Hi, I’m new to options trading and wanted to know what options pricing model do brokers use and does it differ from broker to broker? BSM, Heston model? Thank you.
Idk, I do know thinkorswim (schwab platform) has selection of either BS, Binomial, and Cox-rubentien-other name (sic) for use inside thier Analyze tab modeling section. I also have heard that all brokers will only say they use a “proprietary” formula for calculating the IVs shown on thier option prices tables.
AKdemy on
I’d say any reliable broker will tell you what they use. However, a retail brokers business isn’t accurate analytics. Professional Investors pay thousands a year to get the calculation right. E.g. professionals pay several thousands a year to get this done properly. Most basic traders have access to Bloomberg, which is 25k/year per user. BBG will offer decent pricing and market data. Swap rates will be great, dividend models and IV kind of OK. It will be insufficient if your business is primarily in options trading though.
Some will use just a closed form approximation (like Roll-Geske-Whaley and Barone-Adesi and Whaley, or Bjerksund-Stensland). However, the approximation is quite bad once you have discrete dividend payments.
Cox, Ross and Rubinstein (CRR), is frequently used because it’s easy to implement and accurtate. It’s a recombining binomial tree, and actually a particular case of an explicit FDM scheme (a special case of the FDM for the BS PDE). It’s in essence just a discrete time approximation of the continuous process underlying the BS model. Both approaches, CRR and solving the Black Scholes PDE, will agree with each other. The rest like Leisen and Reiner was just developed to speed up computations.
Nobody should use Heston for vanilla options.
What will matter more than the model will be market data (accurate risk free rates, not just treasury yields, proper dividend modelling, ….).
2 Comments
Idk, I do know thinkorswim (schwab platform) has selection of either BS, Binomial, and Cox-rubentien-other name (sic) for use inside thier Analyze tab modeling section. I also have heard that all brokers will only say they use a “proprietary” formula for calculating the IVs shown on thier option prices tables.
I’d say any reliable broker will tell you what they use. However, a retail brokers business isn’t accurate analytics. Professional Investors pay thousands a year to get the calculation right. E.g. professionals pay several thousands a year to get this done properly. Most basic traders have access to Bloomberg, which is 25k/year per user. BBG will offer decent pricing and market data. Swap rates will be great, dividend models and IV kind of OK. It will be insufficient if your business is primarily in options trading though.
Some will use just a closed form approximation (like Roll-Geske-Whaley and Barone-Adesi and Whaley, or Bjerksund-Stensland). However, the approximation is quite bad once you have discrete dividend payments.
Cox, Ross and Rubinstein (CRR), is frequently used because it’s easy to implement and accurtate. It’s a recombining binomial tree, and actually a particular case of an explicit FDM scheme (a special case of the FDM for the BS PDE). It’s in essence just a discrete time approximation of the continuous process underlying the BS model. Both approaches, CRR and solving the Black Scholes PDE, will agree with each other. The rest like Leisen and Reiner was just developed to speed up computations.
Nobody should use Heston for vanilla options.
What will matter more than the model will be market data (accurate risk free rates, not just treasury yields, proper dividend modelling, ….).