2013/08/30 by Vladimir Filimonov, Didier Sornette, Filimonov, Vladimir +1 · 2 citations
Biochemistry, Genetics and Molecular Biology · Environmental Science · Mathematics · #Diffusion and Search Dynamics #Ecosystem dynamics and resilience #FOS: Economics and business #Point processes and geometric inequalities #Statistical Finance (q-fin.ST)
paper · pdf · doi:10.48550/arxiv.1308.6756
openalex publication_date 2013/08/30 · openalex created_date 2022/10/02 · openalex updated_date 2026/07/28
We present a careful analysis of possible issues on the application of the\nself-excited Hawkes process to high-frequency financial data. We carefully\nanalyze a set of effects leading to significant biases in the estimation of the\n"criticality index" n that quantifies the degree of endogeneity of how much\npast events trigger future events. We report a number of model biases that are\nintrinsic to the estimation of brnaching ratio (n) when using power law memory\nkernels. We demonstrate that the calibration of the Hawkes process on mixtures\nof pure Poisson process with changes of regime leads to completely spurious\napparent critical values for the branching ratio (n~1) while the true value is\nactually n=0. More generally, regime shifts on the parameters of the Hawkes\nmodel and/or on the generating process itself are shown to systematically lead\nto a significant upward bias in the estimation of the branching ratio. We also\ndemonstrate the importance of the preparation of the high-frequency financial\ndata and give special care to the decrease of quality of the timestamps of tick\ndata due to latency and grouping of messages to packets by the stock exchange.\nAltogether, our careful exploration of the caveats of the calibration of the\nHawkes process stresses the need for considering all the above issues before\nany conclusion can be sustained. In this respect, because the above effects are\nplaguing their analyses, the claim by Hardiman, Bercot and Bouchaud (2013) that\nfinancial market have been continuously functioning at or close to criticality\n(n~1) cannot be supported. In contrast, our previous results on E-mini S&P 500\nFutures Contracts and on major commodity future contracts are upheld.\n