EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0886
Title: The Maximal Range-Return Divergence Statistic Authors:  Yifan Li - The University of Manchester (United Kingdom) [presenting]
Ingmar Nolte - Lancaster University (United Kingdom)
Sandra Nolte - Lancaster University (United Kingdom)
Abstract: We propose the Maximal rAnge-rEturn Divergence (MAED) statistic, which is defined as the maximal distance between the price range and the absolute return on a fixed time interval. The statistic can be easily constructed when high-frequency transaction data is available. The MAED statistic summarizes the inward movement of price paths, which contains substantially different information to the candlestick data (i.e., high, low, open, and close price in an interval) that mainly capture the outward movement of prices. We propose a spot volatility estimator based on the MAED-augmented candlestick data and establish its asymptotic properties in the fixed-k asymptotic setting with discrete price observations. Our analytical and simulation results show that our MAED-augmented estimator can reduce the asymptotic variance of the optimal candlestick-based spot volatility estimator by as much as 40%. In the presence of extreme price movements such as a jump or a drift burst, the MAED statistic has very different behaviour from the candlestick-based statistics. This allows us to monitor jumps or drift bursts in real time.