Estimation in Financial Models - RiskLab
Estimation in Financial Models - RiskLab
Estimation in Financial Models - RiskLab
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
where<br />
(X (i,1) ; ) =E <br />
h<br />
(Xi , F (X (i,1) ; )) 2 jX (i,1)<br />
i<br />
; i =1;:::;n: (3.51)<br />
The function G n() is with<strong>in</strong> the class (3.49) <strong>in</strong> some sense "closest" to the<br />
score function based on the usually unknown exact likelihood function.<br />
It should be mentioned here that for small the mart<strong>in</strong>gale estimat<strong>in</strong>g<br />
function ~ G n (), dened <strong>in</strong> (3.48), is a rst order approximation <strong>in</strong> of G n,<br />
that means ~G n () is approximately optimal.<br />
(3) In some cases there are possibly numerical problems <strong>in</strong> calculat<strong>in</strong>g<br />
F _ (x; ) <strong>in</strong> (3.50). One way to solve these problems <strong>in</strong>volves approximat<strong>in</strong>g<br />
F _ (x; ) up to the order O( 2 ). That leads to a third mart<strong>in</strong>gale estimat<strong>in</strong>g<br />
function<br />
G + n () =<br />
nX <br />
_b(X(i,1) ; ) + 1 _<br />
2 2 b(X(i,1) ; )b 0 (X (i,1) ; )<br />
i=1<br />
+b(X (i,1) ; ) _ b 0 (X (i,1) ; )+ 1 2 (_2 (X (i,1) ; )b 00 (X (i,1) ; )<br />
+ 2 (X (i,1) ; ) b _ i 00 (X i , F (X (i,1) ; ))<br />
(X (i,1) ; )) : (3.52)<br />
(X (i,1) ; )<br />
Altogether we have now found expressions for three dierent zero-mean P -<br />
mart<strong>in</strong>gale estimat<strong>in</strong>g functions.<br />
As for the multi-dimensional case, suppose is k-dimensional, fX t g and<br />
b(X t ; ) are d-dimensional, is a dm-dimensional matrix with T positive<br />
denite and the Wiener process fW t g is m-dimensional. Then the k 1-<br />
dimensional mart<strong>in</strong>gale estimat<strong>in</strong>g functions ~G n and G n have the form<br />
and<br />
~G n () =<br />
G n() =<br />
nX<br />
_b(X (i,1) ; ) T (X (i,1) ; )((X (i,1) ; ) T ,1<br />
i=1<br />
X i , F (X (i,1) ; ) <br />
nX<br />
F _<br />
(X (i,1) ; ) T (X (i,1) ; ) ,1 X i , F (X (i,1) ; ) ;<br />
i=1<br />
where is assumed to be positive denite and _ b and _ F denote the d k-<br />
dimensional matrices of partial derivatives with respect to the components<br />
of .<br />
35