FED Data Details Impact of New Mortgage Rules on Mortgage Approvals

August 5, 2014

The Federal Reserve Board recently released its survey of senior bank loan officers. The July 2014 Senior Loan Officer Opinion Survey on Bank Lending Practices (SLOOS) addressed changes in the standards and terms on, and demand for, bank loans to businesses and households over the second quarter of 2014.

In the July 2014 iteration of the Survey, the Federal Reserve Board included a set of special questions on the effects on the approval rates for home-purchase loans from the Ability-to-Repay and Qualified Mortgage Standards under the Truth in Lending Act (the ATR/QM rule), which came into effect early in 2014. The National Association of Home Builders has provided a comprehensive overview of the new rule. NAHB concluded from its analysis that some creditworthy borrowers could be denied access to affordable mortgages as a result of this rule’s implementation.

According to the results of the SLOOS, a fraction of banks reported that their approvals of mortgage purchase loans from individuals were likely lower because of the ATR/QM rule. However, as Chart 1 shows, the impact of the ATR/QM rule was disproportionately felt on approvals of prime non-conforming loans and on non-traditional loans relative to prime conforming loans; regardless of the credit score associated with the prime conforming loan application. As the Federal Reserve explains, “the majority of banks reported that the new rule has had no effect on the approval rate of prime conforming mortgages, in part because those loans qualify for a safe harbor under the exemption for loans that meet the underwriting criteria of the government-sponsored housing enterprises (GSEs)”.


However, while in the aggregate, the majority of banks reported that their approval of prime conforming mortgage loan applications were not impacted by the ATR/QM rule, a disproportionate share of other, smaller, banks were more likely to report a decline in approvals of these loan applications stemming from the ATR/QM rule. As the data below indicates, a plurality of smaller banks, 50%, reported that the ATR/QM rule likely lowered their approval of prime conforming loans for borrowers with a FICO score less than 680 and 44% reported that the ATR/QM rule likely lowered their approvals of prime conforming loans for borrowers with a FICO score greater than or equal to 680. This compares to 22% and 19% respectively for large banks. Moreover, the result that a majority of all banks reported that the ATR/QM rule likely lowered their approvals of prime non-conforming and non-traditional loans, as indicated in Chart 1, reflected the impact of the rule on the approvals of these mortgage applications by smaller banks, as shown in Chart 2.



Contract Rate on New Home Loans Dips Under 4 Percent

June 26, 2014

On average, mortgage interest rates declined in May, according to data released earlier today by the Federal Housing Finance Agency (FHFA).  On conventional mortgages used to purchase newly-built homes, the average contract interest rate dropped by more than 30 basis points, from 4.19 to 3.88 percent.  This is the lowest the new home loan rate has been in a year and the first time it has dipped below 4.0 percent since February.

Contr Rate May 14

Initial fees increased slightly during the month, from 1.22 to 1.25 percent—far from enough to offset the decline in the contract rate.  The result was an average effective rate on new home loans (which amortizes initial fees over the estimated life of the loan) that also dropped by more than 30 basis points, staying barely above 4.0 percent (at 4.01).

Eff Rate May 14

Reversing the trend of the prior two months, the average size of conventional mortgages used to purchase new homes—and the price of the new homes purchased with the mortgages—both declined in May.  The average loan size declined 1.8 percent to $319,800, while the average home price fell by 3.6 percent to $418,800.  Despite the declines, both the average loan size and average new home price remain higher than they had been at any time prior to 2014.

Because the change in price was greater than the change in loan size, the average loan-to-price ratio on conventional mortgages used to purchase new homes increased substantially in May, from 77.0 to 78.6 percent—the highest it’s been since last August, and the first time above 78.0 percent in 2014.

LTP May 14

This information is based on FHFA’s Monthly Interest Rate Survey (MIRS) of loans closed during the last five working days in May. For other details about the survey, see the technical note at the end of FHFA’s June 26 news release.

Housing’s Other Contribution to GDP

June 26, 2014

Housing added 0.7 percentage points to real GDP growth in the first quarter of 2014, however the construction component (residential fixed investment) detracted from growth over the past two quarters owing to the poor weather conditions seen across the country over the period. In addition, housing accounted for 15.6 percent of total real GDP, which is low from a historical perspective.

The price component of GDP is becoming more interesting, however. Starting in 2012, housing has been contributing more to prices, for the most part, as seen in the figure below. In the first quarter of this year, it added over 0.7 percentage points to growth in the GDP price index, led by nearly 0.5 percentage points in the construction component (the most since the third quarter of 2008); the overall price index rose 1.3 percent. This is partly due to increased construction costs and higher rental prices.

