m****r
发帖数: 59 | 1 Suppose the annual variance of stock market returns over the last ten years
has been 0.0225. Also suppose that the T-bill rate has been constant at 3% a
year. What is the minimum positive average stock return for which we can
reject ( at the 5% confidence level ) the hypothesis that the equity premium
over the last then years is zero?
What's the use of 5% confidence level here? And to use the Utility function,
we should know the risk aversion coefficient. It's missing. Or, should we use
other fu | q**i
发帖数: 174 | 2 isn't that simply a one-sided test? find out the coresponding sigma for
a one-sided test at 5% confidence level (95%?). the that many sigma on the
left side of the 3% is the minimum you want.
by the way, you have a better chance getting it answer at a stats /
math forum.
【在 m****r 的大作中提到】
: Suppose the annual variance of stock market returns over the last ten years
: has been 0.0225. Also suppose that the T-bill rate has been constant at 3% a
: year. What is the minimum positive average stock return for which we can
: reject ( at the 5% confidence level ) the hypothesis that the equity premium
: over the last then years is zero?
: What's the use of 5% confidence level here? And to use the Utility function,
: we should know the risk aversion coefficient. It's missing. Or, should we use
: other fu
| m****r
发帖数: 59 | 3 Thanks for your explanation.
The sigma is 0.15.
But I don't quite understand what you meant. I thought to use the Utility
Function: U = E - a*S^2. E is the expected reutern. S is the standard
deviation.
What idea is behind your method? Really sorry for my ignorance.
Any could you tell me how I can know at how many sigmas the confidence level
is 5%? I checked though the web. COuldn't find such info easily.
Thanks a lot!
years
a
premium
function,
use
【在 q**i 的大作中提到】
: isn't that simply a one-sided test? find out the coresponding sigma for
: a one-sided test at 5% confidence level (95%?). the that many sigma on the
: left side of the 3% is the minimum you want.
: by the way, you have a better chance getting it answer at a stats /
: math forum.
|
|
