AuthorTopic: Any Fixed Income guru here?
sugar
@2004-08-20 11:32:27
Let's assume we have three bonds, each has 1 year tenor:

Bond A has 20% coupon paying annually, Bond B has 20% coupon paying semi-annually and bond C pays 20% coupon quarterly. Each is sold at par.

From investor's point of view the C's cashflow (5,5,5,105) is much better than B's (10,110) and A's (120 at the end of year). However the theory teaches us that cashflows should be discounted using the compounding with frequency equal to that of bond's coupon payments i.e. according to, say Fabozzi each bond has the YTM of 20%. That BTW is the simple evidence of the fact that all bonds are sold at par and has the same coupon rate.

So, the question is: given that there are no other instruments on the market (i.e. no spot rates, forward rates etc.) how to compare bonds having different coupon frequency. Say (X,X,X, 100+X), (Y,100+Y) and (100+Z) cashflows. Should YTMs be calculated using the N/Y compounding formula with N equal to bond's coupon frequency or N should be the same for all bonds, say semi-annual (2/Y) or equal to most frequent paying bond periodicity (4/Y in this case)? Probably continious compounfding should be used in this case?

Mahin71
@2004-08-22 17:13:15
I think you've highlighted the flaw of common approach as it does not consider the posibility of different periodicity of payments. I think you should use one discounting method (1/Y, 2/Y or 4/Y) for all the bonds however I can not say which of them exactly. Not sure about continious compounding. By the way, try to think what you will do with disount bonds.

Good luck