FIRE SALES FORENSICS: MEASURING ENDOGENOUS RISK
提出一个可操作框架，量化多资产市场中损失引发的甩卖对投资组合风险的影响，推导出甩卖对波动率和相关性的解析表达式，并开发计量方法用于甩卖事件的取证分析，包括参数可识别性条件、统计检验和估计量。
We propose a tractable framework for quantifying the impact of loss‐triggered fire sales on portfolio risk, in a multi‐asset setting. We derive analytical expressions for the impact of fire sales on the realized volatility and correlations of asset returns in a fire sales scenario and show that our results provide a quantitative explanation for the spikes in volatility and correlations observed during such deleveraging episodes. These results are then used to develop an econometric framework for the forensic analysis of fire sales episodes, using observations of market prices. We give conditions for the identifiability of model parameters from time series of asset prices, propose a statistical test for the presence of fire sales, and an estimator for the magnitude of fire sales in each asset class. Pathwise consistency and large sample properties of the estimator are studied in the high‐frequency asymptotic regime. We illustrate our methodology by applying it to the forensic analysis of two recent deleveraging episodes: the Quant Crash of August 2007 and the Great Deleveraging following the default of Lehman Brothers in Fall 2008.