Hi,This is a bit of a basic question relating to interest rate swap valuation (vanillas).Where is the risk in the swap, is it in the fixed leg or the floating leg?(1). If we view the swap as a fixed rate bond and a floating rate bond then the majority of the risk would be in the fixed part of the swap as (apart from between reset dates) the floating leg values at Par and so has little duration/PV01 etc.To do this we would also assume that notionals are exchanged.(2). If we assume that the swap is just streams of payments of fixed and floating rates (and there is no exchange of notionals) then the majority of the risk is in the floating leg.So which way is the right way to think about the swap or are both ok?Further - is a rough proxy for the PnL from a 1bp change in say the 5y swaprate on a 5y swap approx sum of (accrual factors_i * DF_i) for i = 1y,....5y ? (I am asking this because if we construct our discounting curve out of swaps as well, shifting the 5y swap rate by 1bp should also move the discount factor).Thanks - sorry if these questions are a bit daft - just a bit confused right now.DocToc

On thinking about this a bit more...Doesn't this really depend on how I wish to bucket my risk. Say i construct my discounting curve using (in addition to other instruments) the 5y swap - then i would want to see my risk in the 5y bucket as a result of trading this 5y swap. So constructing my curve like this I would want to see the change in the swap value with respect to the 5y swap rate, other things such as Euribor futures 3y swap rate etc should have some delta, but minimal as these really only affect the value of the swap through the discount factors.Just another thought, would still appreciate experienced swaps/options/exotic rates traders to share their opinion.Thanks

please see this short thread ...http://www.wilmott.com/messageview.cfm? ... SGDBTABLE=

Hi Dave I agree with your previous answer in cases of notional exchange.As in my point (1) the risk is in the fixed leg BUT if we assume no notional exchange then the riskier leg is suddenly the Floating leg. Have you got a mathematical/intuitive reason for this ?Thanks again

I thought thats what my little bit of arithmetic did...

When I say notional exchange I mean:(1). view the swap as a fixed rate bond (so paying principal say 1M) back at the end + short floating rate note (paying 1M) back at the end (here the risk is in the fixed leg)(2). view the swap as a fixed rate bond (no principal back at the end) + short floating rate note (no principal back)I understand the principal exchange (as in most swaps doesn't happen) and has no economic value - because you recieve via one and pay the other so npv = 0, but how you view the risk changes right?I might be completely off - let me know your thoughts.Thanks

no i think you are on the right trackyou think of a swap with p/ex as long position in a bond financed with a frnswap = B - FRNnow you short a zero Z against your long bondswap - Z = (B-Z) - FRNswap = (B-Z) - (FRN-Z)or (B-Z) is the fixed leg and FRN - Z is the floating leg. you should be able to convince yourself that B-Z has a lower duration than FRN-Z.

Right ok - thanks.so in that case: PV(Swap) = sum_(i = 1 to n) *accrual factor * DiscountFactor(i) * Notionalso my "Swap Delta" or whatever the risk to a 1bp change in S (the n year swap rate) is... sum_(i = 1 to n) [1 bp]*accural factor * DiscountFactor(i) *NotionalAgain another problem arises here (maybe its a valid approx) but assuming that this swap pays annually with S% and the forward Libors L% are all paid annually too, DiscountFactor(n) is not independent of the swap rate, but in my approximation above I have assumed it is - so d (DiscFactor(n))/dS = 0Is this a valid statement (assuming that I do build my curve using the n year swap?) or proxy for viewing my swaps' delta/PV01?Thanks again

daveangel, can you recommend a book on these topics of risk and hedge of swaps? eg hedging swap portfolio by bucketing to market risk factors, Xccy swaps, etcI've looked at Pricing, Hedging and Trading Financial Instruments by Carol Alexander but its more of an overview

Have you looked at Bruce Tuckman's book?

No I haven't seen that book, thanks for pointing it out to me, I'll take a look

Turns out Tuckman's book doesn't have anything on Xccy swaps or FRNs. Is Stigum's Money Market any good?Is there a practical hedging book for short term interest rate trading & fx forwards similar to Taleb's Dynamic Hedging? Just interested in something that more clearly lays out the the math between market quotes on basis swaps, valuing a portfolio of existing swaps in multiple currencies, and risk and hedging that portfolio. Like, here's how you bootstrap the forward curve, the discount curve, and why. Preferably with considerations for funding and collateral, and moving to OIS discounting. I can't determine if a book like this exists, maybe this is just industry knowledge, not really codified?

Try Flavell.

QuoteOriginally posted by: qhedgeTry Flavell.+1