Re: [Quantlib-dev] Question about ZeroCouponInflationSwap

[Chris K. <chr...@ya...>]

Hi Alexander,

yes, for ZCIIS you are correct - thanks for
catching it.  The tests are ATM so this didn't show up.

For inflation indexed bonds then you do need an
IndexedCashFlow that does not have the -1, so both versions are
required in general.  Both IndexedCashFlow and ZeroCouponInflationSwap now updated in the SVN for the 0.99 branch.

The interface to IndexedCashFlow has changed from:

IndexedCashFlow(Real notional,
                        const boost::shared_ptr<Index> &index,
                        const Date& baseDate,
                        const Date& fixingDate,
                        const Date& paymentDate)

to:

IndexedCashFlow(Real notional,
                        const boost::shared_ptr<Index> &index,
                        const Date& baseDate,
                        const Date& fixingDate,
                        const Date& paymentDate,
                        bool growthOnly = false)

The bond-focused version (without the -1) is the default, and ZCIIS uses it with growthOnly = true so the desired effect of I(T)/I(0) - 1
happens for the swap.  

Best,
Chris



________________________________
From: "ale...@go..." <ale...@go...>
To: chr...@ya...
Sent: Tue, December 15, 2009 7:35:33 PM
Subject: Question about ZeroCouponInflationSwap

Hello Chris,

I priced a zero coupon inflation swap and can't explain the both legs.

I've noticed that the inflation leg is priced using the formula of the IndexedCashFlow

notional * I(T1) / I(T0)

But in papers and books I found the following formula

Inflation Leg = Notional * I(T1)/I(T0) - 1  

So I am missing this -1. Am I wrong?

Best

Alexander