Mortgage Financing: FHA's Fund Has Grown, but Options for Drawing on the
Fund Have Uncertain Outcomes (Letter Report, 02/28/2001, GAO/GAO-01-460).
The Mutual Mortgage Insurance Fund has maintained an economic value of
at least 2 percent of the Fund's insurance-in-force, as required by law.
GAO's and the Department of Housing and Urban Development's (HUD)
analysis show that the Fund had an economic value of $15.8 billion (3.20
percent) and $16.6 billion (3.66 percent), respectively. Given the
economic value of the Fund and the state of the economy at the end of
fiscal year 1999, a 2-percent capital ratio appears sufficient to
withstand moderately severe economic downturns that could lead to
worse-than-expected loan performance. However, under more severe
economic conditions, the economic value of 2 percent of
insurance-in-force would not be adequate. Because of the uncertainty and
professional judgment associated with this type of economic analysis,
GAO cautions against relying on one estimate or even a group of
estimates to determine the adequacy of the Fund's reserves over the
longer term. HUD could exercise several options under current
legislative authority to reduce the capital ratio for the Fund. It is
difficult, however, to reliably measure the impact of policy changes on
the Fund's capital ratio and Federal Housing Administration borrowers
without using tools designed to estimate the multiple impacts that
policy changes often have. Nonetheless, any option that results in a
reduction in the Fund's reserve, if not accompanied by a similar
reduction in other government spending, would result in a budget surplus
reduction or a deficit increase.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GAO-01-460
TITLE: Mortgage Financing: FHA's Fund Has Grown, but Options for
Drawing on the Fund Have Uncertain Outcomes
DATE: 02/28/2001
SUBJECT: Funds management
Mortgage loans
Econometric modeling
Economic analysis
Mortgage programs
Mortgage protection insurance
Risk management
IDENTIFIER: Mutual Mortgage Insurance Fund
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO Testimony. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
******************************************************************
GAO-01-460
A
Report to Congressional Committees
February 2001 MORTGAGE FINANCING
FHA's Fund Has Grown, but Options for Drawing on the Fund Have Uncertain
Outcomes
GAO- 01- 460
Letter 3 Appendixes Appendix I: Scope and Methodology 34
Appendix II: Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund 38 Appendix III: Development of Scenarios of Adverse
Economic Conditions Used to Estimate the Economic Value
of the Fund 69 Appendix IV: Comments From the Department of Housing and
Urban Development 76 Appendix V: GAO Contacts and Staff Acknowledgments 80
Related GAO Products 81 Tables Table 1: Estimates of Capital Ratios for
FHA's Mutual Mortgage
Insurance Fund by GAO and Deloitte & Touche, End of FY 1999 13 Table 2:
Capital Ratios Under Expected and Historical Economic Scenarios 17 Table 3:
Capital Ratios Under Expected and More Severe Economic
Scenarios in Selected Locations 18 Table 4: Variable Names and Descriptions
48 Table 5: Means of Predictor Variables 50 Table 6: Coefficients From
Foreclosure Equations and Summary
Statistics 54 Table 7: Coefficients From Prepayment Equations and Summary
Statistics 56 Table 8: Premium Schedule for 30- Year Non- Streamline
Mortgages,
by Date of Mortgage Origination 63 Table 9: Premium Schedule for 15- Year
Non- Streamline Mortgages,
by Date of Mortgage Origination 64 Table 10: Premium Schedule for 15- and
30- Year Streamline Refinanced
Mortgages That Originated Before July 1, 1991 64 Table 11: Premium Refund
Rates for Loans That Were Terminated After October 1, 1983 66
Table 12: Alternative Estimates of Capital Ratios for FHA's Mutual Mortgage
Insurance Fund 68
Table 13: Historical Median House Price and Unemployment Experience in
States Representing Regional Downturns 71 Table 14: National Results of
Alternative Scenarios 74
Figures Figure 1: Cash Flows of the Mutual Mortgage Insurance Fund 8 Figure
2: Comparison of Estimated Economic Value and 2 Percent of Insurance- in-
Force, 1989- 2000 10
Figure 3: Capital Ratios Resulting From Applying the Average 1986- 90
Foreclosure Rates in the West South Central Census Division to Varying
Proportions of FHA's Insurance Portfolio in 2000- 04 21 Figure 4: Cumulative
Foreclosure Rates by Book of Business for
30- Year, Fixed- Rate, Nonrefinanced, Mortgages; Actual and Predicted,
Fiscal Years 1975- 99 59 Figure 5: Cumulative Prepayment Rates by Book of
Business, for
30- Year, Fixed- Rate, Nonrefinanced, Mortgages: Actual and Predicted,
Fiscal Years 1975- 99 60
Abbreviations
ARM Adjustable rate mortgage FHA Federal Housing Administration HUD
Department of Housing and Urban Development LTV Loan- to- value
Lett er
February 28, 2001 The Honorable Wayne Allard Chairman, Subcommittee on
Housing and Transportation Committee on Banking, Housing and Urban Affairs
United States Senate
The Honorable Marge Roukema Chairman, Subcommittee on Housing and Community
Opportunity Committee on Financial Services House of Representatives
According to the Department of Housing and Urban Development (HUD), the
economic value of the Mutual Mortgage Insurance Fund (Fund) grew by about $5
billion in 1999, apparently reaching its highest level in at least the last
20 years. 1 Under this Fund, HUD's Federal Housing Administration
(FHA) provides insurance for private lenders against losses on home
mortgages. Borrowers who obtain FHA- insured mortgages pay insurance
premiums, which are deposited into the Fund. Although the Fund has been
financially self- sufficient for most of its history, it experienced
substantial losses during the 1980s, primarily because foreclosure rates on
singlefamily homes supported by the Fund were high in economically stressed
regions. To help place the Fund on a financially sound basis, the Congress
enacted legislation in November 1990 that required the Secretary of HUD
to, among other things, take steps to ensure that the Fund achieves and
maintains an economic value of at least 2 percent of the Fund's insurancein-
force. The 1990 reforms also required that an independent actuarial study be
conducted annually to measure this capital ratio. In January 2000, HUD
reported that, according to an independent study by Deloitte & Touche, the
estimated capital ratio was 3.66 percent at the end of fiscal
1 The economic value of the Fund is the sum of existing capital resources
plus the net present value of future cash flows.
year 1999. 2 With the Fund's capital ratio now substantially above the
minimum required level, proposals have surfaced to reduce the ratio either
by spending some of the Fund's current resources or by reducing the net cash
flows into the Fund.
Concerned about the adequacy of the minimum 2- percent requirement and about
proposals to spend what some were calling excess reserves, you asked us to
determine the conditions under which an estimated capital ratio of 2 percent
would be adequate to maintain the actuarial soundness of the Fund.
Specifically, you asked us to (1) estimate the value of the Fund at the end
of fiscal year 1999, given expected economic conditions, and
compare our estimate to the estimate of the value of the Fund reported by
HUD for that year; (2) determine the extent to which a 2- percent capital
ratio would allow the Fund to withstand worse- than- expected loan
performance due to economic and other factors; and (3) describe some options
for adjusting the size of the Fund if the estimated capital ratio is
different than the amount needed, and describe the impact that these options
might have on the Fund, FHA mortgagors, and the federal budget.
