incorrect treatment of sigma??

[vvaidy]

Hi,

I suspect in your CIR model you are forgetting to scale the brownian motion term. If I am reading your code correctly, instead of:

dr = K*(theta-rates[-1])*dt + sigma*np.sqrt(rates[-1])*np.random.normal()

I would have said:

dr = K*(theta-rates[-1])*dt + sigma*np.sqrt(rates[-1])*np.sqrt(dt)*np.random.normal()

What am I missing?

-VV

[jamesmawm]

you're right, please scale it.