Parts | Days | Selection | Search | Updates | Downloads | Help
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 22: Financial Markets and Risk Management II
SOE 22.1: Talk
Friday, March 31, 2023, 10:00–10:15, ZEU 260
Microscopic origin of the persistent order flows: microscopic data analysis — Yuki Sato1 and •Kiyoshi Kanazawa2 — 1University of Tsukuba, Tsukuba, Japan — 2Kyoto University, Kyoto, Japan
In financial markets, it is a stylised fact that the order flow exhibits persistence (or called the long-range correlation, LRC): if you observe a buy (sell) order, you will likely observe a buy (sell) order even in future. This character can be quantified as the power-law decay of the order-sign autocorrelation function C(τ)∝ τ−γ. In explaining the origin of the LRC, the order-splitting hypothesis was proposed as a promising theory. Further, Lillo, Mike, and Farmer proposed a minimal stochastic model of order-splitting traders in 2005, showing a quantitative prediction connecting the relationship between the microscopic and macroscopic behaviour. However, the LMF quantitative prediction has not yet been verified in the lack of appropriate microscopic datasets. In this talk, we solve this long-standing econophysics problem by analysing a huge microscopic dataset in the Tokyo Stock Exchange market. We apply a strategy clustering to identify the set of splitting traders and then measure the power-law exponent α in the metaorder-length distribution for splitting traders. We finally verify the quantitative prediction of the LMF model (γ=α−1) by providing the scatterplot between α and γ.