To meet these objectives, we reviewed the laws and regulations governing
FHA's insurance program. In addition, to estimate the value of the Fund and
determine the extent to which a 2- percent reserve would allow the Fund to
withstand worse- than- expected loan performance, we developed economic
models. We also met with HUD officials who administer FHA's single- family
insurance program, the independent contractors that have analyzed the Fund
throughout the 1990s, and other experts to better understand methodologies
used to estimate the value of the Fund. In addition, we met with officials
at the Office of Management and Budget and
the Congressional Budget Office who are building economic models of FHA's
insurance program to provide better information about the Fund's impact on
the federal budget. To explore options for changing the size of the Fund, we
met with HUD officials and other interested parties. To determine the impact
of these changes on the Federal budget, we relied on our own experts as well
as budget experts familiar with FHA's program and mortgage models at the
Office of Management and Budget and the Congressional Budget Office. 2
Actuarial Review of MMI Fund as of FY 1999, Deloitte & Touche. In its
January 2001 actuarial review, Deloitte & Touche estimated a capital ratio
of 3. 51 percent at the end of fiscal year 2000. This estimate should not be
compared with its reported estimate for 1999 because, in its latest report,
Deloitte revised downward its estimate of the value of the Fund
at the beginning of fiscal year 2000.
We conducted our work between December 1999 and February 2001 in accordance
with generally accepted government auditing standards. See appendix I for a
further discussion of our scope and methodology. Results in Brief We
estimated that the Mutual Mortgage Insurance Fund had an economic
value of about $15.8 billion at the end of fiscal year 1999. This estimate
implies a capital ratio of 3.20 percent of the unamortized insurance-
inforce. The relatively large economic value and high capital ratio reflect
the strong economic conditions that prevailed during most of the 1990s and
the good economic performance expected for the future as well as the
increased insurance premiums put in place in 1990. HUD reported that
Deloitte & Touche, using a different method of analysis, estimated an
economic value of about $16.6 billion for fiscal year 1999, which on the
basis of its estimate of the unamortized insurance- in- force implied a
capital ratio of about 3.66 percent. Although there is uncertainty
associated with any forecast, both of these estimates easily exceed the
minimum required capital ratio of 2 percent that the Congress set in 1990.
The difference
between these estimates is due in part to differences in the timing of the
analyses and the assumptions used. Nonetheless, given the uncertainty
inherent in forecasting and the number of professional judgments made in
this type of analysis, we conclude that these estimates are comparable.
Given the economic value of the Fund and the state of the economy at the end
of fiscal year 1999, a 2- percent capital ratio appears sufficient to
withstand moderately severe economic downturns that could lead to worse-
than- expected loan performance. That is, such conditions would not cause
the estimated value of the Fund at the end of fiscal year 1999 to decline by
more than 2 percent of the Fund's insurance- in- force. Under economic
scenarios that we developed to represent regional and national economic
downturns that the nation experienced between 1975 and 1999,
the estimated capital ratio fell by only slightly less than 0.4 percentage
points. Some more severe downturns that we analyzed also did not cause the
estimated capital ratio to decline by as much as 2 percentage points.
However, in three of the scenarios with more severe economic conditions, an
economic value of 2 percent of insurance- in- force would not be adequate.
These included (1) a scenario in which the entire nation experiences a
downturn similar to the one New England experienced during the late 1980s
and early 1990s, (2) a scenario in which FHA experiences foreclosure rates
similar to those it experienced in the late
1980s, and (3) a scenario in which 35.6 percent or more of FHA loans
experience foreclosure rates similar to those experienced by FHA in the
west south central portion of the United States in the late 1980s. Because
of the uncertainty and professional judgment associated with this type of
analysis, we caution against relying on any one estimate or even on a group
of estimates at a point in time to determine the adequacy of the Fund's
reserves over the longer term. For example, recent and future FHA- insured
loans might perform worse than our estimates assume for several reasons,
including recent and future changes in FHA's programs and the conventional
mortgage market. To the extent that this is the case, the Fund could be less
able to withstand adverse economic scenarios than some of
our estimates suggest. In fact, the Fund had an even higher capital ratio in
1979 when the economic value of the Fund equaled 5.3 percent of insurance-
in- force, but in little more than a decade- after a national recession, the
substitution of an up- front premium for annual insurance premiums, and
regional real estate declines- the economic value of the Fund was negative.
This report contains a recommendation that the Secretary of HUD address
these limitations in evaluating the health of the Fund.
There are several options available to the Secretary of HUD under current
legislative authority that could result in a lower capital ratio for the
Fund. Other options would require legislative action. However, in either
case, it is difficult to reliably measure the full impact of policy changes
on the Fund's capital ratio and FHA borrowers without using tools designed
to estimate the multiple impacts that policy changes often have. The extent
to which policy changes will affect the Fund's capital ratio and FHA
borrowers is difficult to estimate because the impact often depends not only
on the direct effect of the changes but also on the degree to which the
changes affect the volume of FHA- insured loans and the riskiness of those
loans. A
complete estimate would require that economic models be used to estimate
these indirect effects. At this time, however, neither the models used by
HUD to assess the financial health of the Fund, nor those used by others,
explicitly recognize the indirect effects of policy changes on the volume
and riskiness of FHA's loans. As a result, HUD cannot reliably estimate nor
evaluate the full impact of policy changes on the Fund. Further, the
difficulty of estimating the impact of the various policy options on the
capital ratio also makes it difficult to measure precisely the impact on the
federal budget. Nonetheless, any option that results in a reduction in the
Fund's reserve, if not accompanied by a similar reduction in other
government spending or by an increase in receipts, would result in either a
reduction in the budget surplus or an increase in any existing deficit. This
report contains a recommendation that the Secretary of HUD develop
better tools for assessing the impact of policy changes on the Fund.
Background FHA was established in 1934 under the National Housing Act (P. L.
73- 479) to broaden homeownership, shore up and protect lending
institutions, and stimulate employment in the building industry. FHA insures
private lenders against losses on mortgages that finance purchases of
properties with one to four housing units. Many FHA- insured loans are made
to low- income, minority, and first- time homebuyers.
Generally, borrowers are required to purchase single- family mortgage
insurance when the value of the mortgage is large relative to the price of
the house. FHA, the Department of Veterans Affairs, and private mortgage
insurers provide virtually all of this insurance. In recent years private
mortgage insurers and conventional mortgage lenders have begun to offer
alternatives to borrowers who want to make little or no down payment. 3 FHA
provides most of its single- family insurance through a program supported by
the Mutual Mortgage Insurance Fund. The Fund is organized as a mutual
insurance fund in that any income received in excess of the
amounts required to cover initial insuring costs, operating expenses, and
losses due to claims may be paid to borrowers in the form of distributive
shares after they pay their mortgages in full or voluntarily terminate their
FHA insurance. The economic value of the Fund depends on the relative
sizes of cash outflows and inflows over time. Cash flows out of the Fund
from payments associated with claims on foreclosed properties, refunds of
up- front premiums on mortgages that are prepaid, and administrative
expenses for management of the program (see fig. 1). To cover these
outflows, FHA deposits cash inflows- up- front and annual insurance premiums
from participating homebuyers and net proceeds from the sale of foreclosed
properties- into the Fund. If the Fund were to be exhausted,
the U. S. Treasury would have to cover lenders' claims and administrative
costs directly.
