Forward FX rates

krk · July 20th, 2006, 8:27 am

Thanks for the answer. If I understand you correctly, the approach 2 for getting present value in EUR of a future cashflow in another currency I have described below is the correct one, provided I use the correct discount curves, including the information for SN / ON rates.

arpitbhatnagar1 · July 20th, 2006, 8:35 am

but for valuations will it be right to use diff curves ie df 1 on today and df 1 on spot for fx instruments and swaps respectively?

krk · July 20th, 2006, 12:01 pm

I cannot believe that there is no market convention on such a simple question like how to find the present value in currency 1 of a future cashflow in currency 2....

arpitbhatnagar1 · July 21st, 2006, 6:59 am

u can also assume df at t, t+1 and t+2 are all 1.

johnself11 · July 22nd, 2006, 10:20 am

ok so your model says that it is p/l neutral to lend 1 billion dollars today and receive 1 billion dollars in two days...hmmmmmmmmmmmm

arpitbhatnagar1 · July 22nd, 2006, 12:39 pm

i know its not the best assumption, but then should i have different curves for swaps and fx instruments?

johnself11 · July 22nd, 2006, 1:00 pm

i suggest using two curves: a pricing curve for generating forward rates (df=1 at t+2) and a discounting curve to get the pv of cashflows (df=0 at t)

arpitbhatnagar1 · July 22nd, 2006, 3:28 pm

i am a bit naive so hope u won't find it a prob to answer....if i am discounting back to today then it will result in another problem...most of the swaps are spot started rite! so how can i discount them back to today!

johnself11 · July 22nd, 2006, 5:34 pm

good point, having a pricing curve and discounting curve will mean that a par swap does not value exactly to zero.... in fact this will be even more pronounced if you add a libor spread to your funding curve to refect your institutions credit and cost of funds.... this, however is just a books and records issue.... the practice i suggest is quite useful when valuing off-market trades with clients of lower credit....

arpitbhatnagar1 · July 22nd, 2006, 7:00 pm

this way someone can make money even on day 0...i.e. the day the swap is initiated...similar to using a spread rite!

johnself11 · July 22nd, 2006, 7:15 pm

well if you are using a funding cuve which is different from you forward rate curve (as i suggest) then yes there will be a p\l effect on swaps which are technically "at market"....no one is "making money" here.... a par swap should only have zero NP if you are willing and able to borrow and lend at libor

arpitbhatnagar1 · July 23rd, 2006, 6:54 am

but i can always take an opposite position where i am actually paying/receiving par (the swap rate using df 1 at spot) and make money out of the spread between the 2 swap rates!

johnself11 · July 23rd, 2006, 9:03 am

indeed you could do this p/l vodoo but you are not "making" anything except a painful situation when you have to explain the b.s. p/l to your boss later on.... morover most swap traders manage matched books in which any slightly off-market-valued trades will wash out.....

arpitbhatnagar1 · July 23rd, 2006, 2:12 pm

i don't have any hands on experience with real trading...i was thinking more so from a theoretical angle!but it will be really difficult to tell later on i.e. cause i can just use that for getting a price (swap rate) bt then mtm the swap keepin df 1 @spot..

johnself11 · July 23rd, 2006, 3:22 pm

understood but just think about it....if the market value of the swap is zero and your books and records (because of funding issues) value it slightly different than zero, when you unwind or offset that position in the market, you will negate that imaginary p/l..... if you do not offset of unwind it, the bogus positive value may become an internal "funding arb" against your cash desk but that is just transferrence of value from one member of the team to another....