• 沒有找到結果。

在本論文中,我們使用極值理論(EVT),以靜態與動態的方法,估算具有金 融市場特性的合成資料之風險值與條件風險值,並比較估算出的結果與樣本值、

理論值間的差異,得到以下的結論。對於損失分布為常態分布(normal distribution)

與 Student’s t 分布的靜態資料,利用區塊最大法估計出的風險值與條件風險值比起 樣本值更接近理論值,但對於損失分布為對數常態分布(lognormal distribution)

的靜態資料,卻會超過理論值。對於損失分布為常態與對數常態靜態資料,利用 穿越門檻值法估計出的風險值與條件風險值超過理論值,但對於損失分布為 Student’s t 分布的靜態資料,穿越門檻值法估計出的風險值與條件風險值小於理論 值,顯示出區塊最大法與穿越門檻值法對於不同的損失分布無法維持一定的趨 勢。另一方面,動態模型所估算出的風險值與條件風險值對於具有自我回歸的對 數報酬時間序列,不論其變異數模型是否為定數,皆較於樣本值接近理論值但小 於理論值,具有一定的趨勢。

附錄

定義 1. 風險值(value at risk;VaR)(Jorion,1997)

給定信心水準α∈(0,1),在期間 t 內且信心水準α 下的資產風險值是指最小的 R

k,使得資產在期間 t 內的損失L超越k的機率為1−α 。風險值的數學表示:

{

α

} {

α

}

α( )=inf ∈ : ( > )≤1− =inf ∈ : ( )≥

VaR L l R P L l l R FL l

其中,P(L>l)代表資產損失L大於 lR 的機率函數,FL(l)代表資產損失L小 於或等於 lR 的機率分布函數。

定義 2. 條件風險值(conditional value at risk;CVaR)(Rockafellar & Uryasev,2002)

條件風險值又稱為尾部風險值(tail VaR)、期望差額(expected shortfall)或平均超 額損失(mean excess loss)。根據定義 1.,我們可以定義條件風險值的數學表示為:

=

= − 1 1 1( )

1 VaR 1

1 CVaR 1

α α

α α xdx α FL x dx

其中,α 為信心水準,FL1(x)為資產損失L的機率分布函數的反函數,即分位數函 數(quantile functions)。

參考資料

[1] Artzner, Philippe, Freddy Delbaen, Jean‐Marc Eber, and David Heath.

Coherent measures of risk.

Mathematical Finance 9, no. 3 (1999): 203–228.

[2] Bekiros, Stelios D., and Dimitris A. Georgoutsos.

Estimation of Value-at-Risk by extreme value and conventional methods: a comparative evaluation of their predictive performance.

Journal of International Financial Markets, Institutions and Money 15, no. 3 (2005): 209–228.

[3] Danielsson, Jon, and Casper G. De Vries.

Value-at-risk and extreme returns.

Annales d'Economie et de Statistique 60 (2000): 239–270.

[4] Danielsson, Jon, Laurens de Haan, Liang Peng, and Casper G. de Vries.

Using a bootstrap method to choose the sample fraction in tail index estimation.

Journal of Multivariate Analysis 76, no. 2 (2001): 226–248.

[5] De Haan, Laurens, and Ana Ferreira. Extreme value theory: An introduction. New York: Springer, 2007.

[6] Drees, Holger.

Extreme quantile estimation for dependent data, with applications to finance.

Bernoulli 9, no. 4 (2003): 617–657.

[7] Embrechts, Paul, Sidney I. Resnick, and Gennady Samorodnitsky.

Extreme value theory as a risk management tool.

North American Actuarial Journal 3, no. 2 (1999): 30–41.

[8] Gençay, Ramazan, and Faruk Selçuk.

Extreme value theory and value-at-risk:

relative performance in emerging markets.

International Journal of Forecasting 20, no. 2 (2004): 287–303.

[9] Jorion, Philippe. Value at risk: the new benchmark for controlling market risk. Vol.

2. New York: McGraw–Hill, 1997.

[10] McNeil, Alexander J., and Rüdiger Frey.

Estimation of tail-related risk measures for heteroscedastic financial time series: an extreme value approach.

Journal of Empirical Finance 7, no. 3 (2000): 271–300.

[11] Poon, Ser-Huang, Michael Rockinger, and Jonathan Tawn.

Extreme value dependence in financial markets: Diagnostics, models, and financial implications.

Review of Financial Studies 17, no. 2 (2004): 581–610.

[12] Resnick, Sidney I. Heavy-tail phenomena: Probabilistic and statistical modeling.

New York: Springer, 2007.

[13] Rockafellar, R. Tyrrell, and Stanislav Uryasev.

Conditional value-at-risk for general loss distributions.

Journal of Banking & Finance 26, no. 7 (2002):

1443–1471.

[14] Singh, Abhay K., David E. Allen, and Powell J. Robert.

Extreme market risk and extreme value theory.

Mathematics and Computers in Simulation 94 (2013):

310–328.

[15] Shiller, Robert J. The subprime solution: How today’s global financial crisis happened, and what to do about it. Princeton, NJ: Princeton University Press,

2008.

[16] Taleb, Nassim Nicholas.

Black swans and the domains of statistics.

The American Statistician 61, no. 3 (2007): 198–200.

相關文件