MBS Pricing Snapshot

We've seen this scenario play out many times before and while it frequently amounts to nothing more than a head-fake, it's worth noting that MBS are back to their lows of the day. For Fannie 3.0s, this is still 21 ticks higher than yesterday's close, but is also roughly a quarter of a point off the previous highs of the day.

On the 2 day chart, given the big swing from yesterday to today, this won't look like much and indeed, it's not the basis for any sort of big-picture panic. But it IS the kind of nagging end-of-day weakness that CAN cause a lender or two to reprice for the worse. Bigger risks develop if we slide a few more ticks and if markets beyond MBS seem to be participating in the move. On the edge for now, but stay frosty.

In and of itself, today's 10yr Treasury Auction was quite a bit weaker-than-expected, with yields coming in almost 2 bps higher than expectations and the bid-to-cover at historically low levels. But when you consider that 10yr yields are over 10bps improved from yesterday's close and that they rallied to their best levels of the day BEFORE the auction, a bit of weakness in the auction results is completely palatable, and markets are bearing that out presently.

In fact, had the auction happened half an hour earlier, the metrics used to measure its strength could paint a completely different picture. Reason being, yields are currently UNDER the levels seen from 10:30am - 12:30pm., and with a brief exception during the lead up to the auction, essentially grinding around the lows of the day (also the lows of the past 3 weeks).

For their part, MBS are nonplussed, with MBS briefly revisiting pre-auction lows and then continuing to trade sideways in the day's existing range. Volatility has been generally decreasing since the auction and Fannie 3.0s are hovering around 105-10, which is 25 ticks up on the day.

We've seen one positive reprice on the stability and wouldn't rule out a few more. That said, widespread positive reprices aren't exceedingly likely unless we see more concerted gains.

Last time we had a 10-yr refunding auction on 8/8, it was fairly ugly but probably had more to do with the timing of the auction than the auction itself. Other than that, the historical tendency is for 10yr refunding auctions to stop through at lower than expected yields.

Keep in mind that bid-to-cover is much lower on average for refunding auctions vs reopenings with the last 4 refundings averaging 2.77 vs an average of 3.24 for reopenings over the same time frame.