3 Generally, borrowers are required to purchase mortgage insurance when the
value of the mortgage is over 80 percent of the price of the house. Many
private mortgage insurers will now insure a mortgage up to 97 percent of the
value of the house being purchased. In addition, conventional mortgage
lenders by offering second mortgages of up to 23 percent of the value of the
house, sometimes allow borrowers to borrow more than the value of the house
without obtaining mortgage insurance.
Figure 1: Cash Flows of the Mutual Mortgage Insurance Fund
The Fund remained relatively healthy from its inception until the 1980s when
losses were substantial, primarily because of high foreclosure rates in
regions experiencing economic stress, particularly the oil- producing states
in the west south central section of the United States. These losses
prompted the reforms that were first enacted in November 1990 as part of
the Omnibus Budget Reconciliation Act of 1990 (P. L. 101- 508). The reforms
that were designed to place the Fund on an actuarially sound basis required
? the Secretary of HUD to take steps to ensure that the Fund attains a
capital ratio of 2 percent of the insurance- in- force 4 by November 2000
and maintains that ratio at a minimum at all times thereafter;
? an independent contractor to conduct an annual actuarial review of the
Fund; 5
? the Secretary of HUD to suspend the payment of distributive shares, which
had been paid continuously from 1943 to 1990, until the Fund is actuarially
sound; and
? FHA borrowers to pay more in insurance premiums over the life of their
loans by adding a risk- adjusted annual premium to the one time, upfront
premium. The Federal Credit Reform Act of 1990, enacted as part of the
Omnibus
Budget Reconciliation Act of 1990, also reformed budgeting methods for
federal credit programs including FHA's mutual insurance program. The 1990
credit reforms were intended to ensure that the full cost of credit
activities for the current budget year would be reflected in the federal
budget so that the executive branch and the Congress could consider these
costs when making annual budget decisions. As a result, FHA's budget is
required to reflect the subsidy cost to the government- the estimated
longterm cost calculated on a net present value basis- of FHA's loan
insurance activities for that year. During the 1990s, the estimated economic
value of the Fund- comprised of capital resources and the net present value
of future cash flows- grew substantially. As figure 2 shows, by the end of
fiscal year 1995, the Fund had attained an estimated economic value that
slightly exceeded the amount required for a 2- percent capital ratio. Since
that time, the estimated economic value of the Fund has continued to grow
and has always exceeded the amount required for a 2- percent capital ratio.
4 The Omnibus Budget Reconciliation Act of 1990 defined the capital ratio as
the ratio of the Fund's capital, or economic net worth (economic value), to
its unamortized insurance- inforce. However, the act defined unamortized
insurance- in- force as the remaining obligation on outstanding mortgages- a
definition generally understood to apply to amortized insurance- in- force.
HUD has calculated the 2- percent capital ratio using unamortized insurance-
in- force as it is generally understood- which is the initial amount of
mortgages. Unless otherwise noted, the capital ratios in this report are
calculated using unamortized
insurance- in- force. 5 From 1989 to 1998, Price Waterhouse performed this
review; in 1999, Deloitte & Touche was awarded the contract.
Figure 2: Comparison of Estimated Economic Value and 2 Percent of Insurance-
in- Force, 1989- 2000
20 Dollars in billions
15 10
5 0 -5
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Fiscal years
Estimated value 2% of insurance- in- force
Source: GAO analysis of Price Waterhouse (now PricewaterhouseCoopers) and
Deloitte & Touche data.
As a result of the 1990 housing reforms, the Fund must not only meet capital
ratio requirements, but it must also achieve actuarial soundness; that is,
the Fund must contain sufficient reserves and funding to cover estimated
future losses resulting from the payment of claims on foreclosed mortgages
and administrative costs. However, neither the legislation nor the actuarial
profession defines actuarial soundness. Price Waterhouse (now
PricewaterhouseCoopers) in 1989 concluded that for the Fund to be
actuarially sound, it should have capital resources that could withstand
losses from reasonably adverse, but not catastrophic, economic downturns.
The Price Waterhouse report did not clearly distinguish adverse from
catastrophic downturns; however, they said that private mortgage insurers
are required to hold contingency reserves to protect
against catastrophic losses. In turn, rating agencies require that private
mortgage insurers have enough capital on hand to withstand severe losses
that would occur if loans they insure across the entire nation had losses
similar to those experienced in the west south central states in the 1980s.
Because economic downturns put downward pressure on house prices and
incomes, they can stress FHA's ability to meet its obligations. Thus, it is
reasonable that measures of the financial soundness of the Fund would be
based on tests of the Fund's ability to withstand recent recessions or
regional economic downturns. In the last 25 years, we have experienced a
national recession and regional economic declines that did or could have
placed stress on FHA. For example, the nation experienced a recession in
1981 and 1982 that strained mortgage markets. Regionally, states in the west
south central portion of the nation experienced an economic decline in 1986
through 1989 precipitated by a sharp drop in the price of crude oil.
Similarly, the economic decline experienced by California from 1992 through
1995 placed stress on FHA. Because FHA does substantial business in these
regions of the country, these experiences led to substantial losses for FHA.
In contrast, the economic decline experienced
by the New England states from 1989 through 1991 placed little strain on FHA
because insured mortgages in this region do not make up a large portion of
FHA's total portfolio. However, experiences similar to the New England
downturn, during which the unemployment rate increased by
almost 140 percent and house prices decreased by 5. 5 percent, could place
stress on FHA if they occurred in other regions or the nation as a whole.
The Fund's Estimated On the basis of our economic model of FHA's home loan
program and
Economic Value forecasts of several key economic factors, we estimate that
at the end of fiscal year 1999, the Fund had an economic value of about $15.
8 billion. Exceeds 3 Percent of
This value, which is 3.20 percent of the unamortized insurance- in- force,
Insurance- in- Force
reflects the robust economy and relatively high premium rates prevailing
through most of the 1990s and the good economic performance forecast for the
future. In comparison, Deloitte & Touche estimated that the Fund's 1999
economic value was over $800 million larger than our estimate- or about 3.
66 percent of its estimate of FHA's unamortized insurance- in- force.
Although we did not evaluate the quality of Deloitte's estimates, we believe
that Deloitte's and our estimates are comparable because of the uncertainty
inherent in forecasting and the professional judgments made in this type of
analysis. However, Deloitte's analysis and ours differ in several ways,
including the time when the analyses were performed and some of the
assumptions made.
