I am a bit confused about how I should calculate expected RPI/CPI given a zero coupon inflation curve. The issue I am having difficulty with is the lags... Say we want to calculate the RPI for a date T, say, and we know the inflation rate for that date (r1) and for the date that is say n months before (r2). We dont however know todays RPIm but we know todays reference RPI (call it rpi0) because we have some historical observation of RPI. My current thinking is that the expected rpi is given byrpi = rpi0 * exp(r2*(T-n))or shoudl it be rpi = rpi0 * exp(r1*T)because rpi0 has already been lagged ?

The UK DMO have some stuff on this. Look for anything by Deacon & Derry. May not quite do it for you but worth a look perhaps.Mark