The foreign exchange (FX) market is usually analyzed as a homogeneous sequence of returns r defined as the price difference over a fixed period of time (Dacorogna et al., 2001). This metric is used for a discontinuous time series: over weekends, trading comes to a standstill or, inversely, at news announcements, there are spurts of market activity. Ideally, time should be a dynamic object that adapts itself to market activity. To achieve this goal, we propose an event-based approach that analyzes the time series in terms of price directional changes of a given amplitude λ (Glattfelder et al., 2010). Within this framework, time flows unevenly: any occurrence of a directional change represents a new intrinsic time unit. Intrinsic time flows to the beat of events and is thus better suited to model the dynamics of the underlying processes.
The dissection algorithm identifies the occurrence of a price change λ from the last high or low (i.e., an extrema), whether it is in an up or down mode, respectively. At each occurrence of a directional change, there is the so-called overshoot associated with the previous directional change. The overshoot is defined as the difference between the price level at which the last directional change occurred and the extrema before the next directional change is triggered. Figure 20.1 shows how the price curve is dissected into directional change and overshoot sections.