Y**N 发帖数: 2628 | 1 谁能再帮看看这道题?谢了!
X and Y are lognormal variables. They have a correlation of rho. V[XY] = ? |
k**x 发帖数: 2611 | 2 用covariance的定义 V[XY]=cor[xy]*sqrt(v(x)v(y))
【在 Y**N 的大作中提到】 : 谁能再帮看看这道题?谢了! : X and Y are lognormal variables. They have a correlation of rho. V[XY] = ?
|
G********d 发帖数: 10250 | 3 这也叫题?
【在 Y**N 的大作中提到】 : 谁能再帮看看这道题?谢了! : X and Y are lognormal variables. They have a correlation of rho. V[XY] = ?
|
Y**N 发帖数: 2628 | 4 谢谢!!
【在 k**x 的大作中提到】 : 用covariance的定义 V[XY]=cor[xy]*sqrt(v(x)v(y))
|
n******m 发帖数: 169 | 5 晕,我理解为 Z=XY, 求 Var(Z)... |
z****g 发帖数: 1978 | 6 ...什么时候Cov(X,Y)变成V(XY)了... |
Y**N 发帖数: 2628 | 7 对不起, 题目没有说清楚. 题目是 "Z=XY, 求 Var(Z)..." |
x******a 发帖数: 6336 | 8 first consider Var(ln(Z))
【在 Y**N 的大作中提到】 : 对不起, 题目没有说清楚. 题目是 "Z=XY, 求 Var(Z)..."
|
w******n 发帖数: 309 | 9 but after getting var(InZ) , what should do since the relationship between
Var(Z) and Var(InZ) is not so clear |
j********t 发帖数: 97 | 10 Consider Z = exp(lnX + lnY). lnX, lnY are normal and correlated, I guess (
lnX, lnY) is a joint normal, then lnX+lnY is normal. Utilize moment
generating function of normal to calculate E(Z) and E(Z^2).
But I'm not sure how to get corr(lnX, lnY) from corr(X, Y)=\rho and prove (lnX, lnY) is a bivariate normal? Anyone can help? Thanks! |
f*******g 发帖数: 377 | 11 尽管写的是corr(X, Y)=\rho
但这个rho应该指的就是corr(lnX, lnY)
要不然不太好办
(lnX, lnY) is a bivariate normal? Anyone can help? Thanks!
【在 j********t 的大作中提到】 : Consider Z = exp(lnX + lnY). lnX, lnY are normal and correlated, I guess ( : lnX, lnY) is a joint normal, then lnX+lnY is normal. Utilize moment : generating function of normal to calculate E(Z) and E(Z^2). : But I'm not sure how to get corr(lnX, lnY) from corr(X, Y)=\rho and prove (lnX, lnY) is a bivariate normal? Anyone can help? Thanks!
|