2
$\begingroup$

I am writing an option pricing model for production use. Its not for arb or anything so it doesn't need to be 100% as accurate as possible. Just good enough for "what happens to my book if we jump 10 handles sorta thing."

I want to preface this with, I have to price most of my options same day expiry which kind of leaves out Quantlib or something of the sort unless I want to rewrite a lot of it and recompile it.

1.) I am pricing (futures options) on SPX, Nasdaq, Crude, NatGas 10year treasury note, long term treasury.

2.) What models should I be using? I realize for commodities I need to include cost of carry, but is there a model out there besides Binomial trees that I can use to price the commodities?

3.)As far as dividends, if Im pricing SPX Futures options is binomial tree the best way to do that? and do I really need to price in every discrete dividend into the BT model at the exact date and time it happens? Or will continuous/no dividend be accurate enough?

4.) Are there any additional caveats to pricing same day expiration options? - Currently I'm not sure what to divided the hours left in the day by for my time in the pricing model. I believe it would be from 6pm(open) to 4pm(settle) which is 22 hours if its under 24 hours to expiration. that would open some problems when pricing between 4-5pm say on a tuesday for wednesday expiration but that can be solved pretty easily.

Thanks

$\endgroup$
19
  • $\begingroup$ 1) How could it be for production and not trying to be as accurate as possible ? Sorry your use case is very unclear to me 2) spx options are european 3) same day option will have no issue with dividend since those have dropped by morning time 4)you need to price settlement correctly tho 5) other stuff (risk etc) but its unclear again what you want to $\endgroup$
    – Ezy
    Commented Feb 16, 2019 at 16:52
  • $\begingroup$ @Ezy I was talking about ES options. I need to be as accurate as possible but its not like im doing option arb and my income depends on my models being accurate as possible. This is just for risk metrics for my book. $\endgroup$ Commented Feb 17, 2019 at 3:24
  • $\begingroup$ and what makes you think you need to be less accurate for risk when you are trading 1 day options ? $\endgroup$
    – Ezy
    Commented Feb 17, 2019 at 12:55
  • $\begingroup$ @Ezy im just saying I don't need to be as accurate as say an option trading firm. I just need a general gist of how much ill be down or up. $\endgroup$ Commented Feb 17, 2019 at 14:51
  • $\begingroup$ Your options are european so you can just use BS to get all your greeks. Just be careful with settlements, you need to take them proerly into account $\endgroup$
    – Ezy
    Commented Feb 17, 2019 at 15:15

1 Answer 1

1
$\begingroup$

1) First of all many future option contracts are European, so for those there's no modeling problem. Just use BS. Now certain contracts like quarterly ES option are american for historical reasons

2) Futures do not entitle the long holder to the dividends of the underlying. That's the difference with other type of derived instruments (like SPY or QQQ) and that's also the reason of the existence of the cash-future basis. So for the purpose of pricing american options on futures dividends play no role whatsoever.

3) The source of the early exercise opportunity for american options on futures lies in the possibility of positive cash flow through the margining account. This impacts both the call and the put. By contrast it is never optimal to early exercise an american call option on the forward contract.

4) for some reference on american options on futures you can look here

5) now I have not seen the result anywhere but I believe that it is never optimal to exercise the american option intraday. Therefore if you now restrict to same day expiry option i claim they are fairly priced as european options.

In conclusion for your specific purpose I think you would be good with simply pricing all these instruments as european options when they expire the same day.

$\endgroup$
4
  • $\begingroup$ Interesting. Logical and concise. Thank you for the answer after our back and forth. Now what about 1day-1week. I should mention these are the weekly options on the futures where applicable. Lets say ES for now. $\endgroup$ Commented Feb 18, 2019 at 18:39
  • $\begingroup$ np. well for longer expiries now the early exercise is a possibility so you could use any numerical method of your choice or perhaps some analytical approximation. There's no general answer to that. $\endgroup$
    – Ezy
    Commented Feb 18, 2019 at 18:42
  • $\begingroup$ Okay, makes sense. Im about to post another question about calculating greeks and their intraday characteristics. Do you know much about that? $\endgroup$ Commented Feb 18, 2019 at 19:10
  • $\begingroup$ i am sure there will be plenty of knowledgeable people around here who can answer too. meanwhile feel free to upvote and accept this answer if it seems acceptable to you $\endgroup$
    – Ezy
    Commented Feb 18, 2019 at 19:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.