Tuesday, June 12, 2007

Sir, it's quite possible this market is not entirely stable

Remember in 2002-2003 when the market would rally a little, then accelerate, and every one blamed mortgage servicer hedging their portfolio? I'm becoming strongly suspicious something similar is happening right now. Maybe not just servicers either, but various leveraged MBS buyers. Take a look at recent volatility in 10-year swap spreads...

This spread represents the difference between the 10yr Treasury and the fixed leg of a 10yr plain vanilla interest rate swap. For some background, an interest rate swaps are the most common means of eliminating interest rate risk for leveraged MBS investors.

As interest rates rise, the duration of a mortgage loan increases. Higher rates make refinancing less attractive, as well decreasing the borrower's overall mobility. So when interest rates rise, hedged MBS investors need to own more hedges in order to eliminate duration risk. To see why, imagine that duration of your MBS portfolio starts at 3. You have pay fixed swaps outstanding that have a -3 duration. So the net is zero.

Now interest rates rise and the duration of your MBS portfolio is now 4. The swaps won't change appreciably, so now you need to find another -1 in swaps. Increasing your pay fixed swaps has the same effect as selling bonds. If there is overwhelming demand for the pay fixed side of swaps, the rate has to rise to entice investors to receive fixed. Just like when there is overwhelming supply of bonds, bond yields have to rise.

Back to the graph. Visually, you can see that the swap spread has widened substantially, and the intra-day volatility is noticeably higher over the last 5 sessions or so. That would sure suggest that there is increasing demand for pay fixed, and it looks to me like these players are driving the market. For example, on Friday, when the 10-year oscillated from down 3/4 to up 1/4, swap spreads ranged from +65 to +60. On that day, when the Treasury market was weakest, swap spreads were at their widest and vice versa.

Consider that if the Treasury market was leading the swap market, the opposite would be true. The Treasury rate would rise faster than the swaps market and spreads would seem to compress. When the Treasury was rallying, swaps would lag and spreads would seem to widen.
So what if I'm right? Its just like any other technical influence. Once its out of the market, the market must then search around for the right fundamental price. In this case, we'd likely see a very quick, but probably small, rally. Then it really will depend on where the Fed is going. Which I think we'd all agree, is becoming a tougher and tougher call.

So what else might you do? You could make a bet on swaps tightening, which I think is a no-brainer. Its an easy bet to put on, because just about all high-quality spread sectors are swap-correlated.

The other way to play it is to go long MBS. MBS spreads have widened even faster than swaps, I'd say due to selling by leveraged buyers. Stands to reason: if you need to get your duration down, you can either increase your hedge or decrease your long position. At the very least, no one in that group is buying MBS, and its been widely reported that dealers came into May with very large positions.

So we are in for a volatile market, but hell, that's how traders make money.

Not entirely stable? I'm glad you're here to tell us these things.


Anonymous said...

Some air, but a very small amount, has been taken out of the euro's sails. It's still above $1.30.

I think the $64,000 question is,

Does Bernanke believe that a weak dollar contributes to inflation?

If he does, the Fed is not going to lower rates. Not until the euro is at about $1.10.

Anonymous said...

Great blog, keep it up! I'm learning a lot (= plankton retail bond buyer)

Accrued Interest said...

Welcome to the blog. Hopefully we can help you avoid being fleeced by your broker.