A0210
Title: Financial market efficiency during crisis periods: A long-memory approach based on price ranges
Authors: Cristina Sattarhoff - Kiel University (Germany) [presenting]
Abstract: The intermittency coefficient $\lambda^2$, a parameter of the multifractal random walk model of financial volatility, measures the degree of deviation from a random walk in terms of nonlinear dependence patterns of returns and other price transformations over different time horizons (minute, daily, monthly returns, etc.). The $\lambda^2$ is estimated from price range data using the scaling property of the log-variogram. After assessing the small sample properties in a Monte Carlo simulation study by comparison with well-established estimation procedures for returns, the ranges-based method is employed to track efficiency losses during 2000-2024 for nine international stock market indices using two datasets of daily and minute data, respectively. According to the results, exceptionally large $\lambda^2$ estimates, as classified with the interquartile method, mark financial crisis years throughout. Moreover, $\lambda^2$ is estimated based on 30-minute ranges using a rolling window of 3 months, and exceptionally large values are found, which persist for nearly two weeks before the major price drops during the 2020 COVID stock market crash. These results are very promising for the early detection of crisis events, and this is also the first empirical application of the ranges-based method.
