense to try to forecast volatility in cash flow forecast errors by usage of a GARCH model?

  • $\begingroup$ Technically, this is possible. GARCH is a model of your residuals, ARIMA(X) is a model of your levels. Estimation / Calibration can become tedious, though. $\endgroup$ – Kermittfrog Oct 16 '20 at 11:27
  • $\begingroup$ Are there any lecture slides/papers you could point me to? Or programming packages? $\endgroup$ – shenflow Oct 16 '20 at 12:22
  • $\begingroup$ I shenflow, I would suggest stats quantexchange for this, e.g. here: stats.stackexchange.com/questions/254101/… $\endgroup$ – Kermittfrog Oct 16 '20 at 12:23
  • $\begingroup$ Agreed with @Kermittfrog regarding the technicalities of this estimation. However, if you could elaborate more on the cash flow modelling it would fit here better I think. I'm not sure what you're trying to achieve though. $\endgroup$ – Bob Jansen Oct 16 '20 at 13:44
  • $\begingroup$ I am interested in forecasting the risk, defined as the volatility, in cash flows, as opposed to forecasting the conditional mean @Bob Jansen. $\endgroup$ – shenflow Oct 17 '20 at 6:01

Your Answer

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

Browse other questions tagged or ask your own question.