Originally posted by Riesgo2002
GARCH is a mathematical model used to forecast volatilities based on HVs, it is an autoregressive model, meaning that it reverts back to the mean over time.
If I interpret correctly, autoregressive means past has impact to today, which is the opposite of mean reversing. In the example of one lag volatility, means todays volatility has yesterday's volatility and other components. In a simple term, it can be said that if today's volatility is high (low), tomorrow's volatility is more likely to be high too (low).
