View Full Version : Questions from Hardy: Investment Guarantees

rekrap

01-24-2007, 12:55 PM

This would be a great place to pose questions related to material from the Hardy text.

Who knows, she may even answer your question herself (http://www.actuarialoutpost.com/actuarial_discussion_forum/showpost.php?p=1774381&postcount=9).

carzymathematician

02-22-2007, 02:19 PM

if you need to hedge with a short position in H shares , then you could do that with that many shares, or with that many forward contracts, or with e^(-rT)H futures contracts.

Why it is e^(-rT)*H future contract, not e^(rT)*H future contract ? Thanks

I am somewhat unclear about ur question. Pbly you could say what page and text you're making reference to.

Kazodev

02-22-2007, 02:36 PM

if you need to hedge with a short position in H shares , then you could do that with that many shares, or with that many forward contracts, or with e^(-rT)H futures contracts.

Why it is e^(-rT)*H future contract, not e^(rT)*H future contract ? Thanks

sorry if I'm completely off base here but a future contract is a promise to do something in the future, hence if the price of the future contract at time T is F_{0,T} then the PV of it (assuming cont. rate) is e^(-rT)*F_{0,T} right?

campbell

02-22-2007, 03:11 PM

Hmmm, I'm thinking of it the same way as Kaz. I've got to go home and reread this bit.

Futures contracts always confuse me, the way they "settle" every day.

rekrap

02-22-2007, 05:13 PM

So, according to Hull (http://books.google.com/books?id=THHLr9A4ffgC&pg=RA1-PA89&lpg=RA1-PA89&dq=hedge+ratio+futures+contract+hull&source=web&ots=O711GJ6xzB&sig=ZBBzAwWSflerSsmg-edekdpqtjI#PRA1-PA89,M1http://books.google.com/books?id=THHLr9A4ffgC&pg=RA1-PA89&lpg=RA1-PA89&dq=hedge+ratio+futures+contract+hull&source=web&ots=O711GJ6xzB&sig=ZBBzAwWSflerSsmg-edekdpqtjI#PRA1-PA89,M1) it's e^(rT), but according to Hardy it's e^(-rT)? :-?

Can you provide/cite specific pages in the Hardy text that introduces this issue? I'm having trouble (as are others) following you.

bagheera

04-17-2007, 12:48 PM

On page 53 in "Investment Guarantees", it's written toward the end "The MLE for the variance is n/(n-1)*s_y^2 ..The estimator for sigma is asymptotically unbiased but is biased for finite samples. The standard deviation of the log data, s_y, is an unbiased estimator of sigma for any sample size.

Just above this text "sigma_hat" has been defined with the n in the denominator.

Should the later expression have been s_y^2 = n/(n-1)*sigma_hat^2 instead of n/(n-1)*s_y^2? The use of s_y is confusing..What's the formula for s_y?

rekrap

05-09-2007, 06:29 AM

volatility clustering and bunching, are they the same? they occur for AR and Arch/GARCH. thanks.

Yes.

is these understand correct:

1. long term variance always comes with mean reversion and auto-correlation

2. auto-correlation refer to use past data to compute todays return/variance

3. what is volatility clustering, bunching, and stochastic volatility, are they the same?

4.what is the difference between auto-regression and auto-correlation? Thanks.:popcorn:

Didn't you take Course 4/C? :tfh:

Yes.

Didn't you take Course 4/C? :tfh:

do you sleep?

PS. I like the new avatar ;)

rekrap

05-09-2007, 10:11 AM

do you sleep?

PS. I like the new avatar ;)

Yes. :sleep2:

Thanks. :tup:

vBulletin® v3.7.6, Copyright ©2000-2018, Jelsoft Enterprises Ltd.