Subscription to options data costs many hundreds of dollars per stock – with Greeks, data since 2004 from the CBOE for SPX options is not far short of $400. It would be closer to 1,000 if they had data to offer back to 1985.
I am playing with synthetic data based (in the case of SPX at least) on the Black Scholes Merton process.
It all boils down to what volatility one is going to feed into the option pricing formula. Clearly it is NOT possible to faithfully replicate actual traded option prices. That would be cart before horse. The best one can do is to make some assumptions as to what the implied volatilities might have been on the given day and treat the whole exercise as a production of random prices which may have some use or significance in the production of back tests for some basic option strategies.
Back testing has little predictive ability anyway, so the use of almost randomly produced estimates of implied volatility/prices may have about as much validity as the use of actual option prices.
Implied volatilities for use in the calculations can be derived in many ways. Ideas include the use of historic volatility, the VIX futures structure (since inception of VIX futures in 2004) and the VIX index. Historic volatility could be differentiated in terms of using (say) 5 day historic volatility on a given day for input into an option expiring in 5 days, and so on out to expiries of two years out.
Rather than assuming such data or back tests using such data have any predictive value, the practitioner is better off assuming that such exercises have an educational function in better understanding option pricing and long term option strategies.
Treat it like a convoluted monte carlo simulation.
AM settlement is assumed on the expiry date. I have closely followed the CBOE specification for the price series.
Calculation of option premia is made on the basis of a European vanilla option. Prices at expiry are calculated by reference to the opening price of the underlying.
Data provided consists of:
- Trade Date
- Open – underlying
- Close – underlying
- The 3 month TBill rate
- Volatility of underlying (20 days, annualised)
- Dividend rate applied
- Call Price
- Put Price
Feedback (negative or positive) would be very welcome!