The Estimated Economic Using conservative assumptions, we estimate that at
the end of fiscal year
Value of the Fund Reflects 1999, the Fund had an economic value of about
$15.8 billion. The economic
the Robust Economy and value of the Fund consists of the capital resources
on hand and the net
Increased Premium Rates present value of future cash flows. Documents used
to prepare FHA's 1999 financial statements show that the Fund had capital
resources of about $14.3 billion at the end of that fiscal year. We
estimated the relationship
between historical FHA foreclosures and prepayments and certain key economic
factors to forecast foreclosures and prepayments and the resulting cash
flows over the next 30- year period for mortgages insured by FHA before the
end of fiscal 1999. As a result of this analysis, we estimate that at the
end of 1999 the net present value of future cash flows was about $1. 5
billion. Summing the capital resources and future cash flows gives us
an economic value of about $15.8 billion. See appendix II for a detailed
discussion of the forecasting and cash flow models used to estimate the
economic value of the Fund.
We also estimate that the Fund's capital ratio- the Fund's economic value
divided by its insurance- in- force- exceeded 3- percent at the end of
fiscal year 1999. From the individual loan data provided by HUD, we
calculated that the unamortized insurance- in- force at the end of fiscal
year 1999 was about $494 billion and that the amortized value of that
insurance, an
estimate of the outstanding balance of the loans and thus FHA's insurance
liability, was about $455.8 billion. Therefore, the economic value of the
Fund represented 3.20 percent of the unamortized insurance- in- force and
about 3. 47 percent of the amortized insurance- in- force on September 30,
1999.
The robust economy and the increased premium rates established by the 1990
legislation contributed to the strength of the Fund at the end of fiscal
year 1999. The Fund's economic value principally reflects the large amount
of capital resources that the Fund has accrued. Because current capital
resources are the result of previous cash flows, the robustness of the
economy and the higher premium rates throughout most of the 1990s accounted
for the accumulation of these substantial capital resources. Good economic
times that are accompanied by relatively low interest rates and relatively
high levels of employment are usually associated with high levels of
mortgage activity and relatively low levels of foreclosure; therefore, cash
inflows have been high relative to outflows during this period. The
estimated value of future cash flows also contributed to the strength of the
Fund at the end of fiscal 1999. As a result of relatively low interest rates
and the robust economy, FHA insured a relatively large number of mortgages
in fiscal years 1998 and 1999. These loans make up a large portion of FHA's
insurance- in- force, because many borrowers refinanced their FHA- insured
mortgages originated in earlier years, probably as a result of interest
rates having fallen to relatively low levels in 1998 and 1999. Because these
recent loans have low interest rates and because forecasts of economic
variables for the near future show house prices rising while unemployment
and interest rates remain fairly stable, our models predict that these new
loans will have low levels of foreclosure and prepayment. As a result, our
models predict that future cash flows out of the Fund will be relatively
small. At the same time, we assume that FHAinsured homebuyers will continue
to pay the annual premiums that were reinstituted in 1991. Thus, our models
predict that cash flowing into the Fund from mortgages already in FHA's
portfolio at the end of fiscal year
1999 will be more than sufficient to cover the cash outflows associated with
these loans. As a result, the estimated economic value of the Fund is even
higher than the level of its current capital resources.
Deloitte's Estimates Are As table 1 shows, Deloitte's independent actuarial
analysis of the Fund for
Comparable to Ours, but the fiscal year 1999 estimated a capital ratio that
was somewhat higher than Analyses Differ in Several ours, 3. 66 percent
rather than 3.20 percent of unamortized insurance- inforce.
Ways Although we did not evaluate the quality of Deloitte's estimates, we
did identify some reasons that its estimate of the capital ratio was higher
than ours. The ratio is higher because Deloitte estimates both a higher
economic value of the Fund and a lower amount of insurance- in- force.
Deloitte's higher estimated economic value of the Fund includes a higher
estimated value for capital resources on hand that is somewhat offset by a
lower estimate of the net present value of future cash flows.
Table 1: Estimates of Capital Ratios for FHA's Mutual Mortgage Insurance
Fund by GAO and Deloitte & Touche, End of FY 1999
Dollars in millions
Total capital Future
Economic Unamortized
Capital ratio Estimate resources cash flows value insurance- in- force
(percent)
GAO $14,326 $1,484 $15, 810 $493,990 3. 20 Deloitte 15,331 1, 306 16, 637
454,184 3. 66
Source: GAO analysis and Actuarial Review of MMI Fund as of FY 1999,
Deloitte & Touche.
Our estimate and that of Deloitte rely on forecasts of foreclosures and
prepayments over the next 30 years, and, in turn, these forecasts
necessarily rest on forecasts of certain economic factors. In addition, the
estimates depend on the choices made concerning a variety of other
assumptions. As a result of the inherent uncertainty and the need for
professional judgment in this type of analysis, we believe that our
estimates and Deloitte's estimates of the Fund's economic value and capital
ratio are comparable.
Although the estimates are comparable, Deloitte's estimates of capital
resources and insurance- in- force differ from ours primarily because the
analyses were conducted at different times. Because Deloitte performed its
analysis before the end of 1999, it had to estimate some data for which we
had year- end values. In particular, Deloitte overestimated the 1999 value
of capital resources by extrapolating from the 1998 value. In contrast, we
used
values developed for FHA's 1999 financial statements that were about $1
billion lower than Deloitte's estimate. Using our value for capital
resources, Deloitte's estimated capital ratio would be 3.44 percent rather
than 3.66 percent of insurance- in- force. 6 Similarly, Deloitte
underestimated the number of loans that FHA insured in the fourth quarter of
fiscal year 1999
and, thus, underestimated the value of loans insured for all of fiscal year
1999 by about $33 billion, though this appears to have had little effect on
the estimated capital ratio. Our analysis of the net present value of future
cash flows and that of Deloitte also differ in several respects. Both our
estimates and Deloitte's rely on forecasts of future foreclosures and
prepayments. In turn, these forecasts are generated from models that are
based on estimated relationships between the probability of loan foreclosure
and prepayment and key explanatory factors, such as borrowers' home equity
and interest and unemployment rates. Our model differs from Deloitte's in
the way that it specifies these relationships. For example, Deloitte
specified changes in household income as one of the key explanatory factors,
while we did not. 7 6 In its 2000 actuarial review, Deloitte recognized that
its 1999 estimate of capital resources
was about $1 billion higher than the actual year- end value. In addition,
Deloitte revised upward its measure of up- front mortgage insurance premiums
and downward the net present value of future cash flows calculated at the
beginning of fiscal year 2000. As a result, Deloitte estimates an economic
value of the Fund of $14.1 billion at the beginning of fiscal year 2000,
which would have resulted in a capital ratio even lower than 3.44 percent at
the beginning of fiscal year 2000, that is, the end of fiscal year 1999. 7
We did not include changes in household income because we believe that
unemployment rates are more directly connected to changes in the ability of
borrowers to make mortgage payments and are sufficient to capture those
changes.
