We'll start off with more of an industry-specific curiosity (as opposed to diving right into markets/rates).  Then again, the conforming loan limit is a key consideration for rates, and it will be updated in less than a month!

Today's FHFA home price data was the 2nd to last month needed to calculate the annual change in home prices (which FHFA relies on to determine new loan limits).  But it's not quite as easy as just looking at the headline 8.0% annual growth.  Loan limits are based on FHFA's expanded data set that is only released quarterly.  Fortunately, that expanded data set tends to fall very close to the Home Price Index headlines.  We'll get September's data (and thus the quarterly number) next month.  We got August today, and it was very strong.

But do you really care about the nitty gritty, or do you just want to know what the range of possibilities looks like?!  So without further ado, from worst case to best:

• $537,450 - If prices mysteriously tanked in September and there was no home price growth for the entire quarter.
• $545,975 - If prices mysteriously stopped improving in September, and Q3's gains were solely in Jul/Aug
• $546,435 - If price growth cooled off massively in Sept, and Q3 was in line with average growth over the past year
• $549,190 - If price gains cool a bit in September
• $551,232 - If annual price gains remain at 8.0% (as was reported today)
• $552k+   - To whatever extent September price gains cause annual appreciation to increase beyond the 8.0% seen through the end of August, new conforming loan limits will be $552k or higher.

Moving on to markets

So far this week, the bond market is doing its best to convince us that October's weaker trend has run its course.  After hitting yields of .87% on back to back days at the end of last week, 10yr Treasuries have moved at a decent pace back in the other direction.  They're almost 10bps lower at the moment. 

It's not a huge surprise to see bond bulls push back against the bears here. This isn't a comment on the bigger-picture trend as much as it's a simple little technical trick involving the number of successive days spent moving in one direction or another.  As of last Thursday, yields made new highs on 6 out of the previous 6 days.   Any time we see that happen for at least 5 straight days, it makes more and more sense to keep an eye out for a rebound.

In other words, the more consecutive days spent moving toward higher yields, the more likely we are to see a rally day.  The odds really start tipping in favor of bond bulls after 5 days, as I mentioned in the huddle video last week.  But if today's gains hold, we'll have more than just a token, technical push back against a negative trend.  We may have to look elsewhere for justification.  Here's the first place I'd look:

The *'s in the chart represent fairly subtle nuances in the data.  We've only had 1 real chance to see how markets react to a resurgence in case counts.  March/April doesn't count because it was the initial surge, and October is still too new to draw a conclusion.  In June, we see a prominent increase in cases without a big, immediate reaction in the bond market.  As cases continued higher, bond yields continued lower--ultimately reversing course after cases reversed course (again with a delay--unmarked in that case).  If there's a similar delay this time around, this week would be the time where we'd expect bonds to start pricing in covid concerns.

Of course this isn't the only motivation in play for bonds.  Stimulus continues to loom.  The presidential election keeps the spotlight on uncertainty.  And earnings season for stocks always has a chance to add volatility to both sides of the market.  Best bet is to keep the broader range in mind (0.5-1.0% in terms of 10yr yields), factor in mortgage-specific insulation (mortgage rates can weather a storm of rising rates much better than Treasuries, unless the pace of selling gets extreme), make hay in the sun, and adjust accordingly when the weather pattern shifts.  Rest assured, we'll be talking all about it when it happens.