"2008-10 is when they were buying but it does not inform us what they were paying."
I went looking for a simple MBS pricing model to help me explain how the Fed would make money holding MBSs in an environment where default rates were falling and interest rates on fixed dated cash flow slide down the yield curve. The best I could find was this:
If I ever attempt to explain this, you should put me on ignore and get on with your otherwise happy life. LOL
(There is some humour in there for those who like poking fun at math geeks. Those big funny shaped "E" thingies in the expression (Epsilon), are all about using complex stochastic processes to model the impact of early repayments of loans. The irony about this is that hardly anyone in the US qualifies for refinancing, so modelling early repayments is pretty much a waste of time.)
The discussion about whether the Fed can recover the purchase price on its MBSs assumes that the Fed intends to sell them. I doubt that they will. The only reason for the Fed to sell bonds is to tighten credit. With the current state of the US economy, I wouldn't expect that to happen any time soon.
"Everything depends on whether the QE operations finally result in a return to health for the US economy."
I think that a few people are waking up to the fact that QE hasn't been everything everyone thought it was. It didn't cause any inflation (which is a pity). Apart from causing term rates to drop, it did nothing to stimulate the US economy. Another round of QE might be good for market morale, but not much else.
It will be interesting to see how both gold and equity markets cope with a macro economic outlook that doesn't consider QE. It has been such a security blanket for the thumb suckers.
