Thursday, September 22, 2011
10-yr Treasury yields have never been as low as they are today (1.73% as of this writing). Meanwhile, the effective duration of the MBS market has rarely been lower. The two go hand in hand, since the huge collapse in the duration of MBS (from a high of 4.9 years last April to just 1.5 today, by my estimates) has forced large institutional money managers to buy huge amounts of Treasury notes and bonds in order to keep the duration of their portfolios from collapsing, otherwise they would suffer significant underperformance. Declining MBS duration, in other words, has contributed significantly to the decline in Treasury yields.
The combination today of record-low Treasury yields and almost-record-low MBS duration is like compressing a massive spring: if and when the downward pressure on yields lets up, we will see a massive rebound in yields as the recent process reverses. Managers would be forced to sell massive amounts of Treasury notes and bonds in order to offset the huge increases in MBS duration that will follow any rise in Treasury yields. Thus it is that the mortgage market, which comprises about 28% of the investment grade U.S. bond market, can sometimes be the tail that wags the much larger dog—magnifying greatly the market's underlying volatility during periods of stress. Just a modest nudge from the Fed, in the form of the upcoming Operation Twist, can result—at least temporarily—in a huge decline in yields. And just a whiff of good news could send yields sharply higher. We have seen big reversals before when similar conditions existed (i.e., a big decline in 10-yr yields and a big drop in MBS duration): last Fall, early 2009, the Summer of 2003, and early 1999. The next one could be the mother of all Treasury meltdowns.
The only way that a big increase in yields can be avoided is if the downward pressure on Treasury yields continues. And that can only happen if the economy enters another profound recession, and/or inflation turns negative. Once again we find ourselves on the edge of the same abyss we were staring into at the end of 2008: this market is priced to something akin to the-end-of-the-world-as-we-know-it; nothing less than a catastrophic deterioration from current conditions. No one can rule out a truly worst-case scenario, but to get there requires a whole host of things to go from today's unpleasant to absolutely abysmal. I'm not prepared to bet that the bottom will fall out of everything, so that leaves me bullish compared to where the market is.
Posted by Scott Grannis at 11:53 AM