Latest News From Our Blog

Mortgage rates are on a tear.  As of this week, 2019 is now the best year in more than a decade in terms of the peak-to-trough move.  But there’s a catch: 2019 isn’t over yet, and there’s always a risk that rates move higher between now and the end of the year.

So, will they (move higher)?

There are plenty of opinions on the future direction of rates, but there are NEVER any guarantees.  All we can do is examine the available data and precedent in an attempt to understand what’s possible.

Things that build a case for lower rates

  • Past precedent and market technicals
    • Rates have been a bit lower in the past–both in 2016 and 2012.
    • The yield curve (gap between 2yr and 10yr Treasury yields) inverted recently and some say that’s a harbinger of recession (and recession is good for rates)
    • Rates have undergone an even bigger overall move from peak to trough in 2011/2012 (1st chart below)
    • The super long-term trend in rates is still pointing lower (2nd chart below)

20190906 nl1

20190906 nl2

  • US and Global market fundamentals
    • The economic expansion in the US is now in record territory in terms of longevity.  One school of thought suggests this makes it increasingly susceptible to a shift.  Such a shift would generally coincide with downward pressure on rates (typical of most any economic downturn)
    • Trade war uncertainty continues to promote safe-haven investments like bonds.  Bond buying = lower rates
    • European and Chinese economic data isn’t nearly as strong as US economic data.  Global weakness has spillover effects that can benefit rates in the US.
    • Global central banks haven’t been shy about pushing much of the world’s supply of bonds into negative yield territory.  Although many market participants are forced to buy the bonds of a specific country for certain reasons, investors who have a choice are obviously going to be more interested in keeping their safe-haven assets in the highly liquid Treasury market where yields are still in positive territory across the board.  US 10yr yields are much higher than all G7 countries.

20190906 nl3

Things that build a case for HIGHER rates

  • Past precedent and market technicals
    • Sure, it’s easy enough to highlight the ONLY example in the past 16 years of a bigger move lower in rates, but compared to every other example, 2019 is the biggest and thus the most susceptible to a shift in momentum.  In other words, big, sustained drops in rates only tend to go so far and last so long.
    • Yes, the yield curve inverted, but past precedent suggests it could take 2 years for a recession to materialize in response.  There also aren’t really enough past examples that have occurred in a low, stable rate environment to draw a convincing conclusion.
    • Looking back at the 2nd chart above, notice the red dotted line.  Yes, we’re technically still in the long term downtrend in rates, but we could also be seeing rates level-off and favor a more sideways trend from here.
  • US and Global market fundamentals

    • While it’s true the US economic expansion is long in the tooth, it’s also true that 2015 is seen by many as a “stealth recession”–one that effectively could have reset the expansion’s lease on life.
    • While it’s true that foreign economies decelerated sharply in 2018 and 2019, the losses have been moderating and they may be bottoming out in terms of growth levels.
    • While trade war uncertainty has been a huge motivation for rates to move lower, to whatever extent trade relations can be improved, rates could face some upward pressure as investors anticipate at least a token global growth rebound.

In terms of this week’s underlying market movers, we saw a fairly even split between trade-related updates and US economic data.  The following chart breaks down the 3 most noticeable reactions in Treasuries.  The first was a report citing China’s commerce minister saying that trade negotiations were set to resume in October.  The biggest single uptick in bond yields followed Thursday’s exceptionally strong ISM Non-Manufacturing report.  But bonds were able to recover a bit on Friday after the Jobs report missed expectations.

20190906 nl4

To be fair to the bond market and the jobs report, it was really only the outright job tally that was weaker than expected.  The unemployment rate held at 3.7% despite more workers entering the labor force, and average hourly earnings beat forecasts on a monthly and annual basis.  All told, if bond traders were determined to trade rates higher, this report wouldn’t have prohibited that.  So the fact that rates recovered a bit is somewhat reassuring.

The other way to look at the point above would be to say that bonds have now pivoted above the high yields seen at the end of August and refused to break back below those levels after Friday’s jobs report.  In other words, perhaps the ceiling is once again becoming a floor…

20190906 NL7

Ultimately though, unless next week brings a high-conviction move, we’re likely waiting for the mid-September Fed announcement before we get a clearer idea about the next phase of momentum in rates.  For now, anyway, mortgage rates have greatly enjoyed a bit of sideways weakness in Treasuries.  Up until the past week and a half, they couldn’t keep up.  Now they’re closing the gap!

20190906 nl5

It’s easy enough for media outlets to accurately report the underlying facts when it comes to mortgage rates and housing. It seems far harder to find consensus when it comes to interpreting those facts.  Let’s clear up some of the misconceptions driving this phenomenon as we recap the week’s market movement and housing-related developments.

For many consumers and reporters (and unfortunately, more than a few loan originators), the most glaring source of confusion at the moment is the discrepancy between 10yr Treasury yields and mortgage rates.  The 10yr yield, above all other interest rates, has long been thought to dictate mortgage rate movement.  In a way, it actually does that, but not to such an extent that mortgage rates always follow in lock-step.

