United States - FHFA Terms on Conventional Mortgages: Effective rate

United States: FHFA Terms on Conventional Mortgages: Effective rate

Unit % p.a., NSA
Adjustments Not Seasonally Adjusted
Monthly 1.95 %
Data May 2017 4.02
Apr 2017 4.1

United States: Real Estate

Reference Last Previous Units Frequency
Building Completions Apr 2020 1,176 1,279 Ths., SAAR Monthly
House Price Value for Existing Homes Apr 2020 286.98 289.4 Ths. USD, SA Monthly
House Price Value for New Homes Apr 2020 312,351 326,238 USD, SA Monthly
Residential Building Permits Apr 2020 1,066 1,356 Ths. #, SAAR Monthly
Residential Housing Starts Apr 2020 891 1,276 Ths. #, SAAR Monthly
House Price Index 2020 Q1 454.86 450.97 Index 1980Q1=100, NSA Quarterly
Dwelling Stocks 2019 139,684 138,516 Ths. # Annual

Release Information

The FHFA Monthly Interest Rate Survey (a.k.a. the Monthly Survey of Rates and Terms on Conventional Single-Family Non-farm Mortgage Loans, a.k.a. Terms on Conventional Home Mortgages) reports  interest rates, purchase prices, and loan-to-value ratios for conventional single-family mortgages. The data are collected from a convenience sample of mortgage lenders, including savings associations, savings banks, commercial banks, and mortgage companies. Survey values are computed by weighting individual loans by lender size and geographic location.  Includes the “National Average Contract Mortgage Rate For the Purchase of Previously Occupied Homes By Combined Lenders.”

Survey universe

The sampled mortgages are single-family, conventional, non-jumbo, and (after 2008) exclusively fixed-rate.

The data are collected from a small monthly survey of mortgage lenders which may not be representatative.  It is a convenience sample, not a statistical sample.  The group includes savings associations, mutual savings banks, commercial banks, and mortgage companies.  Data are weighted to reflect the shares of mortgage lending by lender size and type as reported in the latest release of the FRB's HMDA data.

The sample has shrunk progressively:

  • 1999: 200 lenders reported a total of 254,000 individual loans
  • 2007: about 70 lenders and 165,000 loans
  • 2009 and later: 20-50 lenders and 4,000-7,000 loans

Mortgage indicators

  • The composite rates include all mortgage loans in the survey, both FRM and ARM.
  • Contract rate (%)
  • Initial fees and charges (%)
  • The effective interest rate (%) reflects the amortization of initial fees and charges over a 10-year period, which is the historical assumption of the average life of a mortgage loan.
  • Term to maturity (years)
  • Loan amount (thousands of dollars)
  • Purchase price (thousands of dollars)
  • Loan-to-price ratio (a.k.a. loan-to-value ratio, LTV) (%) - Does not equal (100*MLOAN.US/PPRICE.US).  It is a weighted aggregate of the ratios for each loan in the survey.
  • Percent of all loans per loan-to-price ratio class (%) - Some slices of the data present the distribution of loans by LTV: under 70%, 70%-80%, 80%-90%, over 90%.
  • Percent with adjustable rates (%)

"Share of market"

The NADJUST? series report the portion of each slice (denominator) with have adjustable rates (i.e., are ARMs).  For example, LTV70CM.US and NADJUSTCM.US are fractions of the entire survey, but LTV70CXM.US and NADJUSTCXM.US are fractions of the subset of previously-occupied homes.

For the 15- and 30-year data (table 4), the shares are of the total market.  For example, IRFHFBSF30M.US is based on the number of non-jumbo fixed-rate 30-year loans (numerator) and the total loans in the survey (denominator).

Home price

Median sales prices on single-family existing (previously occupied) homes are reported each year by state (since 1991) and for each of the 12 Federal Home Loan Bank districts (since 1973).

Geographic vs. time detail

  • The national level is reported monthly.
  • The state level is reported annually.
  • The metro level is reported quarterly and annually for 33 selected large areas. (A note for Data Buffet users: The areas are not all MSAs, so you'll have to switch between several "geography types" in the Geography Wizard.)


The Monthly Interest Rate Survey asks major lenders to report the terms and conditions on all conventional, single-family, fully amortized, purchase-money loans closed on the last five working days of the month.

The reported information is based on fully amortized mortgage loans used to purchase single-family homes. Loans used to refinance houses are excluded, as are non-amortized and balloon loans.

The survey reports only conventional mortgages, and thus excludes mortgage loans insured by the Federal Housing Administration (FHA) or guaranteed by the Veterans Administration (VA). This exclusion imparts a degree of bias to the house-price data in that, on average, homes financed with FHA or VA mortgages are lower priced than those financed with conventional mortgages, therefore, the reported house prices should not be interpreted as applying to all single-family homes.

