国际清算银行-累积风险溢价（英）-2023.10-73页_1mb

Summary of "The Cumulant Risk Premium" by Albert S. Kyle and Karam Todorov

Cumulant Risk Premium (CRP)

• The paper introduces a methodology to measure the risk premium of higher-order cumulants (related to moments) using leveraged ETFs, which are more liquid than options.
• The Cumulant Risk Premium (CRP) is the difference between physical and risk-neutral cumulants, measuring the risk premium associated with higher-order moments.
• CRP is shown to be negative and substantial in magnitude across various asset classes, averaging -7.4% annualized and more than 100% of the index risk premium (IRP).

Methodology

• Cumulants are used to quantify higher-order moments, and constant-beta assets (leveraged ETFs) are exposed to higher-order cumulants through dynamic rebalancing.
• The short-both strategy (short-selling assets with opposite leverage, e.g., -1 and 1) isolates the even-order CRP (CRPE), revealing that liquidity providers earn positive expected returns.

Key Findings

• Leverage amplifies exposure to higher-order cumulants, with even greater impact for larger |leverage|.
• The even-order CRP (CRPE) averages -4.4% annualized across assets, negatively correlated with market stress indicators like VIX.
• Short-both strategy returns exhibit high Sharpe ratios (above 1 in many markets), and their first principal component serves as a global market stress index, capturing stress in multiple asset classes.

Economic Implications

• The paper challenges the validity of standard linear beta pricing models like the CAPM, which fail in the presence of higher-order cumulants.
• Higher leverage strategies are significantly riskier due to non-linear exposure to higher-order cumulants.
• The CRP framework explains anomalies such as the flatness of the securities market line (SML) and the persistence of certain risk premiums (e.g., variance risk premium).

Practical Use

• The approach provides a cheaper and more precise alternative to options for pricing and hedging higher-order risks.
• The short-both strategy can be used to extract even-order cumulants and gauge market stress in real-time, without relying on complex option-based models.

Figures and Tables Mentioned

• Table 1: Estimates of C RP and CRPE across multiple assets and leverages.
• Table 2: Returns on short-both strategies, CRPE, and IRP.
• Table 3: Regression results linking short-both strategy returns to factors like VIX and momentum.
• Table 4: Cross-sectional asset-pricing tests using average short-both returns.
• Figures 1–9: Visualizing CRP loadings, cumulant dynamics, and market stress measures.