MBS CLOSE: Up On The Day, UNTIL THE LAST MINUTE!

By: Matthew Graham

The title of tonight's close couldn't be a better combination of truth and deception.  On the true side, the 4.5 just over 101-20 was indeed at higher levels late in the say compared to the 101-20 where it opened the day.  And it was indeed in the last minute of the chart below that MBS "fell" 101-10, but the truths end there.

Hopefully such a remarkable downward spike in prices occurring with ostensible regularity at end of the first third of any given month has become recognizable as the settlement of the "front month" coupon where charts pick up the first "off the run" coupon as the new impending delivery.  (read more!) In other words, the price difference between the in-the-process-of-settling December coupons and the still-have-a-month-to-trade January coupons was about 10 ticks at the close of the day.  And so the drop you see is not really a drop in price, but just a change in what's being charted. 

So in terms of relative performance, at just over 101-20, MBS ended more or less where they began the day.  This was still about 8 ticks down from the previous session, but hopefully the chart above shows that even with the adjustment from settlement, this is STILL in line with the downtrend we've experienced on 3 progressively weaker auctions.  All the while, the higher-coupon and hence shorter duration MBS are losing much less by comparison.  In addition, their comparable benchmarks in the short end of the yield curve fared better than the long end.

2's and 3's were only about 2bps higher in yield on the day whereas 7's, 10's, and 30's were all at least 6.8bps worse.  That's the "steepening" we referenced earlier, and though it's at its all time highs, there's plenty of pressure keeping it there during this "year-end-window-dressing" phase in which the shorter end, more cash-like positions are favored for their more salubrious effect on balance sheets.  So that's a bit of a double-whammy for the 'ol 2's vs. 10's curve.  On the one hand, we have the forces that have led to a steep yield curve regardless of seasonal balance sheet factors.  But once those factors are considered, the net effect is just about as wide as it gets.

Maybe the spread goes wider, maybe it corrects, but it doesn't much matter, because in the big picture, and by January, it is quite likely to correct.  But in the meantime, as MBS durations are matched against the yield curve, it fosters an artificial demand for the higher coupon MBS versus the "rate-sheet-influencing" coupons like the 4's through 5's.  Does that mean that rates come down when the spread tightens?  Only inasmuch as the 2's do not push higher in yield.  If short end yields back up, the long end can effectively stay put and still experience spread tightening.  But at the very least, in and of itself, a comparative increase in the preference for longer duration securities is  not a negative factor for mortgage rates.

There is enough data tomorrow to at least justify rolling out of bed to take inventory of the market's reaction.

  • retail sales at 830 ....  much anticipated with all the durm and strang surrounding "consumer led recovery," holiday shopping season, and stagnant stocks...
  • Import and Export Prices also at 830.
  • Consumer Sentiment at 955
  • And Business Inventories at 10am

Any major suggestions from these data stand the chance to be more readily greeted by a market eager to find out if 3.50 is indeed the right line to hold in the 10yr tsy yield.  If things look more bond friendly, the range takes over and we're testing various levels in the 3.4's and perhaps even high 3.3's.  But on the negative side, well, the range ALSO takes over and we're up to the previous yield highs around 3.56.

For today at least, the trade was more about the range, and less about the auction.  Trade took yields right up to 3.5 before the auction and despite the brief blip past that support, it was not revisited until the late, illiquid hours which gave way to 3.499 (no joke).

Stocks need nary and introduction these days and I bet dollars to duckets most of us couldn't tell one day from another if not for the benefit of the x-axis.

We'll sign off with a medium-long term view of 4.5's.  Note that the longer trend remains intact, and just like tsy's, we're right at a pivot point in the 101-10 area (vs. 3.5 area in tsy's).  So although it won't make or break the rest of the year, it would make for less drama to get bond friendly data tomorrow, but like we said, whether it is or it isn't, the range--fresh off more edification for the notion today--is likely to take over.