20230907-华安证券-_学海拾珠_系列之一百五十七_基于隐含波动率和实际波动率的系统风险指标_26页_2mb

报告摘要

Report Summary: Based on Implied and Realized Volatility for Systemic Risk Indicator

Introduction

This report introduces a novel systemic risk indicator, IVRVSRI, derived from implied volatility (IV) and realized volatility (RV), proposed in the "Xuehai Shipuzhu" series No. 157. It analyzes global and national stock market crises from 2000 to 2023 using a combination of market-derived data.

Key Objectives and Findings

Core Innovation: IVRVSRI merges IV and RV to create a simple, robust system risk measure that reduces model complexity and dependency on paid data.
Robustness: The indicator demonstrates sensitivity to various parameters (e.g., memory parameter for RV, IV maturity, percentile choices) but remains effective across different settings, reflecting systemic risk dynamics accurately.
Predictive Power: IVRVSRI outperforms other benchmarks (e.g., SRISK, CATFIN, CISS) in predicting S&P500 index downturns, especially during high systemic risk events, using regression analyses at multiple horizons.
Global Application: It captures regional differences in risk responses (e.g., Japan showed weaker reactions to COVID-19 in 2020) and provides early warnings for major crises like the Global Financial Crisis and COVID-19 pandemic.
Validation: Based on historical data from 2000-2023, the indicator was tested across multiple markets and models, confirming its reliability for real-time monitoring.

Methodology and Data

Approach: Draws from literature on volatility estimation and dynamic ranking, using daily data for IV and RV calculations. IVRVSRI is computed as a weighted sum of IVSRI and RVSRI.
Data Scope: Covers worldwide equity markets, including SP500, EuroStoxx50, Nikkei225, and Bovespa, with focus on major financial events.

Implications and Applications

Practical Use: IVRVSRI can be applied for real-time risk assessment in finance, offering a cost-effective tool for institutions to monitor and mitigate systemic risks.
Limitations: The indicator is based on historical data and does not constitute investment advice.