3

Could someone tell me a common method for pricing VIX Options please?

Do I need to use a Stochastic Vol Model? Or Local Vol Model is suitable as well?

Should the model be modelling S&P 500 and then recalculate the VIX Index? Or model the VIX Index directly?

Thanks!

chengcj
  • 483
  • 4
  • 10
  • Do you understand exactly how the VIX index is derived? – Matt Wolf Nov 06 '13 at 14:34
  • Market makers commonly just treat these as "normal" options that happen to have an unusual skew. This is a roundabout way of asking: what will your hypothetical pricing method be used for? – Brian B Nov 06 '13 at 16:19

2 Answers2

2

I am ignorant to VIX option pricing but I would start with this: The performance of VIX option pricing models: Empirical evidence beyond simulation -- Working paper version

CharlesM
  • 493
  • 4
  • 14
1

There's ongoing debate about whether SPX and VIX options contain different valuable information (see Bardgett Gourier Leippold (2013) - Inferring Volatility Dynamics and Risk Premia from the S&P 500 and VIX Markets among recently contributions). I would say this is slightly counterintuitive (VIX index is calculated as a portfolio of SPX vanilla), but at the same time, is the fact that VIX derivatives provide a pure volatility exposure (and traders know it much better than me) that make them conveying a different information, at least concerning volatility dynamics. This said, I would use a consistent model calibrated on both markets. Moreover, in the context of VIX options, there's no Dupire equation (linking VIX option Prices to a VIX implied local volatiliy), but - in some sense parallel - there's an approach called the standalone approach (see Mencìa Sentana (2012) for an update reference), in which one models directly the VIX (without care of the S&P500 market). I would use this If you need a pure pricing machine. But, a structural model as the one I mentioned before is needed to have a perspective of the market as a whole (e.g. estimate risk premia). To conclude, you could off course even simply interpolate the VIX implied surface as is common practice in SPX context. But I don't think this what you meant.

Gabriele Pompa
  • 700
  • 5
  • 16