MBS Pricing Snapshot

More of an addendum to the last alert which considered the holding or breaking of short term technical levels... Specifically, 1.83 did not, in fact hold as a ceiling in 10yr yields. Stock markets pushed into new highs for the day, and MBS have generally underperformed Treasuries into the sell-off, not to mention for the past several sessions.

All that resulted in Fannie 3.0s hitting their session lows at 104-10 where they've bounced a bit, along with the stock rally stalling out and Treasuries finding at least a temporary ceiling at 1.836. All things considered, these aren't immense moves for mortgages and it's possible that a fair amount of lenders won't need to reprice for the worse considering the fairly tepid morning rate sheets. But we'll definitely get some reprices, and they've already started rolling in.

After hitting their weakest levels of the day at 12:40pm EST, bond markets looked as if they were turning around and bouncing back into positive territory. Instead, those gains are now looking more like a small-scale version of a "dead cat bounce" at the lows. In fact, MBS are through to new lows now with Fannie 3.0s at 104-13. 10yr TSYs are revisiting their highs around 1.83 and equities are near their highs of the day as well. Negative reprice risk is still out there, but we'd also note that trading surrounding the 3pm Treasury close can be relatively more volatile than other times of the day. If 10's don't meaningfully break the 1.83 ceiling, MBS could hold in the green as well. Those ifs/thens look to be a bit up in the air at the moment.

There's not much to be added to topic of the previous alert other than to say it's continuing. Rather than find a firmly supportive ceiling at 1.83, 10yr yields just poked slightly through, pulled along by improving equities. S&P's are about 7 points higher from their morning lows and on the verge of turning positive on the day.

The leak lower for MBS prices has been slow and rhythmic, with prices back at 104-14, matching the previous bounce just before noon. Whereas we only saw one reprice on the earlier weakness--from one of the only lenders that was remotely likely to be considering one--the ongoing weakness, though moderate, increases the risk that the next few early adopters are considering reprices.

Bottom line, we haven't lost enough ground for widespread reprices, but risks are increased that we could see a few more.

Bond markets look like they might be favoring a bit of consolidation at the moment after just having reached their best levels of the year this morning. From the mid-day recap earlier:

" Stocks, however, have been putting in steady gains since then, helping perhaps, to reinforce a "risk-on" bounce that's currently keeping a floor under Treasury yields (1.8114 in 10's) and a lid on MBS at 104-19 in Fannie 3.0s. We'd probably stay cautious against any technical breaks of the morning's weakest levels--104-16 in MBS and 1.83 in 10yr yields as an early indication that we're entering something of a consolidation pattern after more than a week of gains."

We're now seeing those technical breaks on the MBS side only. Fannie 3.0s broke 104-16 and quickly hitched down to 104-14 before bouncing back to 104-16. This adds to the sense that the weakness is the "slow leak" variety and not yet corroborated by a similar break of support in Treasuries. Nonetheless, weakness is weakness, and while the general rally in equities persists, bond markets look like they'll guard against a "risk-on" correction.

So we're faced with the choice to either lower our expectations for supportive levels in MBS or to fear the negative reprice at current levels. The aforementioned "slow leak" doesn't suggest a mad dash to the lock button, though a select few lenders may be considering it at 104-14. Apart from those few lenders, we'd lean toward "lowered expectations" of technical boundaries--say 104-14 (which happily coincides not only with today's lows, but also yesterday's highs) or until 10yr yields break north of their 1.83 ceiling (assuming that Treasuries/MBS continue to be fairly well correlated.