GDP Price

Currently, overall inflation remains subdued and is within the Federal Reserve’s comfort zone. But moving forward, if housing becomes an even larger share of GDP, it would likely place upward pressure on inflation measures in general.


Catch Me If You Can

June 24, 2014

House prices grew by 10.8% between April 2013 and 2014, according to the S&P/Case-Shiller 20-City Composite Home Price Index, which was less than the 12-month growth rate of 12.4% seen in March. Similarly, the Federal Housing Finance Agency’s Purchase-Only Index rose 6.0% compared to 6.4% in March. On a seasonally-adjusted monthly basis the 20-City Composite index increased by 0.2%, while the Purchase-Only index was virtually unchanged. Both indices show that annual house appreciation has slowed over the past five consecutive months ending in April and suggest the housing market may be returning to its long-run trend of growth.

Among the 20 metro areas, Las Vegas experienced the largest annual gains (18.8%), followed by San Francisco (18.2%) and San Diego (15.3%). Meanwhile, cities experiencing the smallest increases include Cleveland (2.7%), Charlotte (4.4%) and New York City (5.4%). In addition, some areas are exhibiting large price gains for lower-priced homes as the chart below demonstrates. In April, Atlanta saw a 35.8% annual increase in homes under $153,000 and San Francisco saw homes under $496,200 rise 30.6%. In contrast, New York City saw an increase of 4.8% and Boston a 9.7% gain.


The growth rates may be inflated, however, due to previously distressed homes being resold. For instance, CoreLogic provides a comparable house price index to the S&P/Case Shiller series and also has an additional one that excludes distressed homes. Foreclosed properties generally sell for a discount, currently at around 18 percent in April, according the National Association of Realtors, which suggests they would weigh down the overall market index. Yet since 2012, it appears distressed sales have added to the market’s growth, as seen the figure below.


What’s occurring is during the housing crisis, many homes were foreclosed upon. In fact, the number of homes in foreclosure eventually peaked in the first quarter of 2010 at over two million, based on data from the Mortgage Bankers Association. These foreclosed homes were bought at deep discounts, ranging up to 40 percent and even higher in certain areas by 2009, as the chart above indicates when distressed sales are included. However, once the homes were repaired and later sold again either in the near-term for profit or when the new homeowners decided to move, the sales price was dramatically higher than the original price, owing to those heavy discounts. And as house prices recovered in general, those who waited longer to sell saw even larger increases in their sales price relative to the original purchase price. For instance, over the 12-months ending in April of this year, the inclusion of distressed properties added over two percentage points to overall growth in the price index, but the actual sales price for those troubled homes was likely at or below those of comparable non-distressed homes. So, this is more of a catch up to the race than a grand spree in house prices.

For full histories of the FHFA US and 9 Census divisions, click here.

For full histories of the composites and 20 markets included in the Case-Shiller composites, click here

MBA Introduces New Index Tracking Mortgages for New Homes

June 16, 2014

According to the Mortgage Bankers Association’s (MBA) Builder Application Survey (BAS), mortgage applications for new home purchases decreased by a not seasonally adjusted monthly rate of 8.4% in May 2014. However, on a 12-month basis, mortgage applications for new home purchases in May 2014 were 4.9% higher than their level in May 2013. This is the fifth consecutive month of year-over-year increases in mortgage applications for new home purchases. The last 12-month decrease in mortgage applications for new home purchases took place in December 2013 (-15.6%), at which point the BAS reached a 22-month low. Since December 2013, the BAS has risen by 57.8%

According to the MBA, its BAS tracks application volume from mortgage subsidiaries of home builders across the country and participants in the survey collectively account for approximately 20.0% of new home sales based on the Census Bureau estimates. The MBA also notes that since the mortgage application is typically made around the same time that the sales contract is signed, capturing the number of mortgage applications will give an indication of new home sales.


The figure above compares the MBA’s BAS, which tracks changes in mortgage applications for new home purchases, with Census’ not seasonally adjusted new home sales data. According to the figure, the BAS rose by 37.2% between December 2012 and March 2013. Over the same period, not seasonally adjusted new home sales rose by 46.4%. Between March 2013 and December 2013, the BAS fell by 38.4% while new home sales declined by 24.4%. Since December 2013, but excluding May 2014, the BAS rose by 72.3%, corresponding with a 32.3% increase in new home sales.