Conventional mortgages are used in 70% of house purchase transactions, and all-cash sales are 7% of purchase transactions. FHA and VA mortgages finance about 23% of house purchases.

Typically, the interest rate is determined 30 to 45 days before the loan closes.

Data on weighted shares of mortgage lending by lender size and type are as reported in the latest release of the FRB HMDA data.

About the NACMIR

The National Average Contract Mortgage Rate is produced as part of the Monthly Interest Rate Survey (MIRS) Many lenders use the rate when adjusting certain ARMs. In the early 1980s, it was the only index rate that federally-chartered S&Ls could use as an ARM index.  It's a lagging indicator because the rate on a loan at closing often reflects a commitment made several months earlier, and because of processing time.

The rate is based on a sample of mortgage lenders (savings associations, mortgage companies, commercial banks, mutual savings banks) who report on existing-home loans closed during a five-day sample window each month. (Since November 1991, the window comprises the last five working days in the month; through October 1991, it was the first five working days.) It lags the reference month by two months, i.e., it's published at the end of the month after the reference month.

Specifically, MIRS includes the terms and conditions on all conventional, single-family, fully amortized, purchase-money loans closed by the respondents; the data thus excludes FHA-insured and VA-guaranteed mortgages, refinancing loans, and balloon loans.

The effective interest rate includes the amortization of initial fees and charges over a 10-year period, which is the historical assumption of the average life of a mortgage loan. The data is weighted to reflect the shares of mortgage lending by lender size and lender type as reported in the latest release of the Federal Reserve Board's Home Mortgage Disclosure Act data (HMDA, also available on Data Buffet).


The data from this survey are used by Fannie Mae and Freddie Mac in setting the "conforming" loan limit -- the maximum size of mortgage loan that these secondary market agencies can purchase or guarantee. The limit for 1999 was $240,000, rising by 2008 to $417,000. Because this excludes buyers in many high-priced urban areas from obtaining conforming loans, in mid-2008 Congress legislated a higher limit -- 50% higher in four high-cost areas (Alaska, Hawaii, Guam, and the U.S. Virgin Islands) and up to $729,750 in other specific areas.

Because NADJUSTCM.IUSA was discontinued by the source, we produce an estimate RXNADJUSTM.IUSA. This estimate equals NADJUSTCM.IUSA from Jan1985 to Oct2008. From Nov2008 onward, we exend the series using MBAACNW.IUSA from the Mortgage Banker's Association Weekly Application Survey, converted to monthly frequency and shifted upward by a constant value of  0.42849088.

No revisions are made to MIRS results.

Sample changes

The survey is conducted with a convenience sample, not a statistical sample, and it may not be representative of the entire market.  Please consult other datasets on Data Buffet for complementary views.

Starting with November 2008 data, figures related to ARMs were no longer reported, because the MIRS no longer had a statistically valid sample; that is, the market for ARMs had shrunk dramatically.  (Concept codes NADJUST?.)

Geographic changes

As of the April 2004 release, the FHFB has implemented the new metropolitan area definitions issued by the Office of Management and Budget on June 6, 2003. For 2004, the Metropolitan areas are Census 2000 Metropolitan areas. They are not identical to the areas used in 2003, and the 2004 and 2003 data are not completely compatible. Information for periods before 2004 use the previous metropolitan area definitions, which are now either CSAs or MSAs under the new definition.

The following areas (Data Buffet geo codes) were discontinued as of 2003Q4: ATL, COU, GRN, HON, IND, KAN, LOU, MIN, PHO, PIT, ROH, SAY, SAN, TAM, WAS, CMBOS, CMCHI, CMCLE, CMDAL, CMDEN, CMDET, CMHOU, CMLOS, CMMIA, CMMIL, CMNEY, CMPHI, CMPOT, CMSAF, CMSEA, CMSTL.


Organizational changes

In July 2008, the Federal Housing Finance Board (FHFB) was absorbed into the Federal Housing Finance Agency (FHFA). As of 2011, the term "Finance Board" still appears on the web site, but not in the news releases.

Prior to October 1989, MIRS was conducted by the Federal Home Loan Bank Board (FHLBB).  It was then run by the Federal Housing Finance Board (FHFB), which in 2008 was (with parts of several other agencies) reorganized as the Federal Housing Finance Agency (FHFA).

The 12 regional Federal Home Loan Banks (FHLBanks) are government-sponsored enterprises (GSEs) created in 1932 to provide low-cost funding for housing finance. They have over 8,100 member financial institutions, including commercial banks, S&Ls, insurance companies, and federally insured credit unions. The FHLBanks are overseen by the FHFB. As of July 2008, the FHFB is part of the newly-created Federal Housing Finance Agency (FHFA), an independent agency in the executive branch.