首页 | 学习中心 | 注册 登录
返回列表 发帖
sabre

UID
223198 
帖子
385 
主题
170 
注册时间
2011-7-11 
最后登录
2016-4-19 
1# 跳转到 » 倒序看帖
打印
tT
发表于 2013-4-28 12:19 | 只看该作者

Portfolio Standard Deviation Question

part, standard, because, average
Portfolio Standard Deviation Question:
On Page 178 of Secret Sauce it says that if the two assets in a two-asset portfolio are perfectly positively correlated (p = +1) then the standard deviation of the portfolio will just be the weighted standard deviations of each asset.  Isn’t this wrong?  My thinking is that if the correlation is ZERO then the portfolio would be the weighted average of the standard deviations, because a correlation of zero would elimanate the 3rd part of the portfolio standard deviation formula (it would make the 3rd part which is 2(w1)(w2)(SD1)(SD2)(P) equal to zero because P=0 and would leave behind the weighted SD)
Can somebody please explain this? or is this a type in the secret sauce?
收藏 分享

pimpineasy

Rank: 2

UID
222246 
帖子
238 
主题
53 
注册时间
2011-7-2 
最后登录
2016-4-19 
2#
发表于 2013-4-28 12:20 | 只看该作者
No… the statement is correct. Variance formula for two assets is (w1*sd1)^2 + (w2*sd2)^2 + 2*w1*w2*sd1*sd2*rho. If rho = 1, then this formula reduces to (w1*sd1 + w2*sd2)^2. Standard deviation is the square root, w1*sd1 + w2*sd2.

TOP

pimpineasy

Rank: 2

UID
222246 
帖子
238 
主题
53 
注册时间
2011-7-2 
最后登录
2016-4-19 
3#
发表于 2013-4-28 12:20 | 只看该作者
Why does rho=1 eliminate the entire third section from the equation though.. wouldn’t rho=0 do this?

TOP

bulosehi

Rank: 1

UID
442300 
帖子
41 
主题
10 
注册时间
2012-6-5 
最后登录
2014-8-5 
4#
发表于 2013-4-28 12:20 | 只看该作者
You are thinking that the formula for variance is (w1*sd1 + w2*sd2)^2 + 2*w1*w2*sd1*sd2*rho. This is wrong (see above).

TOP

andytrader

UID
223231 
帖子
520 
主题
166 
注册时间
2011-7-11 
最后登录
2013-8-21 
5#
发表于 2013-4-28 12:20 | 只看该作者
I appreciate your help but all I’m seeing is that applying a coefficient of zero to the end of the formula would eliminate the last section entirely whereas applying a coefficient of +1 would keep it as some positive value

TOP

kickthatcfa

Rank: 2

UID
223439 
帖子
318 
主题
7 
注册时间
2011-7-11 
最后登录
2014-8-7 
6#
发表于 2013-4-28 12:21 | 只看该作者
I give up.

TOP

kmf229

UID
223327 
帖子
324 
主题
105 
注册时间
2011-7-11 
最后登录
2013-10-21 
7#
发表于 2013-4-28 12:21 | 只看该作者
(match+box)^2 = match^2 + box^2 + 2 match box
now match=w1*sd1
box=w2*sd2
and the 3rd part you are so kindly talking about refers to 2 * match * box * rho.
if rho = 1 - variance becomes (match + box)^2
if rho=0 - variance becomes match^2 + box^2
now figure…
and also closely see the brackets = the way you are thinking about it - and how ohai above has correctly explained it.

TOP

morebeans

Rank: 2

UID
223354 
帖子
192 
主题
39 
注册时间
2011-7-11 
最后登录
2016-4-19 
8#
发表于 2013-4-28 12:21 | 只看该作者
Got it now.  Thank you both.

TOP

返回列表