Key Rate Durations

They are essentially zeros, but with exposure to prepayments. You are correct in that the value of a PO will rise as rates fall (higher prepayment speeds lead to earlier return of principal), but that does not mean that the KRDs are all positive. Some of the lower KRDs are negative according to our readings, however the much larger (in absolute terms) and positive KRDs near the maturity date dominate.

ws Wrote: ------------------------------------------------------- > baocarol Wrote: > -------------------------------------------------- > ----- > > PO has negative key rate duration in the short > and > > intermediate term and turn to positive key rate > > duration in the longer term. which because when > > yield falls, people hold MBS more like to > > refinance, and prepay the principle, which > caused > > PO MBS value underperform compared to bullet > bond, > > but for older PO MBS, the MBS holder may not > > refinace due to the less loan left at the end. > > when yiled increase the PO will perform the > same > > as bullet bond. > > > I thought PO acts like discount bonds, you buy > them 90, 95 cents on the dollar. Therefore, I > would think when rate drops, people are more > likely to refi, therefore principle gets returned > faster. Therefore, we can observe a price > increase in PO. That is still positive key rate > duration. That’s positive effective duration (or something). Key rae durations depend on moving a particular yield (like 3 months).

I know that…however, I when I read the material on CFAI and Schweser, looked at the chart, still doesn’t make a whole lot sense to me, since there was a lack of detailed explanation. I am eager to find out too.

ymmt Wrote: ------------------------------------------------------- > They are essentially zeros, but with exposure to > prepayments. You are correct in that the value of > a PO will rise as rates fall (higher prepayment > speeds lead to earlier return of principal), but > that does not mean that the KRDs are all positive. > Some of the lower KRDs are negative according to > our readings, however the much larger (in absolute > terms) and positive KRDs near the maturity date > dominate. I think I somewhat know what you mean. So, for shorter term (loan is about to mature, ie, I have 3 years left on my mortgage), even there is a rate drop, people are LESS likely to refi because loan is about to mature (less time). For longer term (ie. I have another 15 years left on my mortgage), with a rate drop, people are MORE likely to refi. Even with people LESS likely to refi with only few years left on their mortage, I am still puzzled that why would a PO move in the same direction with rate??? Because at this point, since people are less likely to refi, the prepayment risk is reduced, they will act MORE like a zero. Thanks

I think I know what this is. Holding PO strips is like holding call options on long-term bonds. In fact, the only reason that they have such high effective duration is that mortgage holders hold calls and by buying PO strips you are buying that optionality. That means for early krd’s, the interest rate sensitivity is like rho of the bond call. That gives the negative krd for short maturities and macthes up pretty well with the magnitudes given by ymmt’s desk.

ws Wrote: ------------------------------------------------------- > ymmt Wrote: > -------------------------------------------------- > ----- > > They are essentially zeros, but with exposure > to > > prepayments. You are correct in that the value > of > > a PO will rise as rates fall (higher prepayment > > speeds lead to earlier return of principal), > but > > that does not mean that the KRDs are all > positive. > > Some of the lower KRDs are negative according > to > > our readings, however the much larger (in > absolute > > terms) and positive KRDs near the maturity date > > dominate. > > > I think I somewhat know what you mean. So, for > shorter term (loan is about to mature, ie, I have > 3 years left on my mortgage), even there is a rate > drop, people are LESS likely to refi because loan > is about to mature (less time). For longer term > (ie. I have another 15 years left on my mortgage), > with a rate drop, people are MORE likely to refi. > > Even with people LESS likely to refi with only few > years left on their mortage, I am still puzzled > that why would a PO move in the same direction > with rate??? Because at this point, since people > are less likely to refi, the prepayment risk is > reduced, they will act MORE like a zero. > > Thanks I think we are talking about krd for PO’s from newly minted long-term mortgage pools. At least the Ho article above gives PO krd for PO’s from a brand new GNMA pool. If there is 3 years left in a GNMA pool, the remaining interest and principal are nearly insignificant (there literally might only be 6 mortgages remaining in the pool). I know that in GNMA pool #1, there were only 3 mortgages that made it to final maturity. Since they had been paying down principal all along, the total amount of the pool was like a car loan with three years left.

I can see you comparing bond call with PO, but are you saying bond call has negative krd at the short end? and why is that?

That’s what ymmt’s desk says and what Thomas Ho says, so it’s sounding pretty credible to me. Unless someone can come up with a better idea, I think it is just the usual option rho looking like a key rate duration.