Incorporating the results of the BAS, the MBA estimates a not seasonally adjusted level of 36,000 new home sales and a seasonally adjusted annual rate of 374,000 new home sales in May 2014. This is an interesting project that MBA has taken on, constructing an analogy to NAR’s pending home sales index, for newly constructed single family homes, based on mortgage applications. It’s a relatively new project so the history of the index is short, but it will be interesting to see how the index performs as a predictor over time.

Rates on New Home Loans Remain Stable

May 29, 2014

The average characteristics of conventional mortgages used to buy newly built homes were little changed in April, according to data released earlier today by the Federal Housing Finance Agency (FHFA).  The average contract interest rate declined a scant 2 basis points to 4.19 percent, while the average initial fees and charges increased by a single basis point to 1.22 percent.  As a result, the effective interest rate (which amortizes initial fees over the estimated life of the loan) also declined by 2 basis points, to 4.33 percent.Eff Rate May 14Meanwhile, the average size of the conventional mortgages used to purchase newly built homes increased from $322,600 to $325,800 in April, while the average price of the new homes purchased with the loans went from $427,200 to $434,500. This is the second consecutive month during which both the loan size and new home price both increased, although neither is yet quite back to the peak it reached in January.Loan Size May 14Price May 14This information is based on FHFA’s Monthly Interest Rate Survey (MIRS) of loans closed during the last five working days in April. For other details about the survey, see the technical note at the end of FHFA’s May 29 new release.

Apartment and Condominium Market Shows Positive Growth in First Quarter

May 29, 2014

The NAHB Multifamily Production Index (MPI) increased three points to 53 in the first quarter of 2014, which is the ninth consecutive quarter with a reading of 50 or above.

The MPI measures builder and developer sentiment about current conditions in the apartment and condominium market on a scale of 0 to 100. The index and all of its components are scaled so that any number over 50 indicates that more respondents report conditions are improving than report conditions are getting worse.


The MPI provides a composite measure of three key elements of the multifamily housing market: construction of low-rent units, market-rate rental units and “for-sale” units, or condominiums. The MPI component tracking builder and developer perceptions of low-rent units increased one point to 48 and for-sale units jumped eight points to 54. Meanwhile, the index tracking market-rate rental properties slipped one point to 59, but has remained consistently above 50 since the fourth quarter of 2010.

The MPI shows stable production of apartment units in 2014, which is in line with NAHB’s expectation of a 6 percent increase in multifamily starts this year.

The Multifamily Vacancy Index (MVI), which measures the multifamily housing industry’s perception of vacancies, dropped one point to 37. With the MVI, lower numbers indicate fewer vacancies. The MVI improved consistently through 2010 and has been at a fairly moderate level since 2011 after peaking at 70 in the second quarter of 2009.


Historically, the MPI and MVI have performed well as leading indicators of U.S. Census figures for multifamily starts and vacancy rates, providing information on likely movement in the Census figures one to three quarters in advance.

Remodeling by Zip Code: NAHB Releases 2014 Projections

May 26, 2014

NAHB just released new estimates of spending on improvements to owner-occupied housing by zip code.  The estimates are based on a model relating improvement spending to five key variables (number of homes in the area, the share built in the ’60s, share built in the ’70s, owners’ average income and level of education), projected to calendar year 2014.  As before, the estimates show total spending on improvements, as well as spending per home, in each zip code.

On average, total spending on improvements in a zip code is projected to be about $5.1 million in 2014.  The top 5 total-spending zip codes are all in Maryland, Texas, or Illinois.  Each of these top 5 zips contains at least 15,000 owner-occupied homes and home owners who average at least $145,000 in income and are 60 percent or more college educated.  Most of these top 5 zips don’t have an unusually large share of homes in the key vintage for remodeling (homes built from 1960 to 1979), except for the zip at the very top of the list—#20854 in Maryland, a close-in suburb of Washington DC.  20854 is the only zip where over $60 million in spending on improvements is projected for 2014, and over half the owner-occupied homes in that zip were built 1960-1979.

The map below illustrates improvement spending per owner-occupied home.Remodel_zecta14

In the average zip code, improvement spending per home is just under $1,600.  The top two zip codes by this measure are the same as they were in 2013: #94528 (in a suburb of San Francisco) and #10007 (in Manhattan).  These are the only two zips with projected spending per home of over $5,000 (even over $6,000 in 94528), although several others come fairly close.  This year, unlike last, a zip code in New Jersey cracked the top five—#07078, a suburb within commuting distance of New York City. Homeowners in the top 5 improvement-per-home zips are at least 92 percent college educated and have average incomes of at least $350,000. Nearly half the owner-occupied homes in top zip #94528 were built 1960-1979.