The analyses also differ in the assumptions made about some future economic
values and costs associated with FHA's insurance program. For example, we
assumed lower house price appreciation rates and higher discount rates for
calculating net present values than did Deloitte. 8 In addition, the
analyses differ in the way that they use HUD's data. We used a sample of
individual loans while Deloitte grouped loans into categories to do its
analysis. Although these factors could be important in identifying why the
two estimates differ, we could not quantify their impact because
we did not have access to Deloitte's models. A 2- percent Capital
According to our estimates, worse- than- expected loan performance that
Ratio Appears could be brought on by moderately severe economic conditions
would not cause the estimated value of the fund at the end of fiscal year
1999 to
Sufficient to Withstand decline by more than 2 percent of insurance- in-
force. However, a few more Some Worse- ThanExpected
severe economic scenarios that we examined could result in such poor Loan
loan performance that the estimated value of the fund at the end of fiscal
year 1999 could decline by more than 2 percent of insurance- in- force. Two
Performance
of the three scenarios that showed such a large decline extended adverse
conditions more widely than the moderately severe scenarios and, therefore,
are less likely to occur. While these estimates suggest that the capital
ratios are more than sufficient to protect the Fund at this time from many
worse- than- expected loan performance scenarios, factors not fully
captured in our models could affect the Fund's ability to withstand
worsethan- expected experiences over time. These factors include recent
changes in FHA's insurance program and the conventional mortgage market that
could affect the likelihood of poor loan performance and the ability of the
Fund to withstand that performance. For example, conventional mortgage
lenders and private mortgage insurers have recently lowered the required
down payment on loans. Such actions may have attracted some lower risk
borrowers who would otherwise have insured their loans with FHA. As a
result, the overall riskiness of FHA's portfolio may be greater than we have
estimated, making a given amount of capital less likely to withstand future
economic downturns than we have predicted. 8 Appendix II discusses the
sensitivity of our results to some of the assumptions we made in our
analysis.
At This Time, the Capital Beginning with the robust economy and the value of
the Fund in 1999, our
Ratio Appears Sufficient to analysis shows that a 2- percent capital ratio
appears sufficient to withstand
Withstand Moderately worse- than- expected loan performance that results
from moderately
Severe Economic Scenarios severe economic scenarios similar to those
experienced over the last 25
years. Our model and others that are based on historical experience That Are
Based on Recent
suggest that falling house prices and high levels of unemployment are likely
Historical Experience
to produce poor mortgage performance. Thus, to test the Fund's ability to
withstand worse- than- expected loan performance, we developed economic
scenarios that are based on certain regional downturns and the 1981- 82
national recession.
We tested the adequacy of the capital ratio using economic scenarios that
were based on three recent regional economic downturns- one in the west
south central region of the United States that began in 1986, one in New
England that began in 1989, and one in California that began in 1992- that
produced high mortgage foreclosure rates in those regions. 9 The degree to
which these downturns affected the Fund depended on their severity as well
as on the volume of mortgages insured by FHA in that region. Thus, while New
England suffered a severe downturn in the late 1980s and early 1990s, the
Fund did not suffer significantly because the volume of loans that FHA
insures in New England represents a small share of FHA's total volume of
insured loans. Because regional averages diminish the impact of
the adverse economic experience, from each region we selected a state with
particularly poor experience as the basis for our scenarios. We also
adjusted the scenarios to recognize that the forecasts start from the
economic conditions that existed at the end of 1999. See appendix III for
further discussion of the scenarios that we used to test the adequacy of
FHA's capital ratio. As can be seen in table 2, neither the scenarios that
are based on regional downturns nor the scenario that is based on the 1981-
82 national recession had much of an effect on the value of the Fund. More
specifically, in these worse- than- expected scenarios that are based on
specific historical experiences, the estimated capital ratio never falls
below 2.8 percent, which is only 0. 4 percentage points below our estimated
capital ratio using
9 The west south central region includes those states in the west south
central Bureau of the Census division- Arkansas, Louisiana, Oklahoma, and
Texas. The New England region includes Connecticut, Maine, Massachusetts,
New Hampshire, Rhode Island, and Vermont. California is part of the Pacific
region, which corresponds to the Pacific census division and
also includes Alaska, Hawaii, Oregon, and Washington.
expected economic conditions. However, the national recession had the
greatest impact because it affected FHA's entire portfolio.
Table 2: Capital Ratios Under Expected and Historical Economic Scenarios
Capital ratio a Scenario Description (percent)
Expected economic conditions Unemployment and interest rates vary as
Standard & Poor's DRI 3.20 forecasts; house price growth is adjusted for
constant quality and slower growth. b West south central downturn House
prices and unemployment change as they did in Louisiana from 3.06
1986 through 1990. New England downturn House prices and unemployment change
as they did in Massachusetts 3.14 from 1988 through 1992. Pacific downturn
House prices and unemployment change as they did in California from 2.89
1991 through 1995. 1981- 82 national recession For each state, house prices,
unemployment rates, and interest rates 2.81 change as they did from 1981
through 1985. a In estimating these capital ratios, we assume that the
scenario begins in the first year after the most recent loans were
originated. In our data, the most recent loans were originated in fiscal
year 1999; therefore, scenarios start in fiscal year 2000. We knew when we
performed this analysis that none of these scenarios actually occurred in
fiscal year 2000, but we wanted to test the ability of the Fund to withstand
an economic downturn when many of FHA's loans were new and the borrowers had
not accumulated much equity. In addition, results for scenarios beginning a
year later had less of an effect on the Fund. b Standard and Poor's DRI is a
private economic forecasting company.
Source: GAO analysis.
Under More Severe Although the Fund's estimated capital ratio at the end of
fiscal year 1999
Economic Scenarios, the fell by considerably less than 2 percentage points
under economic Capital Ratio Could Fall by
scenarios that are based on recent regional experiences and the 1981- 82
More Than 2 Percentage national recession, our model suggests that
extensions of some historical Points regional scenarios to broader regions
of the country could cause the capital ratio to fall by more than 2
percentage points. Specifically, to test whether a 2- percent capital ratio
could withstand more severe economic conditions,
we extended the regional scenarios to two regions and then to the nation as
a whole. However, we recognize that these extensions are less likely to
occur than the historical scenarios that affected a single region. As table
3 shows, if any of these downturns simultaneously hit two regions where FHA
has significant business- the west south central and Pacific regions- the
estimated capital ratio would be less than 2 percentage points lower than it
would be with expected loan performance. In addition, even if the
entire nation experienced a downturn similar to two of the three regional
downturns that we analyzed, the estimated capital ratio would still fall by
less than 2 percentage points. However, a national downturn as severe as
that experienced by Massachusetts from 1989 through 1992 would cause
our estimate of the 1999 capital ratio to fall by more than 2 percentage
points.
Table 3: Capital Ratios Under Expected and More Severe Economic Scenarios in
Selected Locations Capital ratio for
Capital ratio for scenarios in two
national scenarios Scenario Description
regions a (percent) (percent)
Expected econom Unemployment and interest rates NA 3. 20 conditions vary as
DRI forecasts; house price
growth is adjusted for constant quality and slower growth. b Extensions of
historical regional downturns West south central
House prices and unemployment 2. 81 2.31 downturn rates change as they did
in
Louisiana from 1986 through 1990. New England downturn House prices and
unemployment 2. 14 0.81
rates change as they did in Massachusetts from 1988 through 1992.