The 10yr is the most widely traded longer-term bond in the US Treasury complex.  It’s a benchmark by which all other long-term rates are judged and measured.  Investors are essentially faced with a choice to buy a risk free bond from the US government at a lower rate of return or a substantially similar bond at a higher rate of return, but that also carries additional risks.

MBS (the mortgage-backed securities that underlie mortgage rates) are just such an investment!  The biggest risk for the MBS investor is a rapid change in refi demand.  In fact, if borrowers were forced to hold their mortgage for the entire term, mortgage rates would be extremely close to Treasury yields and the gap between the two would be almost perfectly constant.

But borrowers can and will continue to refinance when rates drop enough or keep their loans longer than expected if rates are on the rise.  Changes in refi behavior directly affect investors’ returns.  That’s why mortgage rates are higher than Treasury yields.  The spread compensates investors for the extra uncertainty.

When rate volatility is within historic norms, the spread between MBS and Treasuries doesn’t move too quickly.  But when rates are changing massively–as they have in 2019–it can do a lot of damage, and investors will quickly demand much higher spreads before buying MBS.  This translates to mortgage rates being higher than they otherwise would be if they were keeping consistent pace with Treasuries.

Fortunately, MBS are similar enough to Treasuries to have been benefiting from the broader drop in rates.  As such, mortgage rates are indeed at long-term lows.  They’re just not as low as some would expect.  The following chart shows how MBS prices were moving in relative lock-step with Treasuries followed by the big divergence in August.

2019-8-30 NL4

That was August’s big story when it comes to rates.  This week’s big story is much cooler.  It involves MBS beginning to push their spread back in the other direction (something we expected to see as soon as it looked like the drop in Treasury yields was leveling-off).  The following chart has the same two lines, but the y-axis scaling has been adjusted so we can zoom in and look at this week’s relative performance.

2019-8-30 NL5

To bring all this home, here’s how actual mortgage rates have moved relative to Treasuries:

2019-8-30 NL6

Even though mortgage rates may not have fallen as fast as Treasuries, they’ve fallen more than enough to spark the first real refi boom we’ve seen since 2016.  Some news stories from this week called the refi boom into question because the number of refinance applications declined week-over-week, but this is absolutely to be expected.  The following chart shows how rising refi demand will periodically level-off even as rates continue lower.

2019-8-30 NL7

Incidentally, the purchase application data was another source of confusion this week.  Several news outlets cited the decline as evidence that buying demand was shrinking.  Someone should let them know that mortgage application data is not seasonally adjusted and that the August decline looks almost exactly the same almost every year.  Even then, this is WEEKLY data (whereas most reports are monthly), and there will always be some degree of “noise” even if we were viewing it in seasonally adjusted terms.  The noise wouldn’t be nearly as noticeable if this were a monthly report.

But even in monthly housing-related reports, we were faced with some mixed messages.  The highly-regarded Pending Home Sales data was released for the month of July this week, and it showed what some referred to as a troubling or mysterious decline.  Even some industry experts concluded that lower mortgage rates weren’t having the expected effect on home purchase demand.

This one is more debatable than the other examples of ‘misconceptions.’  It is true that Pending Sales declined in July.  It is also true that the decline COULD be indicative of some issue that’s holding back healthier housing demand.  On the other hand, it’s a good idea to remember that even the monthly housing data is noisy, with multiple examples of declines even in broader trends of improvement.

I’m not saying this is definitely going to be a broader trend of improvement and that you should ignore the decline–simply that the single month’s decline doesn’t really tell us anythingconclusive.  Add to that the fact that Pending Sales were still very close to unchanged in year-over-year terms and a better conclusion might be that sales have leveled-off after recovering from late 2018’s weakness.

2019-8-30 NL8

Away from housing and mortgage markets, we may be able to find another misconception or two.  In broader financial markets this week, trade-related headlines continued to drive volatility.  Notably, stocks managed to move slightly higher from mid-week while bond yields held mostly sideways.

2019-8-30 NL0

Correlation between stock prices and bond yields is one of themost common misconceptions there is–largely because it’s so frequently TRUE.  It just depends upon when you look. For instance, if you only charted the 23rd through the 26th above, stocks and bonds would indeed be right on top of each other.  The more time we include in a chart, the more they diverge.

Let’s test this theory by adding even more time.  In the following chart the divergence is obvious.  Each line looks like it has a completely separate set of concerns.  But consider that 2019 has seen a relative sea-change in the Fed’s willingness to cut rates (something that helps both stocks and bonds).  Also consider that bonds are typically early adopters of a gloomy global outlook whereas stocks are still able to benefit from the here and now.  2019-8-30 NL1

In other words, the US economy is still in good shape overall, so stocks are higher, even if they’re looking a bit indecisive.  But investors are concerned the economy could be at a tipping point some time in the next 6-18 months, so longer-term bonds have to account for that risk by moving lower in yield.  Markets will be closed for Labor Day on Monday but apart from that, next week brings several important economic reports that will help investors continue to hone in on an eventual tipping point.

