I’ve been mentioning MBS outperformance and underperformance quite a bit recently, but what do these terms really mean and why do we care?
First off, we always care about strength and weakness in mortgage bonds as they are the primary building block in determining mortgage rates (i.e. higher MBS prices = lower rates, all other things being equal). We also care about strength and weakness in Treasuries because mortgage bonds typically correlate amazingly well with Treasuries (so much so that many mistakenly believe 10yr Treasury yields serve as a basis for mortgage rates).
The type of volatility seen in March exposed a few eye-opening truths. The first was that mortgage bonds can and do deviate from 10yr yields at times. March offered the most extreme example we’ve seen since the financial crisis. To understand why this happened, just consider the amount of investor fear and an uncertainty. Treasuries were a much safer investment at the time and in an illiquid market, a much easier investment to trade compared to MBS. Add in fear about how forbearances would affect mortgage bond cash flows and it made perfect sense for investors to pay less and less for mortgage debt relative to US Treasury debt.
As we know, when investors are paying less for bonds, bond prices are falling and rates are rising. We know that MBS dictate mortgage rates, but a mortgage rate is not the yield associated with MBS in the same way a Treasury yield is a function of Treasury bond prices. Rather, there are actual yields that can be calculated for MBS. These yields rely on some subjective math and guesswork. After all, investors know exactly how and when Treasuries will be repaid whereas mortgage repayment is dependent on the future path of rates and consumer behavior. Still, yields can be calculated. It’s the comparison of those calculated yields and actual 10yr Treasury yields that result in the following chart, which simply measures the MBS yield minus the 10yr Treasury yield.
One way for MBS Live members to get a rough idea of MBS outperformance/underperformance is to compare the day-over-day change in PRICES between 30yr fixed UMBS 2.0 coupons and 10yr Treasury PRICE (not yield). This is far from a perfect comparison, but looking at price vs price is better than trying to interpret price vs yield.
The chart above suggests MBS are doing just fine versus Treasuries and have managed to avoid moving back up into the historically weaker/higher spread range seen in October 2019 (and of course in March 2020). But as we know, mortgage rates took their own sweet time grinding back down to all-time lows even though MBS prices were much higher than they were in 2012 (the reigning champ in the all-time low department until last week). If mortgage rates are based on MBS, what’s up with that?
The answer is very simple: MBS are merely the primary building block of rates. There are other factors. They’re usually so boring and stable that there’s no need to consider the changes in the costs associated with those factors, thus allowing us to follow MBS as an exclusive indicator of mortgage rates. But certain realities of the coronavirus response made for unprecedented volatility in those other factors. These include servicing valuations other things that can impact lender margins (such as uncertainty about unexpected fees and expenses. The 5-7 point hit for selling loans in forbearance is a good example).
To get an accurate idea of just how big the impact has been from these other factors, we can chart average 30yr fixed mortgage rates versus the same MBS yield that fed the calculation in the chart above. Mortgage rates tend to be very well behaved vs MBS yields over the years. We’ve never seen a divergence like the one caused by coronavirus.
As the chart shows, the wounds have begun to heal, but we’re still at extreme levels, historically, even when compared to the previous all-time low rate time frame in late 2012. This is why it’s important for us to follow spreads. It lets us know that there is more room for improvement in rates as temporary barriers continue to be torn down. More simply put, as the coronavirus situation becomes better understood and growth in forbearance continues to slow down, rates could fall even if MBS prices/yields are not improving.
A word of caution though: I wouldn’t look at the chart above and conclude rates are guaranteed to move lower. Spreads could certainly tighten in an environment where Treasury and MBS yields are moving higher fast enough to offset the spread tightening. In order for this to work out, we need the broader rate market to remain fairly stable in at least a somewhat similar range to what we’ve seen in April and May 2020. IF that happens, then yes… mortgage rates should continue moving lower.
As for today specifically, it’s just another day in a week before a 3.5-day holiday weekend. Trading could get weird by the end of it. There’s a lot of uncertainty surrounding the 20yr Treasury auction today (the first time we’ve seen a 20yr bond since 1986). Beyond that, Treasuries could continue keeping an eye on stocks, which are now surging back above yesterday afternoon’s highs. Add to that an ongoing deluge of corporate debt (read this primer to understand that) and we could see a lot of random pressure on Treasuries by the end of the day. If we DON’T, we could choose to view that as a positive indication of our willingness to hold a range under .745%, OR merely the fact that traders are exiting short positions ahead of the long weekend (exiting a short position in bonds = buying = lower yields).