In the 1950s and 1960s, mortgage interest rates didn’t change very much, so a typical homeowner bought a house, got a mortgage, and paid it off over 30 years. However, increased volatility of interest rates since the 1970s means that it is profitable to refinance your mortgage when the mortgage interest rate declines. Certainly, people who bought their homes in the 1980s when interest rates were above 10 percent should have refinanced their mortgage loans at some point. It only takes a small decline in the mortgage interest rate to make refinancing worthwhile. The difference in rates has to make a big enough difference in your monthly payments to make it worth the time, effort, and expense to refinance.
Almost everyone who took out a mortgage before 2009 could have profitably refinanced in 2011, 2012, or 2013. Mortgage interest rates dropped substantially from 2009 to 2013, so a homeowner is certain to pay less by getting a new mortgage than in sticking with the old mortgage. The cost of refinancing is generally about 1 percent of the total amount you are borrowing, but with a drop of 1 to 2 percentage points in the mortgage interest rate, refinancing will pay for itself within a year. So, why doesn’t everyone refinance? Part of the reason is that in the financial crisis, the value of many homes declined sharply, and banks will not refinance a mortgage loan for more than the home is worth. These days, banks are generally willing to only refinance your mortgage up to 80 percent of the home’s value, so the decline in home prices means that many homeowners will not be able to refinance their mortgages.
Given the ability of homeowners to refinance their mortgages, you can see why the percentage of the mortgage market that represents refinancing of existing mortgages (compared with financing new mortgages) rises as the mortgage interest rate declines. This relationship is shown in the chart. The blue dotted line shows that refinancing as a share of all mortgage loans is sometimes as low as 12 percent and as high as 72 percent. As the mortgage interest rate declined from 11 percent in 1990 to 3.5 percent in early 2013, the share of refinancing in the mortgage market generally increased. Whenever the mortgage interest rate rises appreciably, as it did in 1994, 1999, 2005, and mid-2013, you see that refinancing activity drops. And when the mortgage interest rate declines, many people refinance their homes and the market share of refinancing rises.
Chart: Share of Refinancing in Mortgage Market
Source: Federal Reserve Board data on 30-year conventional mortgage interest rate, Mortgage Bankers Association share of refinancing
In the previous blog, I showed you data on mortgage interest rates. But inflation has distorted some of the movements of these interest rates over time. You see much higher interest rates in the late 1970s and early 1980s than in other periods, which was caused by the higher inflation in that period. If we adjust for the difference in inflation by subtracting from the interest rate a measure of expected inflation, then we get a more accurate view of the cost of borrowing for a home, as the chart here shows.
Chart: Inflation-Adjusted Mortgage Interest Rate
Source: Federal Reserve Board data on 30-year conventional mortgage interest rate, adjusted by author for expected inflation using forecasts of inflation from Survey of Professional Forecasters, Federal Reserve Bank of Philadelphia
After adjusting for inflation, we can see that the mortgage interest rate is very low by historical standards today. After a lot of volatility in the mortgage interest rate in the 1980s, it settled down and averaged about 5 percent in the 1990s, then declined to a range of 3 to 4 percent in the early 2000s. After the financial crisis in 2008, the mortgage interest rate dropped dramatically to under 2 percent in 2011, reaching its all-time low at the end of 2012 of 1.1 percent. Most recently (as of the first quarter of 2016), the rate is 1.6 percent, still extremely low.
By adjusting the mortgage interest rate for expected inflation, we arrive at a measure that economists call the “real” interest rate, which is a more appropriate measure than the usual “nominal” interest rate. The nominal interest rate tells you how many dollars you have to pay to cover the interest on your mortgage loan. The real interest rate tells you, more appropriately, the amount of goods and services you need to forgo to cover the interest on your mortgage loan. For example, in the first quarter of 2016, the nominal interest rate was 3.7 percent, the long-term expected inflation rate was 2.1 percent, so the real interest rate was 3.7% – 2.1% = 1.6 percent. You expect to have to pay 3.7 percent more in dollar terms each year to cover the interest on your mortgage, but those dollars are expected to be worth 2.1 percent less each year, so the true cost in terms of goods and services you forgo to pay the interest on your loan is 1.6 percent each year.
The mortgage market is just one of many markets competing for funds in the overall marketplace for credit. As such, funds tend to flow across markets in response to increased demand. Interest rates in all markets are thus correlated, with differences depending on factors such as the risk to any particular sector of the economy. It turns out that the interest rate on mortgages is linked fairly closely to the interest rate on ten-year government bonds. The interest rate on 30-year mortgages averages about 1.7 percentage points above the interest rate on 10-year U.S. Treasury bonds, as Chart 4 shows. During the financial crisis, the difference between the two interest rates increased quite a bit because of the perceived risk to mortgage loans, ranging between 2 and 3 percentage points from November 2007 to March 2009, but has declined since. In May 2016, the interest rate on mortgages was about 3.6 percent, while the interest rate on 10-year government bonds was about 1.8 percent, so the 1.8 percentage point difference between the two was not much above the historical average of 1.7 percentage points.
