skip navigation

Sunday, April 20, 2014

[ – ] Text Size [ + ]  |  Print Page

FHA Lending Activity in the Past Decade: A National Overview

1 | 2 | 3 | 4

Variations in FHA Lending Across States

Print Version of Complete Report

FHA Lending Patterns Nationally and in the Third District States PDF (27 pages, 858 KB)

While all regions in the U.S. experienced declines in FHA volume (and in the FHA’s share of total originations) between 2000 and 2006, the extent of these declines varied greatly for individual states and regions of the country. Maps 1 and 2 provide information on the scale of FHA lending by loan count for each state and on each state’s share of total FHA volume in 2000, while Maps 3 and 4 provide this information for 2003, and Maps 5 and 6, for 2006. Maps 7 and 8 present this picture for 2009 and demonstrate that the FHA’s resurgence after 2006 was nationwide, and not just limited to specific parts of the country.

In 2000, FHA volume was, for the most part, distributed in approximate proportion to the population distribution across states. This is largely the case in 2003 as well, although it is clear that the importance of California, the nation’s largest state, as a source of FHA originations is declining. By 2006, when the FHA’s decline bottoms out, the picture is quite different. FHA loan count in the “sand states”8 has dropped to extremely low levels; this is particularly apparent in how flat Map 6 is for the western third of the country. Lending has dropped in the remainder of the states as well, although generally the drop has not been as extreme. As a result, the distribution of FHA lending across states shifts considerably, with states in the central part of the country, in particular, gaining in their share of total FHA volume. California and Texas illustrate two extremes in FHA lending trends over this period. (See California and Texas.) In 2009, well into the FHA’s resurgence, the distribution of FHA lending originations bears a strong resemblance to the picture in 2000, when that distribution in large part reflected population distribution, although loan levels are obviously considerably higher in 2009. The return of FHA lending to the sand states is particularly notable.

A comparison of the FHA mortgage market with the total mortgage market over the past decade shows some similarities but also some pronounced differences between the two. The differences occur both in the trends in volume and in the relative importance of purchase and refinancings in determining overall trends.

Trends in Volume

Volume for the mortgage market as a whole increased considerably at the start of the decade, fueled by a surge in refinance activity in response to low interest rates. Overall market volume peaked in 2003, earlier in the decade than is commonly believed. In subsequent years, both the volume of refinancings and overall volume declined, although the fall in overall volume was tempered by an increase in purchase volume mid-decade. With the collapse of subprime in 2007, and the broader mortgage market crisis, total market volume (and volume in both the refinancing and purchase segments) fell sharply; in response to extremely low interest rates, refinancing and overall volume increased somewhat in 2009, although total, refinancing, and purchase volume all remain considerably below their decade highs. As in the overall market, FHA lending increased sharply at the start of the decade as refinancing volume grew in response to low interest rates; beyond this point, however, the FHA experience and that of the total market diverged in critical ways. As in the market as a whole, total volume and refinancing volume both fell subsequent to 2003. In the case of the FHA, the fall was accompanied by a fall in purchase volume; furthermore, refinancing activity fell at a faster rate than in the overall market, with the result that the FHA’s shares of the total, purchase, and refinancing markets all fell. Moreover, as volume in the overall market fell in the wake of the subprime collapse, the FHA’s refinancing and purchase volume (and consequently total volume) grew; the increase in purchase volume was particularly sharp. Unlike the case in the overall market, the FHA finished the decade with higher total, purchase, and refinancing volume than at the decade’s start, and its market share was considerably higher than it had been at any point prior to the subprime crisis.

Relative Importance of Refinancings and Purchase Activity in Determining Overall Trends

Throughout the past decade, refinance originations have been the dominant factor in the total market, both in terms of share of overall volume and in determining the overall direction in volume. For example, in 2001 and 2003 this trend is evident where the FHA purchase volume declined while overall volume increased. By contrast, in the case of the FHA, purchase originations have been the dominant factor in every year except for 2003. This finding is congruent with the general perception of the FHA program as primarily a purchase program.

The large differences in volume trends between the overall market and the FHA after 2003 is closely related to the waxing and waning of subprime lending between 2003 and 2007. Evidence strongly suggests that the FHA lost potential borrowers to the subprime sector (whose activity is captured in overall market volume) during this period. (See, for example, Courchane, et al.) A variety of FHA requirements likely contributed to making subprime loans appear more attractive to potential borrowers than FHA loans. For example, the process of getting FHA status for a loan was extremely cumbersome and could work to a purchaser’s disadvantage in hot housing markets.9 Moreover, loan limits in the FHA program stipulated a maximum mortgage amount for which the FHA program could be used. While these limits varied across geographies, they did not keep up with house price appreciation in some areas and relatively low loan limits made the FHA a poor vehicle for mortgage financing in coastal states such as California during the years of rising home prices. With the collapse of subprime lending and the subsequent tightening of underwriting standards in the mortgage sector as a whole, FHA lending rebounded. Higher loan limits authorized by Congress and a streamlined process for gaining FHA loan status have likely contributed to the rebound.

