Financial Return Distributions: Past, Present, and COVID-19

Area of Science:

• Quantitative Finance
• Statistical Modeling
• Financial Market Analysis

Background:

• Analysis of price return distributions is crucial for understanding financial market behavior.
• Previous studies suggested a universal acceleration in market time flow, leading to faster convergence towards normal distributions.
• Recent data (2017-2020) is used to re-evaluate these distribution models.

Purpose of the Study:

• To model and analyze the tails of return distributions for various financial assets (currency exchange rates, cryptocurrencies, CFDs) at different time scales.
• To investigate changes in fitted function parameters over the years and compare with earlier findings.
• To assess the validity of the hypothesized universal constant acceleration of market time flow.

Main Methods:

• Utilized power-law, stretched exponential, and q-Gaussian functions to model return distribution tails.
• Analyzed recent financial data from 2017-2020 for currency exchange rates, cryptocurrencies, and CFDs (stock indices, shares, commodities).
• Compared fitted parameters across different time scales and over years with previous research.

Main Results:

• The inverse-cubic power-law remains a suitable global reference for time horizons up to a few minutes.
• The hypothesized universal constant acceleration of market time flow, leading to faster convergence to normal distribution, was not observed.
• Market dynamics are characterized by alternating regimes, with events like the COVID-19 pandemic significantly impacting short-term behavior.

Conclusions:

• Current market dynamics and short-term processes may mask the previously observed convergence towards normal distribution.
• The speed of market time flow and asset cross-correlation act in opposing directions on return distribution tails.
• Financial market behavior is complex, involving a continuous alternation of regimes with distinct statistical properties.