Hi,I have calculated convexity corrections under a Hull-White process based on a formula in James & Webber. For short maturities, the correction looks reasonable, but for longer maturities (> 10 years) the corrections becomes very high. For maturities between 40 and 50, the correction is even 5%!Now i read that the formula's for convexity corrections are not working very well for longer maturities. I believe that, but what correction can i use then for longer maturities to account for this effect?Greetings,Richard

the simplest and most naive implementation is from Hull's book itself. There is no way that you can get a 5% correction. For a 20 year case,implement the Hull's formula(it is a wrong way to get into the groove,though) in a spreadsheet. It should not take you more than two days to key in the long differentiations(single and double). You must be careful about the percentages(square of a decimal number will be smaller than the number)regardsonce you get this one then implement Hagan (that gentleman in SABR model,you know I hope, very good in maths stuff) or Pelsser's formula(this gentleman is a professor)cheers

You might look at this paper or their book.Hunt Philip, Joanne Kennedy 'On Convexity Corrections' 3/98

Hi,The convexity correction formula from hull's book has an implicit assumption of a lognormally distributed rate. In Hull-White, the (zero-)rates are normal.How does this formula look like when a normally distributed rate is assumed?Greetings,Richard

The Linear Swap Rate Model seems to give larger corrections than the traditionnel adjustment. And the difference increases with volatility which is an additionnel problem.Is there another kind of adjustment in that case ? (High volatilities)

Have you tried computing when ? In my mind, that term should be equal to , rather than (which is the case for GBM)It may be totally off-course so apologies if so!