On the use of the Box–Cox transformation on conditional variance models

We investigate the effects of using the Box–Cox transformation on conditional variance specifications. By deriving its autocorrelation functions, we infer “rich” autocorrelation structures due to the existence of the specification parameter in this non-linear transformation. To illustrate transformation's effects on conditional variance models, we first generate its theoretical autocorrelation function and then investigate model's fit using real financial time-series data.