Pacific downturn House prices and unemployment 2. 59 2.16 rates change as
they did in California from 1991 through 1995.
National scenarios with interest rate changes or high foreclosure rates
Induced refinancing Mortgage interest rates fall,
NA 1. 37 followed by a recession inducing borrowers to refinance, and then a
recession sets in, such that the unemployment rate rises
and house prices fall. Rising interest rate Mortgage and other interest
rates NA 3. 36 scenario are higher from 2000 through 2004
than under expected economic conditions.
Scenario with foreclosure Foreclosure rates in 2000 through
NA 0. 92 rates from the 1980s 2004 equal foreclosure rates from 1986 to 1990
for mortgages
originated in most recent 10- year period c a The two regions are the
Pacific and west south central census divisions. b Standard & Poor's DRI is
a private economic forecasting company.
c This scenario does not vary foreclosure rates for streamline refinanced or
adjustable rate mortgages because there are no data on these products for
the 10- year period prior to 1986. Source: GAO analysis.
Because we were concerned that the historical scenarios we were considering
might not be adequate to test the effect of changes in interest rates, we
developed two additional scenarios: one in which mortgage
interest rates fall and then a recession sets in and one in which mortgage
and other interest rates rise to levels that are higher than those in the
expected economic conditions scenario. The first scenario is more likely to
exhaust a 2- percent capital ratio.
Under a scenario in which mortgage interest rates fall and then a recession
sets in, the drop in interest rates might induce some homeowners to
refinance their mortgages. For those homeowners who refinance outside of
FHA, the fund would no longer be accumulating revenue in the form of annual
premiums; if the homeowners have not had their mortgages for
long, they would receive some premium refunds. Moreover, those borrowers who
use FHA's streamline refinance provision that allows borrowers to refinance
their mortgages without a new appraisal of their home will likely pay annual
premiums for fewer years than if they had not refinanced. 10 So, cash
outflows would have increased and cash inflows
would have decreased before the recession hits. When the recession hits,
cash outflows would increase further because of increased foreclosures among
the remaining borrowers. As table 3 shows, our model predicts that
the capital ratio would fall substantially- by almost 2 percentage points-
under this scenario.
A scenario with rising mortgage interest rates will affect various loan
types differently. Because the payments on adjustable rate mortgages
increase as interest rates rise, there is an increased likelihood that
borrowers with these types of mortgages will default. However, since FHA-
insured mortgages are assumable, rising interest rates make fixed- rate
mortgages
10 Borrowers with 30- year mortgages who borrow more than 95 percent of the
value of their home are required to pay annual premiums equal to 0.5 percent
of the remaining balance of their mortgage for 30 years. However, if these
borrowers refinance their mortgages using
FHA's streamline program, they will have to pay annual premiums for no more
than 7 years. In “Mortgage Refinancing, Adverse Selection, and FHA's
Streamline Program”, Journal of Real Estate Finance and Economics
(vol. 21, issue 2), David Brickman and Patric Hendershott estimated that
borrowers who borrowed 97 percent of the value of their homes in 1994 and
immediately refinanced would have reduced the net present value of their
premium payments by 32 percent.
more valuable to those borrowers holding them. This decreases the likelihood
that borrowers with these types of mortgages will default. Insurance on
loans originated in 1998 and 1999 make up 42 percent of
FHA's portfolio at the end of fiscal year 1999, and the insured loans are
predominately fixed rate mortgages. Consequently, it is not surprising that
a rising interest rate scenario leads to an increase in the value of the
Fund. Because our economic model did not predict regional or national
foreclosure rates as high as those experienced during the 1980s in any of
our scenarios, we estimated cash flows using foreclosure rates that more
closely matched regional experience during the 1980s. Specifically, we
assumed that for mortgages originated from 1989 through 1999, foreclosure
rates in 2000 through 2004 would equal those experienced from 1986
through 1990 by FHA- insured loans that originated between 1975 and 1985 in
a given region. As table 3 shows, the capital ratio fell to 0.92 percent
under this scenario. To test an even more severe scenario, one similar to
that used by rating agencies for private mortgage insurers, 11 we also
calculated future cash flows assuming that foreclosure rates in 2000 through
2004 extended the very poor performance of the west south central mortgages
in the 1980s to ever larger portions of FHA's insurance portfolio.
As figure 3 shows, we found that if 36.5 percent of FHA- insured mortgages
experienced these high default rates, the estimated capital ratio for fiscal
year 1999 would fall by 2 percentage points. If about 55 percent of FHA's
portfolio experienced these conditions, a less likely event, the capital
ratio would be 0. 11 For example, Standard & Poor's assumes that over the
next 10 years, default rates for fixed rate mortgages will equal those from
the west south central states in the 1980s. In addition, they assume that no
new loans will be insured. The rating for private mortgage
insurers depends on how much capital a company has at the end of this 10-
year experience.
Figure 3: Capital Ratios Resulting From Applying the Average 1986- 90
Foreclosure Rates in the West South Central Census Division to Varying
Proportions of FHA's Insurance Portfolio in 2000- 04
4 Capital ratio (percent) 3 2 1 0
020 40 60 80100 -1
-2 -3
Percentage of FHA's portfolio
Point at which the capital ratio would fall by 2 percentage points Capital
ratio with 1986 to 1990 foreclosure rates
Note: West south central mortgages made up 9 percent of FHA's portfolio in
1999. This analysis does not change foreclosure rates for streamline
refinanced or adjustable rate mortgages because there are no data on these
products for the 10- year period prior to 1986. Source: GAO analysis.
Other Factors May Affect Because our models are based on the relationship
between foreclosures
the Adequacy of the Capital and prepayments and certain economic factors
from fiscal years 1975
Ratio through 1999, they do not account for the potential impact of recent
events,
such as changes in FHA's program or in the behavior of the conventional
mortgage market. In addition, our models assume that no additional changes
in FHA's program or the conventional mortgage market that would affect FHA-
insured loans originated through 1999 take place during the forecast period,
which extends from fiscal years 2000 through 2028. To the extent that any
such changes cause foreclosure and prepayment rates on existing FHA- insured
loans to be higher or lower than we have predicted,
the Fund's capital ratio would be different under the various scenarios we
have discussed. Furthermore, our analysis does not attempt to predict how
loans insured by FHA after fiscal year 1999 will behave. Future changes in
FHA's program, such as the premium changes adopted as of January 1,
2001, 12 or in the conventional mortgage market may make future loans
perform better or worse than we might expect from past experience. In
addition, these changes may increase or reduce the amount of cash flowing
into the Fund and thus its ability to withstand worse- than- expected loan
performance in the future.
Changes in FHA's Insurance HUD and the Congress can change FHA's insurance
program in a variety of Program
ways, including changes in refund policy and underwriting standards. In
fact, HUD and the Congress have taken the following actions in recent years
that could affect the Fund in ways that are not accounted for in our models:
? HUD has suggested that it will reinstitute distributive shares and
Members of Congress have introduced bills requiring HUD to take that action.
