01.01.2015 Views

Does Tail Dependence Make A Difference In the ... - Boston College

Does Tail Dependence Make A Difference In the ... - Boston College

Does Tail Dependence Make A Difference In the ... - Boston College

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

have<br />

(<br />

)<br />

MES(τ) = −E X|Y V aR y (τ)<br />

= −<br />

= −<br />

= −<br />

= − 1 τ<br />

N∑<br />

i=1<br />

N∑<br />

i=1<br />

(<br />

)<br />

x i P r X = x i |Y V aR y (τ)<br />

(<br />

)<br />

P r X = x i , Y V aR y (τ)<br />

x i (<br />

)<br />

P r Y V aR y (τ)<br />

)<br />

(Y V aR y (τ)|X = x i<br />

N∑ P r<br />

x i<br />

i=1<br />

N∑<br />

i=1<br />

(<br />

) P r(X = x i )<br />

P r Y V aR y (τ)<br />

)<br />

x i P r<br />

(Y V aR y (τ)|X = x i P r(X = x i )<br />

(5)<br />

Let P r(Y = y i ) = v and P r(X = x i ) = u, we have already proved in <strong>the</strong> beginning of appendix<br />

that <strong>the</strong> conditional distribution F (y i |x i ) is equal to <strong>the</strong> first order derivative of copula function<br />

C(u, v) with respect to u<br />

Therefore<br />

)<br />

P r<br />

(Y y i |X = x i =<br />

)<br />

P r<br />

(Y V aR y (τ)|X = x i =<br />

∂C(u, v)<br />

∂u<br />

∂C(u, τ)<br />

∂u<br />

Substituting equation (6) into <strong>the</strong> equation (5) yields<br />

(<br />

)<br />

MES(τ) = −E X|Y V aR y (τ)<br />

(6)<br />

= − 1 τ<br />

N∑<br />

i=1<br />

∂C(u, τ)<br />

x i P r(X = x i )<br />

∂u<br />

38

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!