Chart: Mortgage and Ten-Year Government Bond Interest Rates
Source: Federal Reserve Board data on interest rates
If we look at the total amount of mortgage debt outstanding as a percent of GDP (see chart), we see that it has grown steadily over time. Mortgage debt rose from 20 percent of GDP in 1955 to about 30 percent in 1965, and then stayed near 30 percent until 1985. In the 7 years from 1985 to 1992, mortgage debt grew another 12 percentage points to 42 percent of GDP. It didn’t rise much over the next 8 years, reaching 46 percent of GDP by 2000. But then the combination of low interest rates and sub-prime lending led mortgage debt to rise very sharply, hitting a peak of 73 percent of GDP in 2009. It has declined significantly since the financial crisis, dropping 21 percentage points from 2009 to 2015, standing at 52 percent of GDP in late 2015. But the level of mortgage debt outstanding relative to GDP is still far above where it was before 2000 when the housing boom accelerated, so it may still need to drop before the housing market is restored to full health.
Chart: U.S. Household Mortgage Debt as a Percent of GDP
Source: Federal Reserve Board data on mortgage debt, rescaled by author using GDP from Bureau of Economic Analysis
From an economic point of view, Chart 1 (in the last blog) is a bit misleading because a dollar in 1975 is worth a lot more than a dollar in 2015, thanks to inflation. To adjust for inflation, we can use a measure of consumer prices. The resulting chart of inflation-adjusted home prices (Chart 2) gives a different picture than Chart 1. You can see that, in inflation-adjusted terms, housing prices grew rapidly in the second half of the 1970s, were fairly stable in the 1980s and early 1990s, and then began rising rapidly again in the second half of the 1990s until 2007. But, beginning early in 2007, inflation-adjusted home prices began to decline substantially. Over the five-year period from the peak in 2006 to their low point in 2012, inflation-adjusted home prices declined an average of 6 percent per year.
The data in Chart 2 can also be used to answer the question: is investing in a home a good financial investment? The answer is a clear NO! Using the data and calculating the inflation-adjusted annual return on homeownership, from 1975 to 2015, home prices rose only 1 percent per year, far less than stocks and bonds over the same period. Even at their peak in 2006, inflation-adjusted home prices had appreciated only 2 percent per year. So, if you buy a home, you are buying it because it provides shelter, not because it is a great financial investment.
Chart 2: U.S. Home Prices (Inflation-Adjusted)
Source: Federal Housing Finance Agency (FHFA) All-Transactions Home Price Index, rescaled by author and adjusted for inflation using personal consumption expenditures price index from Bureau of Economic Analysis
In the housing bubble in the early to mid-2000s, sensible ideas about how much debt a home buyer should take on were ignored by everyone—bankers and homeowners. Both foolish lenders and greedy homeowners were swayed by the fact that home prices had been increasing substantially over the past 10 years. If we look at Chart 1 showing quarterly data on national housing prices, you can see why they might be fooled. The chart plots a measure of prices of homes nationwide from 1975 to 2015. In the chart, you can see that until 2007, national house prices had never declined for more than one quarter in a row, and had increased in every year; so both lenders and homeowners were fooled into thinking that homes were a safe asset. Second, the appreciation in house prices from 2000 to 2005 averaged 8% per year. This led lenders and homeowners into falsely thinking that a homeowner’s income and ability to pay were no longer relevant. After all, if the home appreciated by 8% per year, it would be easy to sell the home and pay off the mortgage without a loss to either side. So, lenders started making sub-prime loans, which allow people to qualify for loans at percentages well below standard rules, such as those that limit the monthly housing payment to be less than 28 percent of a household’s income, and limit the total of a household’s monthly housing payment and payment on other debts (credit cards or student loans) to be less than 36 percent of its income. Some loans were even known as NINJA loans (because the homeowner had No Income, No Job, and no Assets). Those loans relied entirely on the appreciation of the home to be valid; as the home appreciated in value, the homeowner would borrow even more to pay off the original mortgage loan. But all of this sub-prime lending turned out to be foolish because it relied on the continued appreciation of house prices. When home prices began to decline in 2007, panic set in. Everyone began to realize that these loans would never be paid off. The worst part of the story is that sub-prime loans had the potential to greatly increase the homeownership rate among low-income households, but after the bubble burst, that dream vanished. So, the main benefit of allowing sub-prime lending in the first place, which was to increase homeownership among the poor, actually made things worse.
Chart 1: U.S. Home Prices
Source: Federal Housing Finance Agency (FHFA) All-Transactions Home Price Index, rescaled by author
This is a blog with ideas for a book on using economics in daily life. The idea is that many people could benefit from ideas and data known by economists, but we economists do not always communicate them in ways that people can understand. Economists tend to talk to each other, instead. So, this blog is my attempt to remedy that. I’m starting with discussions about housing and mortgages.