壓力測試：厚尾偏態與混合分配模型的應用

Improving the Stress Tests: Fat-tail, Skewness and Mixture-distribution Models

本文使用混合分配方法延伸 Kim and Finger（2000）的模型以發展一個可抓住厚尾，偏態（報酬分配不對稱）的壓力測試模型。在資料上，我們以希臘國債與S&P500 指數作為核心資產（風暴來源），臺灣市場的標的資產作為邊緣資產（風暴的受災者）。在與資料的配適能力上，本文發展的模型確實優於過去假設常態分配的壓力測試模型。在實際執行壓力測試中，本研究比較了本文使用的混合分配模型與過去模型的差異，我們發現壓力測試結果的差異相當大，因此肯定了能抓住厚尾及偏態現象模型的重要性。

In this paper, the authors extend the broken-arrow stress test, introduced by Kim and Finger (2000), to include the fat-tail, and skewness of asset return into the model. Then they estimate correlation levels in stress situations and apply these correlation levels to obtain the expected loss for the peripheral assets in the stress test. To identify correlation in stress situations, the joint distribution of core assets (S&P 500 and Greek government bonds) and peripheral assets (16 Taiwan Financial assets) is specified as a mixture of skewed-t family. The mixture of skewed-t (or t), in which it captures the fat-tail and skewness of asset return, produces the best fitness with data and a more reasonable expected loss than other parametric models.