In the attached file, it mentions a "repo spread" rate to be added to the continuously compoundingrate. Question 1: How can I extract the repo spread from the all-in repo rate in the market?Question 2: How can I apply this discounting for all the cashflows as the repo market is only short-dated?Also, a trader I am working with likes to simply subtract the Repo Rate from the Bond's Coupon andprice a bond with that subtracted coupon to give him a sense of how much money he is going to make.Question 3: Is this the correct way to think of it? Is it simply an estimate? Changing the coupon doesn't affect the principal - is that ok?Question 4: Can someone explain the difference between the method described in the lecture notes attached and the trader's intuition?RNP

Sorry, I wasn't allowed to upload pdf or gif files. Basically, the modified DF was adjusted like:Z(t,T) = exp[ - Integral(t,T){ r(x) + r_repo(x) } dx ]Hope that helps.

Only zipped files can be uploaded...

I'm guessing by repo spread you are talking about the specialness of a repo. ie, right now in the Canadian market, the overnight rate (continuously compounding rate) is 4.25. However, for the 10-year bond that is cheapest-to-deliver, the repo rate is like 4.10, making the repo spread -0.15. Though I also see people talk about repo spread as in spread from LIBOR borrowing (unsecured) to borrowing costs in the repo market (secured). Basically, at the end of the day, what's in the integral (r(x) + r_repo(x) should be what you can actually get money at (or lend money at).In either cases, it's pretty tough to model the repo spread. I don't know of any work in this area. I don`t know, just try and estimate via historical average maybe.For Q3, I would much rather use yield-to-maturity instead of coupon if they intend to hold to maturity. For instance, how much would he make if he bought a zero-coupon bond? Using only coupon misses the pull-to-par effect. In most cases, though, this is negligible.