Abstract:Cascades of events and extreme occurrences have garnered significant attention across diverse domains such as financial markets, seismology, and social physics. Such events can stem either from the internal dynamics inherent to the system (endogenous), or from external shocks (exogenous). The possibility of separating these two classes of events has critical implications for professionals in those fields. We introduce an unsupervised framework leveraging a representation of jump time-series based on wavelet coefficients and apply it to stock price jumps. In line with previous work, we recover the fact that the time-asymmetry of volatility is a major feature. Mean-reversion and trend are found to be two additional key features, allowing us to identify new classes of jumps. Furthermore, thanks to our wavelet-based representation, we investigate the reflexive properties of co-jumps, which occur when multiple stocks experience price jumps within the same minute. We argue that a significant fraction of co-jumps results from an endogenous contagion mechanism.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the classification problem of price jumps in financial markets, especially to measure the reflexivity of these jumps by analyzing their time series. Specifically, the author introduced an unsupervised framework based on wavelet coefficients to discover new price - jump categories and explored the occurrence mechanism of co - jumps. #### Main research questions 1. **Distinguish between endogenous and exogenous events**: The paper attempts to distinguish between events caused by the internal dynamics of the system (endogenous) and those caused by external shocks (exogenous). This is of great significance in fields such as financial markets, seismology, and social physics. 2. **Identify new categories of jumps**: By introducing a time - series representation method based on wavelet coefficients, the author hopes to identify new price - jump categories. These categories include: - **Reflexive jumps**: Classified according to the time - asymmetry of volatility. - **Mean - reverting jumps**: Classified according to the change in the sign of returns before and after the jump. - **Trend - aligned jumps**: Classified according to the consistency of trends before and after the jump. 3. **Endogeneity of co - jumps**: The author also explored whether price jumps (i.e., co - jumps) that occur in multiple stocks within the same minute are mainly caused by endogenous contagion mechanisms rather than driven by external factors. #### Method overview - **Wavelet transform**: Use wavelet coefficients \(W_jx(t)\) and wavelet coefficients of volatility \(W_{j_2}|W_{j_1}x|(t)\) to characterize the time series of jumps. - **Principal component analysis (PCA)**: Extract the main feature directions of the jump time series through PCA to achieve unsupervised classification. These directions include: - **Volatility - asymmetry direction \(D_1\)**: Used to measure the time - asymmetry of volatility. - **Mean - reverting direction \(D_2\)**: Used to measure the change in the sign of returns before and after the jump. - **Trend - behavior direction \(D_3\)**: Used to measure the consistency of trends before and after the jump. #### Key findings - **Volatility - asymmetry**: The time - asymmetry of volatility is one of the key features for distinguishing different jump types. For example, endogenous jumps are usually more symmetric than exogenous jumps. - **Mean - reverting and trend - behavior**: Mean - reverting and trend - behavior are two other key features that help to further refine the classification of jumps. - **Endogeneity of co - jumps**: The research shows that many co - jumps are not triggered by news or other external factors, but are caused by the contagion mechanism within the market. Through these methods, the author can not only better understand the nature of price jumps, but also provide valuable insights for the prediction and stability of financial markets.

Riding Wavelets: A Method to Discover New Classes of Price Jumps

Identifying jumps in high-frequency time series by wavelets

Dynamic Asset Allocation with Jump Risk

Double Smoothed Volatility Estimation of Potentially Non‐stationary Jump‐diffusion Model of Shibor

Application of Wavelet Analysis in Time-series Analysis

A Wavelet Method for Panel Models with Jump Discontinuities in the Parameters

Jump Clustering, Information Flows, and Stock Price Efficiency

Understanding jumps in high frequency digital asset markets

Common price and volatility jumps in noisy high-frequency data

Exogenous and Endogenous Price Jumps Belong to Different Dynamical Classes

Jump detection in high-frequency order prices

Studying the impact of fluctuations, spikes and rare events in time series through a wavelet entropy predictability measure

Wavelet Analysis of Cryptocurrencies -- Non-Linear Dynamics in High Frequency Domains

Price Jumps and the Related Liquidity Dynamics: Evidences from Chinese Stock Market

The dynamical relation between price changes and trading volume: A multidimensional clustering analysis

Pattern recognition in trading behaviors before stock price jumps: new method based on multivariate time series classification

The Economic Value of Volatility Timing with Realized Jumps

Investigating Impacts of Self-Exciting Jumps in Returns and Volatility: A Bayesian Learning Approach

Jumps at ultra-high frequency: Evidence from the Chinese stock market

Volatility Dynamics and Mixed Jump-GARCH Model Based Jump Detection in Financial Markets

Self-Exciting Jumps, Learning, and Asset Pricing Implications