The immediate consequence of this action would be that cash flows out of the
Fund would be higher than our estimates. ? During the late 1990s, the
Congress required that FHA implement a new
loss mitigation program that encourages lenders to take actions to lower
defaults on FHA- insured mortgages. The program requires that lenders
provide homebuyers with certain options to avoid foreclosure. While it is
hoped that losses from foreclosures will decline as a result of this
program, if foreclosure is simply delayed as a result of forbearance,
losses could ultimately be larger in the long run. In either case, actual
cash flows would likely be different than our estimates. ? FHA has also
reduced up- front premiums for new homeowners who receive financial
counseling before buying a home. If the program reduces the likelihood that
these homeowners will default, losses would be lower than we have estimated.
13 12 These changes include reducing the up- front premium for all
homebuyers to 1.5 percent (a reduction of up to 0. 75 percentage points) and
canceling the annual mortgage insurance
premiums for most borrowers when the value of the mortgage reaches 78
percent of the original price of the house.
13 As a result of the reduction in up- front premiums for all borrowers on
January 1, 2001, borrowers receiving financial counseling before purchasing
a home will no longer pay lower up- front premiums than other borrowers.
? HUD has taken action to improve the oversight of lenders and better
dispose of properties and is continuing to implement new programs in these
areas. Better oversight of lenders could mean that losses on existing
business would be lower than we have predicted, and better practices for
disposing of property could reduce losses associated with foreclosed
properties below the level we have estimated.
Our models do not look at cash flows associated with loans that FHA would
insure after fiscal year 1999. However, recent and future changes in FHA's
insurance program will affect the likelihood that these loans will perform
differently than past experience suggests they will. If, for example,
FHA loosens underwriting standards, there is a greater likelihood that
future loans would perform worse than past experience suggests. In addition,
changes in premiums, such as the recent reductions in up- front premiums,
could reduce cash inflows into the Fund and, therefore, reduce the Fund's
ability to withstand poor loan performance. However, this premium change
could also lower the riskiness of the loans FHA insures. Changes in the
Conventional Recent changes in the conventional mortgage market, especially
changes in Mortgage Market
FHA's competitors' behavior, may also affect the estimates we have made
concerning the Fund's ability to withstand adverse economic conditions over
the long run. Homebuyers' demand for FHA- insured loans depends, in part, on
the alternatives available to them. In recent years, FHA's competitors in
the conventional mortgage market- private mortgage insurers and conventional
mortgage lenders- are increasingly offering products that compete with FHA's
for those homebuyers who are
borrowing more than 95 percent of the value of their home. These
developments in the conventional mortgage market may have increased the
average risk of FHA- insured loans in the late 1990s. In particular, by
lowering the required down payment, conventional mortgage lenders and
private mortgage insurers may have attracted some borrowers who might
otherwise have insured their mortgages with FHA. If, by selectively offering
these low down payment loans, conventional mortgage lenders and private
mortgage insurers were able to attract FHA's lower- risk borrowers, recent
FHA loans with down payments of less than 5 percent may be more risky
on average than they have been historically. If this effect, known as
adverse selection, has been substantial, the economic value of the Fund may
be lower than we estimate, and it may be more difficult for the Fund to
withstand worse- than- expected loan performance than our estimates suggest.
In addition, should these competitive pressures persist, newly insured loans
are likely to perform worse than prior experience would suggest, and then
any given capital ratio would be less able to withstand
such performance. FHA is taking some action to more effectively compete. For
example, FHA is attempting to implement an automated underwriting system
that could enhance the ability of lenders underwriting FHA- insured
mortgages to distinguish better credit risks from poorer ones. Although this
effort is likely to increase the speed with which lenders process FHAinsured
loans, it may not improve the risk profile of FHA borrowers unless lenders
can lower the price of insurance for better credit risks.
The Impacts of Options Several options are available to the Secretary of HUD
under current
for Reducing FHA's legislative authority that could result in reducing FHA's
capital ratio. Other
options would require legislative action. Reliably measuring the impacts of
Capital Ratio Are these options on the Fund's capital ratio and FHA
borrowers is difficult Difficult to Predict without using tools designed to
estimate the multiple impacts that policy changes often have. While HUD has
substantially improved its ability to monitor the financial condition of the
Fund, neither the models used by HUD to assess the financial health of the
Fund, nor those used by others, explicitly recognize the indirect effects of
policy changes on the volume and riskiness of FHA's loans. As a result, the
impacts of the various policy
options on the federal budget are difficult to discern. However, any option
that results in a reduction in the Fund's reserve, if not accompanied by a
similar reduction in other government spending or by an increase in
receipts, would result in either a reduction in the surplus or an increase
in any existing deficit.
The Secretary of HUD and There are several changes to the FHA single- family
loan program that could
the Congress Have be adopted if the Secretary of HUD or the Congress
believes that the Numerous Options
economic value of the Fund is higher than the amount needed to ensure
Available to Reduce the actuarial soundness. For example, actions that the
Secretary could take that could reduce the value of the fund include
lowering insurance
Capital Ratio premiums, adjusting underwriting standards, and reinstituting
distributive
shares. 14 However, congressional action in the form of new legislation
would be required to make other program changes that are not now 14 Between
1943 and 1990, FHA rebated these so- called excess funds to borrowers as
distributive shares. In 1990, however, the Congress suspended the payment of
these shares
until the Secretary of HUD determines that the Fund is actuarially sound.
HUD has announced that it will resume paying distributive shares. HUD
officials said that they are developing systems to facilitate the payment of
these shares and expect to be ready to resume paying them in mid- 2001.
authorized or clearly contemplated by the statute. These would include
actions such as changing the maximum amount FHA- insured homebuyers may
borrow relative to the price of the house they are purchasing and using the
Fund's reserves for other federal programs. 15 Generally, the Secretary
of HUD, in making any authorized changes to the FHA single- family program,
must meet certain operational goals. These operational goals include (1)
maintaining an adequate capital ratio, (2) meeting the needs of homebuyers
with low down payments and first- time homebuyers by providing access to
mortgage credit, (3) minimizing the risk to the Fund
and to homeowners from homeowner default, and (4) avoiding adverse
selection.
Potential Effects of Options Reliably estimating the potential effect of
various options on the Fund's
on the Fund's Capital Ratio capital ratio and FHA borrowers is difficult
because the impacts of these and FHA Borrowers Are
policy changes are complex and tools available for handling these Difficult
to Measure
complexities may not be adequate. Policy changes have not only immediate,
straightforward impacts on the Fund and FHA's borrowers, but also more
indirect impacts that may intensify or offset the original effect.
