3
$\begingroup$

What's the correct discount curve to use for exchange traded products? Would these be discounted at the OIS rate (because of the central clearing house)?

E.g. the E-Mini S&P500 Future @ CME: I'm trying to model liquidity preferences and implied dividends based on listed futures and options- so I need an accurate curve for the index's discounting. Some back-of-the-envelope tricks (treasury curves + basis) don't show enough structure in longer durations.

The S&P500 future is a unique market with its own characteristics. What about an equity forward curve with uncertain liquidity and dividends? Should I compute the repo cost (could be hard-to-borrow) of the stock from box spreads in order to discount the single equity (vs discounting the index)? Is there a better method in practice?

Collateralized OTC products are discount at OIS because this is the rate paid on the collateral. What about exchange products? What about equities with interesting financing characteristics?

Edit: is it the case that the financing I am giving up is the secured overnight rate for centrally cleared products when trading the index, so I should use the risk-free rate? What about for stock options with more interesting borrowing markets?

$\endgroup$
  • 1
    $\begingroup$ what are you discounting? S&P500 futures are immediate instruments, if you buy at 2850 and sell hours later at 2900 the gain is paid immediately (end of day settlement). If you are inferring the discount factor used to price the instrument in the first place (i.e. why it is at 2850) then I suspect that is related to the industrial financing available on the basket of stocks. Much like a bond future is priced by considering the repo rate applicable to the underlying cheapest to deliver bond. $\endgroup$ – Attack68 Aug 7 '19 at 17:33
  • $\begingroup$ In both cases, I am doing research on the options market written on the future and on the equity (the hypothetical stock with peculiar repo curves). I think this is the same discount curve for the options and the underlying (due to risk-neutral pricing and the change of numéraire). $\endgroup$ – Jared Aug 8 '19 at 12:30
  • $\begingroup$ @jared are the options margined? $\endgroup$ – will Aug 11 '19 at 11:04
  • $\begingroup$ @will yes they are, also centrally cleared $\endgroup$ – Jared Aug 11 '19 at 22:10
1
$\begingroup$

This question Setting the r in put-call parity? shows the details are subtle and nuanced, but the answer is to use the rate paid on the collateral or margin.

$\endgroup$
1
$\begingroup$

if you are asking how CME collateral is discounted, then you have two considerations:

  1. What does the CME give you on your USD cash? That's simple, it's OIS. You don't get the interest immediately, but instead I think once a month. I'm not sure if they compound it - but I would imagine they do as it references OIS.

  2. What's your funding situation. For example, your treasury might charge you a spread of OIS+20 on the collateral. So, when modeling the convexity factor of a future, you need to incorporate the spread if you want to be super accurate. But usually, for things like ESU9, it's less than 1/100 of a bp, so you can throw it away.

Now, if you are asking, how to compute funding cost for the underlying reference asset, that's different. Options with early exercise are more complicated - but European options and futures are simple. Front month dividends are fairly well telegraphed. On Bloomberg you can simply enter "ESU9 [Index] FAIR":

ESU9 Index FAIR

Just pay attention to section 14, in the middle, where it shows ESU9 in blue. That shows a four cent difference between spot and futures, $3.51 in divs, and an implied rate of 2.30.

For a base rate reference itself, pick your poison: OIS, Interpolated Libor, 1m libor, o/n libor, your personal credit card rate, etc...

Does that help with your question?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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