I'd like to implement the case where volatility parameters are time dependent.I search more information about branchings ,probability....Do we obtain a recombining tree with this method?

I had found this article...But when he says :" we match these by branching from x*i,j to one of x*i+1,k-1 x*i+1,k and x*i+1,k+1 where k is chosen so that x*i+1,k is as close as possible to x*i,j+Mi x*i,j deltaT....for node i,j+1 you won't necessary branch from x*i,j+1 to one of x*i+1,k x*i+1,k+1 and x*i+1,k+2 ...with the same k, will you?

Hi guys,I have a question on the trinomial tree of Hull and White. If we look at the H-W articles of the 90 94 and 96 or the book, they propose 3 branching alternative to avoid negative probabilities. We have to find the Jmax and Jmin and change the branching at these Nodes. I'm also looking at the article The general Hull-White and super Calibration. In that article they do not talk about the 3 different branching. Instead the central node is found by a formula and we continue the branching. in the exemple of the article there: 1 node a t=0, 3 node at t=1 9 node at t=2 5 node at t=5 Im wondering the relation between these two methods... are they equivalent if the spacing of time is constant. Thank you. If you have any other refernces on the calibration please let me know...

QuoteOriginally posted by: jonath024Yes, I believe it is recombining.Have you looked at Hull & White "Using Hull-White Interest Rate Trees", Journal of Derivatives, Vol. 3, No. 3, (Spring 1996), pp. 26-36 ? It is available at John Hull's web sitehttp://www.rotman.utoronto.ca/~hull/Downloadab ... ions/There is an example of fitting a time dependent vol structure in this paper.Jon Does anyone has example of implementing hull white tree with time-dependent vol ? please help!!!!

There is a better Gaussian model that allows calibration without trees. See Andersen and Piterbarg 'Interest Rate Modeling, Volume 2' Chap 12.