Homepage Masthead
Liberty Street Economics Blog
E-mail alerts
RSS feeds
YouTube
FOLLOW US:

 
 
Maps of Nonprime Mortgage Conditions in the United States

Technical Appendix for Maps
The dynamic maps show conditions and density of owner-occupied subprime and alt-A mortgage loans for the states, counties and zip codes in the United States.

These maps are based on data for owner-occupied mortgage loans that have been placed into a security that has been assigned a grade of subprime or alt-A.

The underlying data do not represent every subprime mortgage or alt-A mortgage loan, whether in portfolio or in a security. We estimate that as of year-end 2007, there were a total of about 7 million subprime loans. The underlying data contained 3.3 million active subprime loans, suggesting a coverage ratio of 47 percent.

We estimate that about 25 percent of securitized subprime loans do not appear in this data set. By definition, all alt-A mortgages are securitized. Our best guess is that 2.4 million loans in this portion of the data cover more than 90 percent of the pools marketed as alt-A.

The loan data are drawn from reports by the Board of Governors of the Federal Reserve System based on data from FirstAmerican CoreLogic, LoanPerformance Data. The number of housing units in each zip code is drawn from the U.S. Census 2000.

Compared with prime mortgages, subprime mortgages were typically made to borrowers with blemished credit history or who had provided only limited documentation of their income or assets. Originations of subprime mortgages fell sharply in the second half of 2007 and have been extremely light so far in 2008.

Loans marketed in the alt-A securities are typically higher-balanced loans made to borrowers who might have had past credit problems—but not severe enough to drop them into the subprime category—or who, for some reason, such as a desire not to document income, chose not to obtain a prime mortgage. Although the term “alt-A” applies technically only to securities, not mortgages, it has become common practice to refer to near-prime or nontraditional mortgages as “alt-A” loans.

Loans/1000 housing units means the number of owner-occupied loans per 1000 housing units in the area. This measures the density of loans in the area. As noted above, we estimate that these data cover 47 percent of subprime loans. It is reasonable to assume that the covered loans are representative of the entire subprime market but we can not say for sure how accurate that assumption is.

In foreclosure/1000 housing units means the number of owner-occupied loans per 1000 housing units where the lender has initiated the foreclosure process but has not completed it. The length of the foreclosure process varies by state, so two otherwise similar areas in different states could record different foreclosure densities if the foreclosure process takes longer in one state than the other. Thus, this field measures the stock of loans in foreclosure at a particular time, not the rate of completed foreclosures. At the end of 2007, there were just over 250,000 owner-occupied subprime loans in foreclosure in these data.

REO/1000 housing units means the number of real estate owned properties per 1000 housing units. REO status indicates that the lender has taken legal title to the property, through foreclosure or transference of title from the borrower. The file identified nearly 140,000 owner-occupied subprime loans in this status at the end of December of 2007.

Users interested in knowing more about REO loans can consult the Mortgage Bankers Association® list of property preservation contacts for numerous large mortgage servicers from around the country. See external links on the right-hand side of this page to access the list. While the list is not exhaustive, it provides a starting point for locating appropriate contacts who may control REO properties around the county.

Share ARMs (adjustable rate mortgages) means the share of owner-occupied loans that have a variable rate of interest that will be reset periodically in contrast to loans with interest rates fixed to maturity. ARMs are given special consideration because they traditionally have a higher likelihood of being delinquent or foreclosed upon than fixed rate loans. This is true in the prime, alt-A and subprime markets.

Share current shows the percentage of owner-occupied loans where the borrower’s payments are up to date. Loans in this category may have at some time been delinquent but as of the end of our period they are caught up. While slightly more than 60 percent of owner-occupied subprime loans were current in December 2007, only 50 percent of these loans missed a payment in 2007.

Share 90+ days overdue means that the loan payment is 90 or more days overdue but the loan is not in foreclosure or REO. The variable here is the share of owner-occupied subprime loans that are 90 or more days overdue. Loans are either current, 30 to 89 days overdue, 90 or more days overdue, in foreclosure or in REO.

Share in foreclosure means where the lender has initiated the foreclosure process but has not completed it. The length of the foreclosure process varies by state, so two otherwise equal areas in different states could record different foreclosure shares if the foreclosure process takes longer in one state than the other. Thus, this field measures the stock of loans in foreclosure at a particular time, not the rate of completed foreclosures.

Median combined LTV stands for the median Loan to Value ratio, that is, the ratio of the loan amount to the value of the property at origination. Some properties have multiple liens at origination because a second or “piggyback” loan was also executed. Our data capture only the information reported by the first lender. Only if the same lender originated and securitized a second lien is it captured in our combined LTV measure. Home equity lines of credit are not captured in our LTV ratios. The median LTV includes the second liens as noted above and is the value at which half the LTVs are higher and half the LTVs are lower.

Share low FICO and high LTV are here defined as the share of owner-occupied loans with both FICO scores below 620 and LTVs above 90 percent. Loans with these two attributes together are commonly considered among the riskiest of all loans. FICO is a credit bureau risk score. The higher the FICO score, the lower the likelihood of delinquency or default for a given loan. Also, everything else being equal, the lower the FICO score, the higher will be the cost of borrowing/interest rate.

For more information about the relationship between FICO score and mortgage interest rate see the external link on the right-hand side of this page. The FICO scores we note are the scores at the time the loans were originated. The average FICO score at origination for owner-occupied subprime loans in the data was just under 620 compared with slightly more than 700 for alt-A loans.

Share low or no documentation refers to the percentage of owner-occupied loans where the borrower provided little or no verification of income and assets in order to receive the mortgage.

Share ARMs resetting in 12 mo. means the share of adjustable rate mortgages where the rate of interest is scheduled to undergo its first rate reset within the next 12 months. Many subprime ARMs, especially so-called “2/28s” and "3/27s" carried initial rates below their fully indexed rate. The fully indexed rate on a subprime ARM is typically the six-month LIBOR (London Interbank Offer Rate) plus the margin, typically about 6 percent for subprime loans. At their first rate reset, rates on subprime ARMs can move up sharply, depending on the current level of the six-month LIBOR.

Share late payment in 12 mo. means the share of owner-occupied loans where at least one payment has been late over the past 12 months. Difficulties in paying on time often precede more serious defaults.

Notes on Underlying Data

Conversions for zip code: The loan-level information contain only zip codes as geographic identifiers. Zip code boundaries do not conform exactly with county or even state boundaries. We endeavor to be as accurate as possible in converting data from the zip code to county level, but there may be some imperfections in our conversion.