A study of persistence of price movement using High Frequency Financial Data

We present a methodology to discover how often the price deviates from established levels without sound reason. We look at the distribution of these so called “rare events” during the trading day and analyze the speed of the price bouncing back to the levels established before the rare events. We provide methods to calibrate trading rules based on the detection of these events and we exemplify for a particular trading rule. In order to draw comprehensive conclusions we group the equity into classes depending on the volume of daily trades. We find the behavior these classes to be very different. We also find possible evidence of algorithmic trading related to these events. ______________________________________________________________________________________