股價報酬之財務時間序列方法風險值評估

Abstract

隨著經濟日益發展、生活水準提升，投資資訊也愈來愈普及與透明化，人民運用各種投資理財工具去創造更大的報酬，但是於1994年美國加州橘郡銀行與1995年英國霸菱銀行皆發生破產事件，起因於交易員的不當操作而宣告倒閉，所以風險管理也愈來愈受到重視，1988年巴賽爾銀行監理委員會開始允許銀行以風險值（VaR）模型計算市場風險資本，並於1996年提出回溯測試，準則是以250個營業日與估計值做比較，我們選定的樣本時間為2004年1月1日至2011年12月31日的日本東京日經225指數、香港恆生指數和美國紐約史坦普爾500股價指數來預測2012年1月1日後的250個營業日，其中以短期歷史模擬法（short term）、長期歷史模擬法（long term）、風險矩陣（RiskMetrics）、GARCH-N、GARCH-t、IGARCH-N與IGARCH-t來做分析預測，最後使用風險評估表來解釋各方法是否合適。

Along with the economy develop increasingly, the standard of living improve fast, and investment information is also universal and transparent. People use financial management tools to create the largest rewards. But in 1994 and 1995, Orange County’s Bank at California and Barings Bank at UK went into bankruptcy because traders operated improperly, so they managed the risk much attention attention . In 1988, the Basel Committee on Banking Supervision began allowing banks to use value-at-risk model to calculate market risk capital. In 1996, it proposed backtesting guideline which compared with the estimated values and the 250 business days . The data set was collected in January 1, 2004 to December 31, 2011 which is available at TEJ. We collect Nikkei 225, Hang Seng and S&P500 index and predict the 250 business days after 1 January 2012. There are short-term historical simulation method, long-term historical simulation method, RiskMetrics, GARCH-N, GARCH-t, IGARCH-N and IGARCH-t in our final model to analysis and predict. Finally, we also use the risk assessment table to explain if methods are appropriate.