I want to model stochastic interest rate for my option pricing software. I think the simple Vasicek model should be good enough for my purpose.According to my understanding, I can obtain the spot rate r(t) for a particular date with these steps:1. obtain deposit rates and swap rates for that date2. Fit a Nelson-Siegel curve to it.3. r(t) should then be the yield when t=0By obtaining a time series of r(t), I can then use the method described here:http://www.sitmo.com/doc/Calibrating_th ... ck_modelto calibrate for the parameters of the Vasicek model.Do you think I am going in the right direction? Also, since I am using real world data, do I need to adjust for market price of risk? If so, what will the new stochastic process look like?Thanks a lot in advance!

You use the yield curve slope and calibrate to implied vols. Nowhere should you calibrate to historical data. Consult the book by Martingale Methods in Financial Modelling for more details.

I bet you've done quite a research on this topic, and I believe your method is feasible. However, it seems a little bit devious using estimated short rates to estimate Vasicek model. Firstly, this might induce seriously deflected estimates since estimated short rates might've contained systematic errors; Secondly, it has been advocated for a while not employing N-S model to capture the over time dynamics of forward curve, because it has been proved (Filipovic (2000)) that N-S is not compatible with any nontrivial term structure of interest rate models, the over time dynamics results from N-S might be problematic. And we could avoid these problems by directly estimating Vasicek model from zero bond prices (or LIBOR&Swap rates), since based on Vasicek model we have closed form expression for bond prices. Equating the model prices to the market data in terms of least square will give you the parameter estimates (including the market price of risk parameter).

http://www.quantnet.org/forum/showthread.php?t=3977If you read that thread, you'll find matlab code to do what you are asking about (fit the Vasicek model to a yield curve). I implemented this for a thesis I completed in 2008.

What do you want to do with your model?

yes you can, but it is more appropriate for multifactor modelscheck that papers by Chen and Scott, and also Duan and SimonatoKyriakos

Also I guess your approach will calibrate to the true dynamics rather than the risk neutral ones (which you need for option pricing)K

QuoteOriginally posted by: RezAlso I guess your approach will calibrate to the true dynamics rather than the risk neutral ones (which you need for option pricing)KHi, I have the same problem of the thread starter.I have to price stock options under a stochastic spot interest rate described by the Vasicek model. I estimated the physical (true) Vasicek parameters by using daily observations of the 1week-Euribor from 1999 to 2012 (did I do it right?). The estimated long-term mean should implicitly include the price of risk. How can I obtain the long-term mean under the risk-neutral measure? Thanks in advance.