Why the 10 Year Yield is Dropping, and Why the 30 Isn't Catching Up

In response to RW's post yesterday on the drop in the 30 Year fixed rate mortgage to 4.375%, I commented:

Ultra-low mtg rates are why the Fed announced a coupon swap today on to-be-settled Agency MBS. On-the-run 5.5's are expensive because of low supply. Doubt these were purchased prior to the Mar 31 termination of the purchase program, but as a result of dollar rolls. Each transaction is a hidden subsidy to favored PD's.

Another unintended consequence: MBS duration models are triggering long term Treasury purchases to compensate for increased prepayment risk. These models are slow to adapt, however, and many data sources suggest refis are slowing, not growing (incl. Consumer Metrics Institute). 10's could go to 2%, but the ultimate unwind will be enhanced when the artificial demand evaporates.

Today, ZeroHedge runs with a Morgan Stanley report, part mea culpa, part defensive posturing with regard to Jim Caron's prior (and continuing) non-deflationary stance. After some hilarious predictions about 3.5% annual growth for the US in 2010 (globally 4.8%!), Caron reasons as follows:

But for now, here are some of the reasons the UST 10y has broken down below 3.00%

1. Fundamental reasons: The most common reason is that people feel the mkt is headed into a disnflationary double dip and buying UST 10s at this level is a good play for that scenario.

2. Hedging reasons: as one long/short equity manager put it to me, "USTs are big and liquid, you've gotta own them". This was in the context of owning USTs as a preferred hedge against their equity longs. They felt owning USTs was akin to owning tail risk. And at least you got paid some carry to own this hedge. If mkt conditions improved, you could exit easily.

3. Asset Allocation and Indexers - these reasons are more technical:

• people were short/underweight USTs through May. Recently, fund managers have been getting back to a neutral weighting, which implies they will need to buy USTs in the process.
• as the universe of USTs increase, indexers need to buy more and more USTs to remain at the prescribed weighting of their index.
• the Fed announced yesterday it was going to swap the 5.5% coupon mortgages it bought during QE for 4.5% coupon mortgages. This was done for liquidity to alleviate delivery fails in the 5.5%s. The impact of moving to a lower coupon is that it's duration is longer. So, if you sell the 4.5% mortgage to the Fed in exchange for a 5.5%, then you are effectively short ~$2Bn UST 10y equivalents. This means you need to BUY $2Bn 10s. This caused the mkt to drop 6bps in UST10y yields in a flash. Also, people speculated this may be a signal QE may re-start, which we think is nonsense.
• Month-end/Quarter end window dressing. Although there is not much of an extension, managers want to show investors they own nice safe USTs.
• people view the roll-off of the €442Bn 1yr LTRO in Europe on July 1 as a potential event risk. We disagree with that notion (as per Mutkin's last weekly).

4. Rumors: there have been some articles in the UK Telegraph suggesting that the Fed will enter QE again and buy over $2Tr more in assets. This has not been verified as a reliable source.

A Counter Point to the Deflationistas

If true deflation fears were at work in the market then one should expect the longest duration points on the curve to rally most and we would see a significant flattening of the yield curve, especially the UST 10s30s segment of the curve. We use this as a check against deflation fears. Instead, we see the opposite with the UST 10s30s curve right up against its 20-year highs. If deflation is truly at work in the market, then someone ought to tell the UST 10s30s curve. And for that matter, all curve segments should be much flatter. Yet the curve flattening has been slow and grudging. Thus the yield curve is more representing an overall shift to lower yields as the Fed is expected to be on hold for a long time, not that deflation is truly at work. Don't confuse the two. It's too early to conclude that deflation is truly at work in the marketplace.

His justifications for the yield drop are pretty much spot on (though the conditions will likely persist longer than he thinks); however, his observations on the 10s30s spread are leading to an erroneous conclusion. That segment will be the last to contract because (a) there is a real and rising risk of US default as evidenced by credit default swaps (30 years is 3 times as long as 10 years, after all), (b) the 30 year is not used as a primary hedge instrument against duration, nor is it materially a necessary embedded component in the half quadrillion in interest rate swaps (IRS) (whereas the 2, 5 & 7 tenors--and to a lesser degree the 10--are), and (c) the greatest demand for the 30 year will be the Fed itself when it restarts Treasury QE.

Until Gentle Ben's QE 2.0 capitulation, the 10s30s spread should not be read as a non-confirmation of deflation as there is simply not the same artificial structural demand for the 30 year as there is with the 10. This is also in line with my previous explanation of how artificial distortions in the yield curve will hinder its predictive power.

Long term, it's going to be inflationary turtles all the way down, but for the time being, Treasurys are more than likely to rally (that is, unless you think the S&P 500 puts in a quadruple bottom here).