The increase in FHA volume appears to come from two sources. First, some purchasers who might have gotten subprime loans mid-decade are now getting loans from the FHA.10 However, the increase in FHA volume cannot be attributed solely to a return of borrowers who might have found subprime a more attractive option mid-decade. The credit score distribution of recent FHA borrowers11 suggests that, with the tightening of the credit market, some borrowers who might in the past have gotten prime conventional loans may not be able to do so and are turning to the FHA as an alternative; that is, much of the FHA’s increased market share is crisis driven.

The national patterns explained here regarding total originations and FHA originations provide a useful illustration of the mortgage market in the nation as a whole. To the extent that the FHA is acting as a stabilizing factor to the market, and its increased use in recent years is primarily crisis-related, then as the housing market improves, the FHA’s market share and volume are likely to decline. It is possible, however, that some of the buyers who in a stronger market might be expected to get prime conventional mortgages will continue to find the FHA to be a better alternative. Factors such as the availability of private mortgage insurance and the strictness of post-crisis underwriting standards are likely to affect the decision-making calculus.“Brand loyalty” may also be a factor: Following the economic crisis in oil patch states in the early 1980s, the FHA program’s market share in Texas remained high compared to the rest of the nation. Given such factors, the future scale of FHA lending will be interesting to follow.

A Brief Note on the Lending Environment

During the 2000-2009 period, overall mortgage lending and mortgage loan performance were, unsurprisingly, influenced by a range of factors external to the FHA itself. We note three factors of particular importance in understanding FHA lending patterns and how these patterns differed from those of the overall mortgage market over the period in question.12

Interest rates. As shown in Chart 2, interest rates have been relatively low during the decade in contrast to previous periods. Rates were particularly low in 2003 as a result of Federal Reserve actions in response to the recession in the early years of the decade, and since mid 2008, when the Fed took action in response to the poor economic conditions of the last few years.

Home prices. As shown by the Case-Schiller home price index (Chart 3), house prices increased sharply for much of the period and then, beginning in the third quarter of 2006, declined sharply, an indication of the onset of the housing crisis.

Subprime lending.13 The pattern of subprime lending during the decade provides a particularly important context for understanding trends in FHA lending. Subprime lending increased sharply in 2003 and, by 2006, had surpassed $600 billion.14 It accounted for one-fifth of mortgage originations in 2006, before it collapsed in 2007.15 In this period, mortgage credit became more readily available to borrowers with low credit scores, and there was, more generally, a relaxing of overall credit standards.16 Because of relaxed underwriting, as well as other factors discussed later in this report, subprime loans may have appeared to be a more attractive option than FHA loans in the middle of the past decade. In contrast, the collapse of subprime lending — and the broader housing market problems that followed — set the stage for the FHA’s resurgence at the end of the decade.

California and Texas

California and Texas, which together had a share of FHA originations ranging from about 13 percent to more than 20 percent between 2000 and 2009,17 illustrate two extremes in FHA lending trends over the course of that decade.

California had the largest percentage decrease in FHA lending and began to experience that decrease earlier than other states (Chart 10). According to HMDA figures, in 2000, California accounted for 11.8 percent of all mortgage originations but nearly 13 percent of FHA loans. This 13 percent share was the highest in the nation. By 2006, California accounted for more than 14 percent of all mortgage originations but only 1.23 percent of all FHA originations. One factor contributing to the FHA’s decline in California over this period may have been the low level of FHA loan ceilings compared with house prices in many parts of the state. In the aftermath of the subprime collapse, California’s share of total FHA originations rebounded to about 10 percent by 2009; in this year, California’s FHA lending originations had increased by a factor of about 37 from their 2006 low. A 2008 increase in FHA loan ceilings, from less than $400,000 in 2006 and 2007 to its current level of about $729,000, has likely contributed to the size of the rebound.

