Although it is clear that the volatility of asset returns is serially correlated, there is no general agreement as to the most appropriate parametric model for characterizing this temporal dependence. In this paper, we propose a simple way of modeling financial market volatility using high-frequency data. The method avoids using a tight parametric model by instead simply fitting a long autoregression to log-squared, squared, or absolute high-frequency returns. This can either be estimated by the usual time domain method, or alternatively the autoregressive coefficients can be backed out from the smoothed periodogram estimate of the spectrum of log-squared, squared, or absolute returns. We show how this approach can be used to construct volatility forecasts, which compare favorably with some leading alternatives in an out-of-sample forecasting exercise.

ABSTRACT Variance‐ratio tests are routinely employed to assess the variation in return volatility over time and across markets. However, such tests are not statistically robust and can be seriously misleading within a high‐frequency context. We develop improved inference procedures using a Fourier Flexible Form regression framework. The practical significance is illustrated through tests for changes in the FX intraday volatility pattern following the removal of trading restrictions in Tokyo. Contrary to earlier evidence, we find nodiscernible changes outside of the Tokyo lunch period. We ascribe the difference to the fragile finite‐sample inference of conventional variance‐ratio procedures and a single outlier.