macrocephalopod

is that a bond bought at par does not appreciate much if held to maturity then reinvested. And if one is running a not-very-long-duration benchmark, pull to par can be fierce.

You didn't read the rest of the thread, did you.

Tl;DR of my comments below: Yes I did. I may have gotten muddled, but there is weirdness involved at very long duration on higher vol which bothers me with the analysis structurally, and now I'm gonna go bang my head against a wall.

All log returns so it decomposes into duration/rolldown nicely.

Very interesting that the 30y have the highest *rolldown* returns of the bunch. I would have expected them to have the same or less, more than compensated of course by duration returns.

There are two components, the pure rolldown effect (which is small for 30s and sometimes even negative) and pure yield level which is captured by duration shortening, which is more significant (that’s my intuition, haven’t checked)

I assumed that was where you were capturing carry. What is the periodicity of duration re-lengthening in that model?