Maximum likelihood estimation of autoregressive model parameters

Motivation Autoregressive (AR) models are a type of time series model in which the value at any point in time is a linear function of the previous values and white noise. AR models have many applications including, but not limited to, mathematical finance. In this short article, we derive the maximum likelihood estimator (MLE) for the coefficients and variance of an autoregressive model when the white noise is normally distributed. We will do this by appealing to the finite memory of the AR proc...

Submit