Various speeches from members of the Federal Reserve contributed to upward pressure on rates throughout the week. But then on Friday, all that pressure was erased in an instant or two.  What has that kind of power?  

If you were forced to guess Friday’s big market mover in advance based on a schedule of events, the most logical choice would be the Fed’s Jackson Hole symposium.  The Fed Chair often delivers a speech at Jackson Hole that’s worthy of a fairly big response.

But this year, Trump trade tweets trumped the Fed, and it wasn’t even a close call.  In fact, between China announcing tariffs on the US early in the morning and Trump’s tweets a few hours later, the trade war news on Friday completely erased an entire week’s worth of movement in both stocks and bonds.

2019-8-23 NL1

Here’s a closer look at Friday itself so we can see how Fed Chair Powell’s big speech stacked up against the trade news.

2019-8-23 NL3

Up until then, there was some question as to rates remaining inside their super low range and stocks potentially breaking above a glass ceiling marked by recently recurring highs.  Afterward, bonds look like they’re right where they want to bewhile stocks are now much closer to the other side of August’s consolidation pattern.

2019-8-23 NL4

If Friday’s brand of drama continues impacting markets, both stocks and bonds could end up breaking their respective floors seen in the chart above.  Market technicians believe that such breakouts carry an implication for additional momentum.  Paradoxically, that might be too much of a good thing for the mortgage market.

When it comes to housing and mortgages, low rates are obviously a great thing in general. But MBS (the mortgage-backed securities that dictate the rates lenders can offer) don’t cope well with huge moves and volatility.  In short, they’ve had a hard time keeping pace with the move lower in Treasury yields (which are typically almost perfectly correlated).

2019-8-23 NL2

As seen in the chart above, there are several points in time where the orange line is higher than it was in the recent past while the blue line is lower.  Today versus August 7th is a prime example.

So what’s the fallout?  It’s actually not terrible.  More than anything, this is simply something to be aware of if you need a mortgage or if your work depends upon the mortgage market in some way.  Simply put: rates are super low in general, but they’re not any lower than they were earlier this month–even though the 10yr Treasury yield is trying to convince you otherwise.

Given enough stability in the bond market (which may or may not happen any time soon) and enough passage of time (which is pretty much guaranteed if you can manage to wait for it), this paradox should begin to heal itself.  Just know that “time,” in this case, could be weeks or even months.  If Treasury yields happen to be higher by then, mortgage rates wouldn’t likely be any lower than they are now.

Fortunately, today’s mortgage rates have been just fine as far as recent data is concerned.  The “refi boom” we discussed last week continued in full swing according to this week’s data.  While we’re nowhere near the epic levels seen in 2012 and early 2013, we’ve rapidly regained the still-impressive 2016 territory.

2019-8-23 NL7

Purchase mortgage applications declined, but only due to seasonal factors (the chart above is NOT seasonally adjusted.  We always see a drop at this time of year).  Were we to examine year-over-year changes in purchases, we’d see some very promising trends in housing.  In so doing, we could set the record straight on some unnecessarily downbeat headlines regarding New Home Sales.

Please understand, the following is not an attempt to be a cheerleader for home sales–simply to focus on facts.  At the headline level, The Census Bureau reported a 12.8% decline in July–far below the forecast calling for a 0.2% decline.  But this is misleading.  June’s number was initially reported as 646k (annual pace of sales).  July’s tally of 635k would therefore be less than a 2% decline.  What gives?

The New Home Sales data is well known for huge margins of error and substantial revisions.  June contained just such a revision.  It catapulted that initial 646k number to a mighty 728k!  The interesting thing about 728k is that it’s the highest reading since the Financial Crisis.  Moreover, it’s a REVISED number and thus more “locked-in” than July’s 635k.  And if all that is too much to think about, try this: June and July combined for more New Home Sales than any other 2 months except Nov/Dec 2017.  You’d have to go back to 2007 to see anything higher.

2019-8-23 NL5

Existing Home Sales constitute a far bigger part of the housing market and the news is good there too.  July’s sales brought the annual pace to 5.42 million from a positively-revised 5.29 million in June.  It was the first time in more than a year that Existing Sales have been in positive territory versus the same month in the previous year.

2019-8-23 NL6

Long-story short: the rebound in housing that this newsletter discussed extensively in early 2019 has come to fruition.  Where we go from here remains to be seen and likely depends on economic variables and rates’ ability to stay in “refi boom” territory.

Rates themselves–although obviously susceptible to trade war surprises–will also take guidance from economic data.  In that regard, next week is more active.  Traders will be looking at manufacturing-related data to see how much the trade-related uncertainty continues to filter through to the real economy.