Serving the Quantitative Finance Community

 
User avatar
kokoon
Topic Author
Posts: 0
Joined: September 18th, 2003, 11:30 pm

Dividend problem?

May 12th, 2004, 3:25 pm

I need to calculate IV from option prices. Since the volatility I want to calculate are five to ten years. I don't have any dividend data for this period, the only dividend quoted are for a few months. I would like to know how I could manage this problem. Any article would be welcome.Thanks in advance.KOKOON
 
User avatar
kokoon
Topic Author
Posts: 0
Joined: September 18th, 2003, 11:30 pm

Dividend problem?

May 13th, 2004, 6:41 am

There are about 50 views, but no reply. Has anybody ever encountered this problem?
 
User avatar
Watchman
Posts: 0
Joined: November 13th, 2003, 1:52 pm

Dividend problem?

May 13th, 2004, 7:16 am

if you've got enough prices, back out the divs from the calls and puts. If not, I don't think there is a definitive solution. You'll have to make some assumptions, potentially trying a few different scenarios with respect to how the dividends might grow (or not) over the next few years.
 
User avatar
capcapo
Posts: 0
Joined: July 14th, 2002, 3:00 am

Dividend problem?

May 13th, 2004, 11:46 am

If dividends are not too high call/put parity gives a reasonable approximation
 
User avatar
Graeme
Posts: 7
Joined: April 25th, 2003, 5:47 pm

Dividend problem?

May 14th, 2004, 7:12 pm

The put-call parity idea is nice, but it may have the following problems1. non-contemporaneous put and call prices2. using a continuous dividend yield in general is 'wrong', see the article by Lewis and 2 * Haug, in Wilmott mag, for example.You might need to create a model of how dividends will evolve in the future. There are very simple models, such as that the dividends which have recently occured will repeat (modulo grossing up for time value, etc). There are more sophisticated models. I know there is a model in PW's two volume set, which looks interesting, but I haven't looked at it closely to make any comment besides just that. As above point 2., the required outputs may need to be discrete rather than continuous.
Last edited by Graeme on May 13th, 2004, 10:00 pm, edited 1 time in total.
 
User avatar
Gmike2000
Posts: 0
Joined: September 25th, 2003, 9:49 pm

Dividend problem?

May 15th, 2004, 9:14 pm

you cannot use actual dividends paid. the market prices expected dividends.you need to back out the expected dividend yield. if you have spot and futures price for the underlying asset (recorded on the same date as the optinos), you could use spot/forward parity and back out the dividend yield priced in. You don't need to know the risk free rate....just use whatever you get as the yield differential, because this is what you feed into the black sholes formula to get implied vol.if you don't have a futures price, use put call parity to back out an implied futures price. then use spot/forward parity F=S*exp([r-d]*t) to get the expected dividend yield. For the put call parity, use only liquid options.
Last edited by Gmike2000 on May 14th, 2004, 10:00 pm, edited 1 time in total.