HIGH-FREQUENCY TIME SERIES

A time-series analysis generally assumes that the time lag between two consecutive observations is equally long; these are referred to as equidistant observations (data). For example, the daily closing rates of listed shares can be regarded as equidistant. In the case of equidistant data, standard time-series analysis methods can be used to analyze data.

If the (high-frequency) intraday rates of shares are analyzed rather than their closing rates, they are generally no longer equidistant, which means that the standard methods of time-series analysis can no longer be used.

Research therefore focuses on methods and models that explicitly take the non-equidistant and high-frequency data structure into account.