ARCH models in Python

Model Confidence Set (MCS) implementation in Python

Easily evaluate your forecasts with (multivariate) Diebold-Mariano and (multivariate) Giacomini-White tests of equal predictive ability and MCS.

This repository includes the scripts to replicate the results of my paper entitled "A False Discovery Rate Approach to Optimal Volatility Forecasting Model Selection".

The MCB for variable selection identifies two nested models (upper and lower confidence bound models) containing the true model at a given confidence level.