12M EUR/USD FX swaps are currently -80. Calculating the theoretical value using 12M USD rates of 1.5% and 12M EUR rates of 0.80% would return -96. In order to make the swap -80, the EUR rate would have to be reduced by approx. 13 bps. This difference of -13 pips should be due to basis but the basis on 1 year EUR/USD XCCY swap is currently -25 bps.Why the difference? It has been larger a few months ago.

Hi,Been a while since I've done this but as a (very) rough back of the envelope, think about the following set of trades:1) 1yr EURUSD Xccy Basis swap: pay 3m Euribor +s vs receive US Libor flat with notional exchange at beginning and end. On bbrg I see -27.5bps for his.2) 1yr EURUSD FX Swap: Buy & Sell EURUSD with the interest amounts being paid on the far leg of the swap (at effective interest rates X$ and Xe) at -81.7pts.3) 1yr US IRS: pay 3mL vs recieve Annual fixed at rate R$ = 0.376%4) 1yr EU IRS: receive 3mE vs pay Annual fixed at rate Re = 1.236%If you draw the cashflows for these swaps you'll see that everything cancels out apart from a quarterly debit of s (the basis spread), a final debit of (X$-Xe) [plus the final quarterly s], and a final receipt of (Xe-Re).The choice of USD discount curve in this case is purely arbitrary as we are looking at the relative spread, so you can assume that X$=R$ (again, this is a very rough approximation just so you can see how the mechanics work, to do this properly you should build all the curves properly). In which case the equation for the NPV reduces to:Sum(s*dFi) ~ (Xe-Re)*dFTwhere dFi are the quarterly discount factors and dFT is the final discount factor.Feeding the US swap rate into the standard FX Fwd formula (we are using a purely arbitrary rate as above) you get an estimate for the implied EUR rate Xe, which using the above at I make to be 0.965%. Comparing that with the EU swap rate gives -27bps which is pretty much where the basis swap is quoted.

Sorry, I forgot to mention, I think the reason why yours is so different is that you are using 12m deposit rates which have an embedded premium over 3m rates, and for the arbitrage constraint to apply you need to replicate the cashflows of the basis swap which means you need to use Annual fixed vs 3m floating swaps in the calculation (a 12m deposit wouldn't match off the quarterly payments). My rough example doesnt account for the value of the quarterly basis spread payments, but they don't alter the breakeven rate that much, but you should be able to calculate it properly with an appropriate discount curve.

Good reply. Thx very much

Hello,How would you hedge the RESET risk, risk materialising whn you reset the Spot rate every quarter and exchange the MtM withthe counterparty...well what's the risk in the first place : spot risk and basis risk on the amount of cash that is exchanged at each reset, is that correct? How do you hedge this?Cheers

there wouldn't be any net resets, the floaters from the xccy basis swap would wash against the floater from the interest rate swaps