When valuing illiquid alt-a and subprime MBS, what are some common ways that investors come up with discount rates? I assume it wouldn’t be proper to use a static rate, that instead you should apply some sort of yield curve to the cash flows? From what I have seen, the discount rate affects the valuation more than the negative home price appreciation scenarios.

anyone?

If you are valuing a sub-prime or Alt-A MBS and all you have is a bunch of discount rates, you have nothing. The key is to model default rates and default correlations. There is monster literature on this. Home price appreciation is about recovery and needs to be included too. Modelling this stuff is a morass, which is why we are having such a problem.

i have the cash flows based on cpr, cdr and severity… i am just wondering what discount rates i use for the cash flows

correction: Forass, not Morass.

not that I am an expert, but it would be an assumption on default probability distributions, and running monte carlo simulations on default assumptions, interest rate paths, etc… In the perfect world, if I lent someone money and they posted an asset as collateral, i don’t care the value of the collateral as long as they continue to pay me back… however, where I think many investors may have missed in their modeling, is that there is a correlation between the asset value and defaults. (ie. If a subprime borrower’s house depreciates, they are more likely to quit paying their loans…go figure.) This is a double whammy…

drs Wrote: ------------------------------------------------------- > i have the cash flows based on cpr, cdr and > severity… i am just wondering what discount > rates i use for the cash flows ohh… can’t you just use a spread close to implied liquid MBS? discounted for illiquidity?

implied liquid mbs?

drs Wrote: ------------------------------------------------------- > implied liquid mbs? maybe they don’t exist anymore? i dunno… you said illiquid, so i assume their are some liquid ones? or not…

I’d say you want to separate out the sources of uncertainty. 1. Cash flows, adjusted by probability. You’ve calculated those from cpr/cdr/severity. 2. Discount rate. 3. Illiquidity discount. I’d say that, relative to a few years ago, #1 has declined (due to deteriorating underlying fundamentals), #2 should be about the same, and #3 has increased (perhaps best expressed as bid/ask spread). When projecting uncertain cash flows, many ppl just merge #1 and #2 – but as any corp fin textbook will tell you, it’s misleading to do this – it leads to inappropriate aggregate discount factors. I think your consideration of a term structure for discount rates indicates you’re thinking about them in the wrong way. I’ve never seen bid/ask spreads (illiquidity) folded into discount factors. (Illiquidity can be addressed with patience; discount rates, not.)

the margin on these mbs are based on libor or cmt plus a certain amount. So why wouldnt i discount them based on libor plus a spread or something to that effect?

drs Wrote: ------------------------------------------------------- > the margin on these mbs are based on libor or cmt > plus a certain amount. So why wouldnt i discount > them based on libor plus a spread or something to > that effect? are your CF’s based on hard probabilities or do you have a range? If you have a schedule of cash flows based on defaults, (the spread is already built into the Cash flows,) then just discount them back using Swap curve.