The improvement spending projections for 2014 are based on a statistical model developed by NAHB economists using data from the HUD/Census Bureau American Housing Survey and summary statistics for the Census Bureau’s approximation of zip-code areas from the American Community Survey. Members of NAHB Remodelers can log onto NAHB’s web site with their usernames and passwords and access improvement spending projections in more than 25,000 zip codes here. Projections aggregated to the state level are available to everyone, and are reproduced below:

State Remodeling 2014

Efficient Windows Top List of Green Products Used by Remodelers

May 22, 2014

Energy efficient windows emerged as the leading green product among remodelers responding to NAHB’s Remodeling Market Index (RMI) survey for the first quarter of 2014, as nearly 9 out of 10 remodelers surveyed said they’d commonly used low-e windows during the past year.  Next on the list were high efficiency HVAC systems and programmable thermostats at 70 percent each, closely followed by ENERGY STAR appliances at 69 percent.

Green Remod

Although the features at the top of the list all involve energy eficiency, the term “green” is usually defined more broadly than that.  Moisture control, for example, is classified as green here, because it results in some components of the home needing to be replaced less often, reducing environmental impacts associated with manufacturing, transporting and installing those components over time. The list of 23 green products and practices used in the RMI survey is based on the major sections of the National Green Building Standard (which can and should be applied to remodeling as well as new construction).

Given the difference in cost, it’s perhaps surprising that use of program-mable thermostats is no more common than use of high efficiency HVAC among remodelers.  Anecdotally, several NAHB members have reported that a small but discernible share of their customers tend to resist devices that require programming.  A similar result was found in a survey on green products and practices used by single-family builders.

It’s also interesting that, across the two surveys, the same four green features appear at the top of the list and in the same order for both remodelers and builders.  The remodelers’ percentages tend to be a little lower, but this is natural, because not every remodeling project involves every home component. High efficiency HVAC systems, for example (the second ranked green feature for both builders and remodelers) are commonly used by 90 percent of builders, compared to 70 percent of remodelers.  But remodelers who specialized in projects like replacing windows or building decks in 2013 may have seldom if ever needed to install HVAC systems, while builders of new homes would have, at some point, dealt with every aspect of HVAC.

Mortgage Delinquency Rates Fall

May 20, 2014

Data released by the Mortgage Bankers Association (MBA) indicates that the delinquency rate for mortgage loans on one-to-four-unit residential properties, considered single-family properties, decreased to a not seasonally adjusted rate of 5.69% of all loans outstanding at the end of the first quarter of 2014, 106 basis points below the 6.75% delinquency rate recorded in the first quarter of 2013. This level represents the lowest level since the first quarter of 2008.

According to MBA’s National Delinquency Survey, the four-quarter decrease in the delinquency rate reflected a decline across each stage of delinquency. In addition, the foreclosure inventory also fell. As Figure 1 below illustrates, the percentage of all loans past due fell by 106 basis points over the past four quarters. Loans 30-59 days past due fell by 45 basis points, loans 60-89 days past due fell by 15 basis points, and loans 90 or more days past due decreased by 45 basis points. The foreclosure inventory fell by 90 basis points over the past four quarters. In sum, the serious delinquency rate, the portion of loans either 90 or more days late or in the foreclosure inventory decreased by 135 basis points over the past year.


The recent four-quarter decline in the portion of loans that are seriously delinquent furthers a trend in place since 2010. Figure 2 compares the share of loans that became seriously delinquent based on their origination year with the annual average serious delinquency rate. According to this illustration, the share of single-family loans considered seriously delinquent peaked in 2010 at 9.0%. Since 2010, the proportion of seriously delinquent mortgages has fallen to 5.8%.

The extended decline in the percentage of loans considered seriously delinquent partially reflects a decrease in serious delinquency among loans originated in more recent years. According to the figure below, an average of 4.0% of loans originated in 2005 became seriously delinquent. However, 10.6% of loans originated in 2006, 11.5% of loans originated in 2007, and 6.4% of loans originated in 2008 ultimately became seriously delinquent. In contrast, 1.0% of loans originated in 2009 became seriously delinquent. Since 2010, less than 1.0% of originated loans have become seriously delinquent.