It was a volatile, but positive day for bond markets with 10yr yields closing well outside the recent range at 2.06.  Fannie 3.5s gained 8 ticks to end at 104-09 and 3.0s were up 10 ticks at 101-09.

Most of the 'back and forth' volatility occurred during the overnight session.  Yields drifted slightly lower by the time European markets opened only to launch themselves quickly higher as stocks/oil/etc bounced back.  In general, there were simply the same sorts of "risk-on/risk-off" tradeflows across multiple markets.  For example, when "risk is on," stocks, commodities, and bond yields are rising.  When "risk is off," bonds rally and stocks/commodities are falling.

Away from all that wide-angle logic, there are no great explanations for each of the day's individual trends as far as data and news are concerned. At the risk of oversimplifying, we could say that bonds initially followed weakness in Asian markets where the Nikkei fell much more quickly than 10yr Treasury yields.  From there, we could account for the reversal by saying European markets found bottom after 30 minutes of quick selling at the open (2am -230am).  The next few hours saw Treasury yields move higher with stocks and oil leading the way.

Up until that point, everything actually made decent enough sense.  Even as domestic markets opened, slumping stocks and oil prices were logical companions for falling bond yields.  But then things got a bit weird--or at least they'll be weird until the last few sentences here.   Stocks and oil recovered, but bonds did NOT do much to follow the move this time around.  There was a small reaction to a component of the Consumer Confidence data at 10am (the "labor differential"--an anecdotal indicator for NFP-- hit it's best level since 2008), but bonds were generally bulletproof for the rest of the day.  Why?

First of all, bond traders know that the risks of a big move lower in yield outweigh the risks of a big move higher.  If EU CPI and domestic data are bond-friendly tomorrow, we'll be starting out from the best levels in more than a month.  That usually results in a major snow-ball-style move.  It's also very likely reinforced by the month-end buying needs (money managers making compulsory trades to bring portfolios in line with their published duration exposure).  Finally, there were NO new corporate deals brought to market today, making for a lopsided demand/supply equation.  Without the month-end trading environment and the absence of corporate issuance, the day might have been noticeably worse.

MBS Pricing Snapshot
Pricing shown below is delayed, please note the timestamp at the bottom. Real time pricing is available via MBS Live.
MBS
FNMA 3.0
101-09 : +0-10
FNMA 3.5
104-09 : +0-08
FNMA 4.0
106-21 : +0-05
Treasuries
2 YR
0.6530 : -0.0190
10 YR
2.0600 : -0.0380
30 YR
2.8580 : -0.0180
Pricing as of 9/29/15 5:56PMEST
Today's Reprice Alerts and Updates
A recap of Alerts and Updates provided to MBS Live subscribers.
10:06AM  :  ALERT ISSUED: Once Again, NYSE Open Juices Bond Rally
MBS Live Chat Highlights
A recap of featured comments from the Live Discussion on the MBS Live Dashboard.
Ira Selwin  :  "FNMA going to 95% LTV come December on High Bal"
Compliance is Watching Me  :  "I have already practiced my script: Mr. Customer, we are using a new disclosure that was last changed when Reagan was in his first term. Please allow for some possible technical issues as we work through these mandatory regulatory changes. Any complaints may be directed to your congressman."
Matt Hodges  :  "the LE specifically excludes aggregate adjustment"
Matt Hodges  :  "i'm still hung up on the fact that cash to close due to AA won't be same on LE as fee sheet or 1003"
Sung Kim  :  "its going to be like a fireworks show on friday"
Joel Marks  :  "I just know this is the most important NFP since the last one."
Matthew Graham  :  "to approach the example from a different angle, let's say I buy 3.0s the day before the roll at 101-08. You buy the next month's 3.0s at 101-00, the day after the roll. Despite the fact that our purchases were only 2 days apart, I get my first payment a month before you! The closer you get to your first payment, the more implied appreciation in your MBS purchase--effectively getting you back to that "all things being equal" price of 101-08."
Matthew Graham  :  "Spencer, the impact of the roll isn't cumulative. In other words, it doesn't keep adding up. It washed out every month over the course of the month. All things being equal, there is a gradual appreciation through the course of the month that would get prices back to where they were before the previous roll. E.G. Fannie 3.0's are at 101-08. Let's say that's where they are before the roll and the roll takes them to 101-00. They'd then slowly appreciate back to 101-08 before the next roll, assuming absolutely no external impact on bond markets or the relative value of MBS."
Spencer Packer  :  "Question: As I look at the 6 month history for the 3.5, I'm seeing that we're real close to best levels since the beginning of May. I'm wondering if we might actually be better because of all of the rolls we've had since then. Is there any truth to this? I mean, most rolls are what, 8 - 12 ticks? if we've had 4 of them since May, that would make our current adjusted levels better than then? "