By 
Vital Statistics:
Stocks are lower this morning after Chinese developer Evergrande called off talks with creditors and looks set for bankruptcy. Bonds and MBS are down.
The upcoming week has quite a bit of data, with house prices / new home sales on Tuesday, GDP on Thursday, and PCE inflation data on Friday. We will also get some Fed-Speak.
There was an interesting interview in Housing Wire with Doug Duncan, chief economist at Fannie Mae. MBS spreads are a huge topic these days, and he was discussing who will be the marginal buyer to step up and replace the Fed’s buying.
Kim: Spreads in the mortgage space are wide. What are the reasons for that?
Duncan: There are several reasons for that. If that business flow for a time period helps them cover the variable costs, then it can be effective.
For one thing, no fixed-income investor thinks that mortgage-backed securities with 7% mortgage rates will be there when the Fed finishes the inflation fight. They’re going to cut rates and that will prepay. So you’re having to encourage investors with wider spreads to accept that.
It’s also the case that the Fed is running its portfolio off because they don’t talk about it much. But somebody has to replace the Fed, and the Fed is not an economic buyer. That is they weren’t buying for risk-return metrics; they were buying to affect the structure of markets. So they are a policy buyer.
They were withdrawing volatility from the market, and they were lowering rates to benefit consumers. When [the Fed] is replaced, it’s likely to be by a private investor who’s going to have yield expectations. They may require wider spreads than the Fed because the Fed is not an economic buyer.
While I believe he is correct in that the new buyer of MBS will require a higher spread than the Fed, which had no such requirements, I think he overstates the effect the Fed’s buying had on MBS spreads in the first place. Take a look at the chart below, which is the 30 year fixed rate mortgage rate minus the 10 year.
This is not exactly MBS spreads, but it is a close enough approximation. The thing that sticks out to me is that MBS spreads in the era of QE are not that much different than they were before the real estate bubble. If the Fed’s massive buying of MBS didn’t make that dramatic of a difference, how is slowly letting the portfolio run off going to do it?
Once the Fed is out of the way with rate hikes, we should see a dramatic drop in bond market volatility as the uncertainty over monetary policy disappears. Since fixed income investors are looking at option-adjusted spreads (OAS), as volatility dries up in the bond market, we should see MBS become more attractive to other credit-risk free assets. Yes, prepays might increase, but rates have to fall a lot to trigger any sort of refi boom.
From 12/31/99 – 12/31/06, the difference between the 10 year and the average 30 year fixed rate mortgage was 1.79%. This was pre-Fed intervention. If spreads return to that level, we would be looking at a 30 year fixed rate mortgage around 6.3%, or 100 basis points lower than here.
Comments