Zapisane w:
| 1. autor: | |
|---|---|
| Format: | Recurso digital |
| Język: | angielski |
| Wydane: |
Zenodo
2025
|
| Hasła przedmiotowe: | |
| Dostęp online: | https://doi.org/10.5281/zenodo.18018292 |
| Etykiety: |
Dodaj etykietę
Nie ma etykietki, Dołącz pierwszą etykiete!
|
Spis treści:
- <p>LRD v6.0.1 is a reproducible, audit-driven empirical suite for long-range dependence / fractal memory in real-world data, centered on Bitcoin volatility as a benchmark for AI forecasting and risk systems.</p> <p> </p> <p>The pipeline estimates scaling via DFA-2, quantifies uncertainty via block bootstrap confidence intervals, and applies phase-randomized surrogate tests as a falsification step. If scaling is not distinguishable from surrogates or collapses under scale-range sensitivity checks, the "LRD" claim is rejected.</p> <p> </p> <p>What's new in v6.0.1: Two short addenda link the macroscopic volatility-memory signal to modern market microstructure universality results:</p> <p> </p> <p>Square-root law (SRL) of price impact — Sato & Kanazawa (Phys. Rev. Lett. 135, 257401, 2025) provide high-precision evidence that the SRL holds with exponent δ ≈ 1/2 across all liquid stocks and traders on the Tokyo Stock Exchange over eight years.</p> <p>Theoretical framework — As Bouchaud argues in an APS Physics Viewpoint, universal laws can emerge in financial markets from the aggregation of diverse microscopic behaviors, analogous to critical phenomena in physics.</p> <p>This provides a mechanistic anchor for interpreting cross-market fractal memory while maintaining the falsifiable empirical protocol.</p> <p> </p> <p>Domains covered: Bitcoin/crypto volatility, earthquake activity, HRV (physiology), genomics.</p> <p> </p> <p>Audit protocol: DFA-2 + block-bootstrap CI + phase-randomized surrogate tests + scale-range sensitivity.</p>