While trying to model interest rates using the CIR model, which of these are used? dr = a(mu-r)dt + sqrt(r)*e*sigma* sqrt(dt) or dr = a(mu-r)dt + sqrt(r)*dwn*sigma What is the difference using the 2? The R package uses the second of the above expressions, but I've come across an excel illustration with the first. Specifically, which one is to be used in what situations?
Sweeting, McNeil Frey and Embrechts and the IFoA Tables all use versions of the second formula. CIR was covered in CT8 (Chapter 17) Simon
Our trusty CT8 tutor has pointed out that they are in fact the same. In the 2nd one we have dWt ~ N(0,dt) The excel version is just starting with a standard normal variable, e ~ N(0,1), and multiplying by sqrt(dt) to get the same distribution.