Mortgage Rates Fall Again

Information provided by the Federal Housing Financing Agency’s Mortgage Interest Rate Survey indicates that mortgage rates on purchases of newly constructed homes fell over the month of May to 4.00 percent, from 4.02 percent in April. Since rising by 64 basis points between October 2016 and February 2017, rates have now slipped 18 basis points over 3 consecutive months.

A previous post illustrated that the longer-term trend in the 30-year fixed rate mortgage reflects changes in the rate on the underlying 10-year Treasury Note. Over the October 2016 to February 2017 period, the U.S. 10-Year Note rate rose by 66 basis points to 2.42 percent. The 66 basis point increase in the 10-Year Treasury Note rate over this time period would imply that the average mortgage risk premium, the difference between the rate on the conventional 30-Year Fixed Rate Mortgage for the purchase of newly constructed homes and the 10-Year Treasury Note rate, fell by 2 basis points over the same period. Over the more recent February-May period, the rate on the 10-Year Treasury Note fell by 12 basis points to 2.30 percent, twice the implied change in the risk premium*.

The figure above suggests that the trend in the 10-Year Treasury Note rate has had a larger impact on changes in the mortgage rate on purchases of new construction, although the risk premium associated with mortgage lending has also played a role, albeit a smaller one. With some assumptions, researchers can decompose the 10-Year Treasury Note rate into its inflation component and its “real” or inflation-adjusted component by comparing the yield on the 10-Year Treasury Inflation Protected Securities with the rate on the 10-year Treasury Note. Since the yield on the 10-Year Inflation Protected Securities is indexed to the Consumer Price Index (CPI), then the spread between these two yields should also provide information about the market’s view of inflation**.

The figure above indicates that the increase in the 10-Year Treasury Note rate over the October 2016 to February 2017 period reflected a nearly equal increase in both the market’s view of inflation expectations and the real return on the 10-year Treasury rate. However, over the February to May period, the decline in the 10-Year Treasury Note rate overall reflected a decrease in the markets expectations for inflation. Meanwhile, the real return on the 10-Year Treasury Note rose.

However, relying solely on the bond market for information about trends in inflation may not yield accurate results. In addition to the assumptions underlying this calculation, other factors may be impacting the recorded yields on these securities. For example, the Federal Open Market Committee’s quantitative easing programs, which involved significant purchases of longer-term Treasury securities, may mask the private market’s view of inflation as implied by bond yield spreads. Comparing results from bond yields with survey data on consumer prices can help to confirm whether inflation accelerated toward the end of 2016 and whether it has slowed in recent months.

As shown in the table above, 12-month changes in the CPI indicate that inflation had accelerated between October 2016 and February 2017, from 1.64 percent to 2.80 percent. Since February 2017, inflation has slowed to 1.87 percent. Similarly, using changes in the price index on the Personal Consumption Expenditures component of the National Income Product Accounts (i.e. GDP) shows that inflation had largely been accelerating between October 2016 and February 2017 before slowing in each month since February***.

Taken together, the recent rise and, so far, smaller decline in the mortgage rate largely reflected changes in the 10-Year Treasury Note rate, although, to a smaller extent, the implied risk premium played a role as well. Over the end of 2016, the increase in the 10-Year Treasury Note rate reflected both an increase in the expectations for inflation as well as the real yield on Treasury securities. However, the decline in mortgage rates over the February-May period likely reflected, in large part, a slowdown in inflation. This analysis then suggests that there are at least two links between inflation and housing demand, the house price channel and the mortgage rate channel.

* According to the FHFA, “The indices are based on a small monthly survey of mortgage lenders, which may not be representative”. When we substitute the rate on the 30-Year Fixed Rate Mortgage rate as provided by Freddie Mac, a more commonly used rate, the impact of the risk premium over the February-May period shrinks even more.

** More directly, calculating this “breakeven rate” provides information on the markets expectations for inflation over the 10-year window. However, an idea known as Rational Expectations Theory posits that expectations of price changes are equal to prices changes in the present, absent any shocks (i.e. energy price volatility).

*** In fact, inflation had been accelerating since the summer.

Tags: , , , , ,