The option analytics is a wonderful and much appreciated addition to the TWS platform. But of course there's always room for enhancements. The following is a few things I'd like to see in an upcoming version. Maybe def could relay this to the people in charge of this at IB.
- Option pricing model. Per the manual, IA uses plain vanilla Black-Scholes for European and a variation of Black-Scholes for American style options. Unfortunately, IB does not disclose exactly what approximation is used which makes me reluctant to rely too much on the values produced. I assume it's probably Whaley for stocks and/or Barone-Adesi/Whaley for stocks and indexes? It would also be nice if the binomial model (Cox, Ross, Rubinstein) could be added as an alternative option for american options in the future.
- Option model initialization. The initial values chosen by IA's vertical volatiliy skew model to compute theoretical option values are beyond anything even remotely useful with default volatilities in the 1600% range. Why aren't the volatility assumptions initialized to default to the average implied volatility when IA is used on a specific options class for the first time? That would provide much more reasonable default theoretical values. Of course, automatic adjustment to historical volatility would be even nicer but I guess that's beyond the intended scope of the application.
- Position risk graphs. It would be really nice if a future version could include position risk/profit charting. That would be a logical extension of the portfolio analytics already included in IA.
Dave
- Option pricing model. Per the manual, IA uses plain vanilla Black-Scholes for European and a variation of Black-Scholes for American style options. Unfortunately, IB does not disclose exactly what approximation is used which makes me reluctant to rely too much on the values produced. I assume it's probably Whaley for stocks and/or Barone-Adesi/Whaley for stocks and indexes? It would also be nice if the binomial model (Cox, Ross, Rubinstein) could be added as an alternative option for american options in the future.
- Option model initialization. The initial values chosen by IA's vertical volatiliy skew model to compute theoretical option values are beyond anything even remotely useful with default volatilities in the 1600% range. Why aren't the volatility assumptions initialized to default to the average implied volatility when IA is used on a specific options class for the first time? That would provide much more reasonable default theoretical values. Of course, automatic adjustment to historical volatility would be even nicer but I guess that's beyond the intended scope of the application.
- Position risk graphs. It would be really nice if a future version could include position risk/profit charting. That would be a logical extension of the portfolio analytics already included in IA.
Dave
