MBS Day Ahead: New Month, New Trend? Reasons to be Cautious

The end of November was a supportive environment for longer-dated Treasuries like 10yr notes and 30yr bonds.  The MBS that most directly affect rate sheets tend to track more with this so-called "long end" of the yield curve.  As such, rates have been stable to slightly lower for the past few weeks.

Now we must ask ourselves if that was purely a byproduct of 2 temporary factors.

The first factor would be the "month-end" trading environment.  Most fund managers use industry benchmark indices (like Barclay's Aggregate Bond Index, which also includes specific indices for MBS and Treasuries) to determine the right mix of bonds in their portfolio. In other words, if a fund purports to hold 40% Treasuries, they're not just holding any random Treasuries, but a specific mix that averages out to the index value.  If they never bought more Treasuries, the average duration (i.e. the number of years left that the bonds will be paying a return on investment) would slowly decrease.  Month-end buying keeps the average duration more stable. 

The index doesn't stay perfectly flat.  Although it does generally need to increase (or in industry jargon, "extend"), some extensions are bigger than others.  This month's was particularly big, suggesting a higher-than-normal amount of month-end bond buying.  The natural fear is that positive or stable trend at the end of November subsequently reverses course as the new month begins.

The second factor is the big movement that's been seen in the yield curve.  Simply put, the "long end" (again, that's stuff like 10-30yr Treasuries, and to some extent, MBS) has been outperforming because the short end (stuff like 2-3yr Treasuries) suffers more as Fed rate hike expectations increase.  This has made for an environment where 10yr yields were seemingly stagnating in a non-threatening, sideways, slightly stronger pattern.  At the same time, 2yr yields were very much on the move.  The following chart shows 10yr yields overlaid with the spread between 2 and 10yr yields.  The lower the purple line, the closer the two have become.

That purple line is also known as the 2s/10s curve, and it's getting close to its lowest (or more appropriately, "narrowest") levels since before the financial crisis.  This was one of the quickest adjustments of the year for the yield curve as markets came to terms with the FOMC meeting minutes in mid-November.  Paradoxically, rapid selling in the short end of the yield curve can actually help or stabilize the longer end of the yield curve as bond market investors keep their holdings in the bond market. 

In other words, they're selling the short end and buying the long end.  Once the quicker moves are over though, any further contraction in the 2s10s curve historically tends to correlate with rising longer term yields, especially when it contracts from current levels.  This is easy to see in the chart below, where 1994 is a more extreme example that resulted in significant pain for longer term rates and 2004 resulted in a more gradual rise.

Of course there's no rule that says history must repeat itself in the same way every time, but with so few past examples of a move below current yield curve levels, it's a concern.  Combined with the potentially changing technical outlook in 10yr yields themselves, it could be enough to change our recent trend as the new month begins.  The recent technical environment looks a bit like May 2015 where yields leveled-off after a quick run higher, but paused after hitting the 21-day moving average (middle bollinger band in the chart) and subsequently moved higher after MACD (a momentum indicator) ran out of steam (i.e. "stopped moving lower).

It may very well be the case that the data coming out over the next 4 days will decide how quickly we see such a deterioration (or lack thereof).  I'm thinking more about the ECB on Thursday and NFP on Friday, but today's ISM report at 10am is always a potential market mover if it deviates enough from the consensus.  That said, manufacturing has been stagnating this year, so market participants may not be as surprised as we'd like by an unexpectedly weak report.  I'm not sure the opposite is true for a strong report.