Negative OAS - what does it mean?

Singlestrand · October 31st, 2005, 8:56 pm

ok, dumb question - but please clarify.what does a Negative OAS mean?I have a CMO with a negative OAS calculated over the treasury curve.Now,Z-spread = OAS + option cost.Z-spread risks = credit + liquidity + callOAS risks = credit + liquidityDoes it mean that this bond has less credit + liquidity risk than treasuries? Or does it mean that the interest rate model is garbage?Or does it mean that the whole OAS calc procedure is garbage?

JWD · November 1st, 2005, 1:15 am

Hi Singlestrand,Not a dumb question. Assuming you’ve checked your OAS MC calculator for bugs, have normalized it with other CMOs to make sure it’s reasonable, that the MC convergence is OK etc., a negative OAS sure seems weird all right. Are you getting this result for one particular tranche; if so, what is the nature of the tranche? Also, is this negative OAS a one-time problem today or have you been getting it systematically? ----------

Perpetual · November 6th, 2015, 2:32 am

I actually ran into same question for Structured IO. What does negative OAS really mean for Strucutred IO?

bearish · November 6th, 2015, 6:03 pm

QuoteOriginally posted by: PerpetualI actually ran into same question for Structured IO. What does negative OAS really mean for Strucutred IO?It means that your model suggests that the IO is very rich.

Perpetual · November 6th, 2015, 9:28 pm

Sorry. what does that mean IO is very rich?

bearish · November 7th, 2015, 2:41 am

QuoteOriginally posted by: PerpetualSorry. what does that mean IO is very rich?In the world of securities, rich is the opposite of cheap. The market price is higher than what your model suggests that it should be. This is one of the (if not the) primary reasons to create a model in the first place.

Perpetual · November 10th, 2015, 12:34 am

Thanks!I am struggling to understand the OAS and OAD for IO. For example, given two models (1 and 2) and assume model 1 over predict prepayment than model 2. Here is my understanding. The cash flows and number of cash flows from model one will be less than model 2 when pricing. So the OAS1<OAS2 no matter if OAS is positive or negative, correct? What about OAD? How to undersand it?