Compared with the situation in California, FHA lending originations were relatively stable in Texas, where the decline in FHA originations was much lower than was typical elsewhere. In 2000, Texas accounted for 6 percent of all originations and 8.5 percent of total FHA originations. As in the nation as a whole, FHA originations in Texas peaked in 2003 and then fell to a low point in 2006. However, FHA lending originations in Texas that year were still above 47,000 in total count, while FHA originations in California fell below 5,000; that is, there were almost 10 times as many FHA originations in 2006 in Texas as in California. In that year, Texas had a 12 percent share of all FHA originations, the largest in the nation. (Unlike in California, house prices in Texas tend to be low compared with the rest of the country. In addition, Texas had little house price appreciation in the first part of the past decade. In turn, to the extent that FHA loan limits were a factor in the FHA’s decline during that period, one would expect less impact on FHA volume in Texas than in California.) With the resurgence of the FHA at the end of the decade, Texas’s share of FHA originations fell. It stood at 7 percent in 2009, close to its 2000 share.


Avery, Robert, Kenneth Brevoort, and Glenn Canner. “The 2007 HMDA Data,” Federal Reserve Bulletin (December 2008).

Courchane, Marsha, Rajeev Darolia, and Peter Zorn. “Industry Changes in the Market for Mortgage Loans External Link,” Connecticut Law Review, 41:4 (May 2009), pp. 1143-76.

Federal Financial Institutions Examination Council. “Who Reports HMDA Data?” FFIEC (2010); available online at External Link.

Federal Reserve Bank of Dallas. “The CRA and Subprime Lending: Discerning the Difference External Link,” Banking and Community Perspectives, 1 (2009).

Mayer, Chris, and Karen Pence. “Subprime Mortgages: What, Where, and to Whom?” Finance and Economics Discussion Series, Washington: Board of Governors of the Federal Reserve System, Division of Research and Statistics and Division of Monetary Affairs (2008-29); available at External Link PDF Document.

Newburger, Harriet B. “FHA Lending: Recent Trends and Their Implications for the Future,” Federal Reserve Bank of Philadelphia Community Development Discussion Paper (forthcoming).

U.S. Department of Housing and Urban Development. Federal Housing Administration (2010); available at External Link.

1 | 2 | 3 | 4

  • 8 The “sand states” are California, Arizona, Nevada, and Florida.
  • 9 This process was streamlined in 2005.
  • 10 It should be noted, however, that because FHA underwriting criteria are stricter than those of the subprime sector, not all borrowers who might have gotten subprime loans in the past can now get them from the FHA. Current lender underwriting policies may also limit the extent to which borrowers who might have qualified for a subprime loan in the past can now get an FHA loan; for example, evidence suggests that in 2009, lenders made few loans to potential buyers with FICO scores below 620. For a more complete discussion, see Harriet B. Newburger, “FHA Lending: Recent Trends and Their Implications for the Future,â€� Federal Reserve Bank of Philadelphia Community Development Discussion Paper (forthcoming).
  • 11 More information on this trend is available in a subsequent report on FHA delinquency trends.
  • 12 While discussed separately, patterns that appeared in these factors over the decade were related to some degree.
  • 13 Although there is more than one definition of subprime lending, in this report, subprime refers to higher priced home mortgage credit as reported in the HMDA loan application register (LAR). In 2004, information was added to the HMDA LAR requiring that mortgage loan rate spreads be reported when the rate spread on the loan is above the threshold when compared with a comparable-maturity Treasury rate for first-lien mortgages with an annual percentage rate (APR) 3 percentage points over the Treasury benchmark and for junior liens with an APR 5 percentage points over the benchmark. These mortgages with a reported spread are commonly called “higher-priced,” or subprime, loans. For a more complete discussion, see the study by Chris Mayer and Karen Pence (2008). These reporting rules were changed in 2009 with amendments to the Home Mortgage Disclosure Act (HMDA), Regulation C.
  • 14 See the report by the Federal Reserve Bank of Dallas External Link (2009).
  • 15 See Federal Reserve Bank of Dallas (2009). Another source notes that the subprime market virtually disappeared in 2007, a year in which 169 lenders, primarily subprime nondepository lenders, ceased reporting HMDA data. See the article by Robert Avery, Kenneth Brevoort, and Glenn Canner (December 2008) for a more complete discussion of this topic.
  • 16 For example, the Federal Reserve Board’s Senior Loan Officer Opinion Survey External Link provides evidence that lending standards became less stringent during the early years of the housing boom. More information on this trend is provided in a subsequent report on delinquency trends. Additionally, this trend is evident in Chart 11.
  • 17 Overall, according to HMDA data, California accounts for 9.31 percent of total FHA originations from 2000-2009, and Texas accounts for 8.1 percent of the FHA total for this period
  • Last updated: Tuesday, February 15, 2011