Implementing these options could affect both the volume and the average
riskiness of loans made, which, in turn, could affect any future estimate of
the Fund's economic value. As a result of this complexity, obtaining a
reliable estimate would likely require that economic models be used to
estimate the indirect effects of policy changes. In 1990, the Congress
enacted legislation designed to provide better information on the Fund's
financial condition. The Omnibus Budget Reconciliation Act requires annual
independent actuarial reviews of the Fund and includes credit reforms that
require HUD to estimate, for loans originated in a given year, the net
present value of the anticipated cash flows over all the years that the
loans will be in existence. The models developed by HUD to comply with these
requirements are based on detailed analyses of the Fund's historical claim
and loss rates and have improved HUD's ability to monitor the financial
condition of the Fund. At this time, however, neither the models used by HUD
to assess the financial health of the Fund, nor those used by others,
explicitly recognize the indirect effects of policy changes on the volume
and riskiness of FHA's loans. As a result, HUD cannot reliably estimate the
impact of policy changes on the Fund. Although it is difficult to predict
the overall impact of a change on the Fund's capital ratio 15 During the 106
th Congress, legislation was introduced that proposed using the Fund's
resources to fund affordable rental housing (see S. 2997).
and thus on FHA borrowers as a whole, different options would likely have
different impacts on current and prospective FHA- insured borrowers.
Effect on the Fund's Capital Many of the proposals to reduce the capital
ratio, such as lowering
Ratio premiums or reinstituting distributive shares, will reduce the price
of FHA insurance to the borrower. If no change in the volume of loans FHA
insures is considered, then the effect of lowering premiums, for example,
clearly
would be to lower the economic value of the Fund. However, for two reasons,
this price reduction is likely to increase the volume of FHA loans
originated, which would increase both premium income and claims against the
Fund when some of these new loans default. First, by lowering the
price of FHA insurance relative to the price of private mortgage insurance,
this premium reduction would likely induce some borrowers who otherwise
would have obtained private mortgage insurance to obtain FHA insurance
instead, thereby increasing FHA's market share. Second, people who were
deferring home purchases because of the high price of FHA insurance might
buy homes with FHA insurance once the price is lower. Without a complete
analysis of the impact on the volume of loans, reliably
estimating the effect of lowering the premiums on the Fund's economic value
is difficult.
Furthermore, the economic value of the Fund is influenced not only by the
volume of loans FHA insures, but also by the riskiness of those loans.
Therefore, determining the effect a policy change will have on the economic
value of the Fund requires determining how the policy will affect the
riskiness of FHA- insured loans. In the case of lowering up- front premiums,
for example, the new FHA- insured loans could be less risky than
FHA's existing loans. As a result, the new loans would be profitable and
offset the direct impact of lower premiums. Generally, private mortgage
insurers require that borrowers meet higher credit standards than does FHA.
So, to the extent that these new FHA borrowers would have obtained private
mortgage insurance without the lower premiums, they are likely to have lower
risk profiles than the average for all current FHA borrowers. At
the same time, lowering up- front premiums is not likely to attract many
additional higher- risk borrowers who would previously not have qualified
for FHA- insured loans. 16 Because HUD does not have adequate tools to
handle the complexities of
estimating the ultimate impact of policy changes on the volume of FHAinsured
loans and the riskiness of those loans, these factors are not always
considered in assessing the impact of policy changes. For example, assuming
that the volume and riskiness of FHA- insured loans will not change, HUD
estimates that the recent reductions in up- front premiums
combined with the introduction of mortgage insurance cancellation policies
will lower the estimated value of the Fund by almost $6 billion over the
next 6 years. Because this estimate does not consider the possible changes
in the volume of loans that will be insured and the riskiness of those
loans, it is an estimate only of the direct impact rather than the full
impact of policy changes. Similarly, a recent study presents estimates that
lowering up- front premiums to 1.5 percent would result in an almost
fivefold increase in the likelihood that cash inflows would be less than
outflows over a random 10- year period. 17 However, this study notes that it
did not look at how these changes would affect the riskiness of new loans.
The complexity of estimating the impact of policy changes on the Fund
implies that economic models would be needed to reliably estimate the likely
outcomes. The most likely sources for such models would be the studies that
compute the economic value of the Fund; however, the models
HUD and others have been using to assess the financial health of the Fund do
not explicitly recognize the impact of policy changes on the economic value
of the Fund. Instead, they assume that FHA's market share remains
static. Effect on Borrowers Although it is difficult to predict the overall
impact of a change on the
Fund's capital ratio and thus on FHA borrowers as a whole, different options
would likely have different impacts on various FHA- insured borrowers. Some
proposals would more likely benefit existing and future 16 Most FHA
borrowers finance their up- front premiums over the life of their loans
(generally 30 years), rather than paying them up- front. As a result,
lowering the up- front premiums would not substantially reduce the amount of
cash required from borrowers at settlement. However, those borrowers who
could not qualify for a mortgage because their monthly payments would have
been too high relative to their income, might then qualify because the
reduction in up- front premiums would lower the monthly payment. 17 See
“Credit Risk, Capital, and FHA Mortgage Insurance ,” by Charles
A. Capone, Jr., forthcoming in the Journal of Housing Research (vol. 11,
issue 2).
FHA- insured borrowers, while others would benefit only future borrowers,
and still others would benefit neither of these groups. One interpretation
of the higher premiums that borrowers paid during the period in which the
economic value of the fund has been rising is that borrowers during the
1990s “overpaid” for their insurance. Some options for reducing
the capital ratio, such as reinstituting distributive shares, would be more
likely to compensate these borrowers. Paying distributive shares would
benefit certain existing borrowers who voluntarily terminate their
mortgages. If these policies continued into the future, they would also
benefit future
policyholders. Alternatively, reducing up- front premiums, reducing the
number of years over which annual insurance premiums must be paid, or
relaxing underwriting standards would tend to benefit only future borrowers.
Policy options that propose to use some of FHA's capital
resources for spending on other programs would benefit neither existing nor
future FHA- insured borrowers, but would instead benefit the recipients of
those programs receiving the new expenditures. For example, reducing the
capital ratio by shifting funds from the Fund to subsidize multifamily
housing may primarily benefit renters rather than single- family homeowners.
However, over time such a policy could be sustained only so long as FHA
borrowers continue to pay premiums higher than the cost to FHA of insuring
single- family mortgages.
Potential Impact of Options Because of the difficulty in reliably measuring
the effect of most actions on the Federal Budget Is that could be taken
either by the Secretary of HUD or the Congress on the
Difficult to Discern Fund's capital ratio, we cannot precisely measure the
effect of these
policies on the budget. 18 However, any actions taken by the Secretary or
the Congress that influence the Fund's capital ratio will have a similar
effect on the federal budget. Specifically, any proposal that results in a
reduction in the Fund's reserve, if not accompanied by a similar reduction
in other government spending or by an increase in receipts, would result in
either a reduction in the surplus or an increase in any existing deficit. If
the Secretary or the Congress adopts policies, such as paying distributive
shares or relaxing underwriting standards, that could reduce the
profitability of the Fund, both the negative subsidy amount reported in
FHA's budget submission and the Fund's reserve would be lower. Some of 18
However, the Congressional Budget Office has begun building a model that it
believes will
allow it to forecast simulations and what- if analyses with the objective of
making FHA's budget reporting more transparent and informative.
these policies- such as paying distributive shares- would affect FHA's cash
flows immediately. Thus, the amount of money available for FHA to invest in
Treasury securities would be lower. The Treasury, in turn, would
have less money available for other purposes, and any overall surplus would
decline or any deficit would rise. If the amounts of cash flowing out of the
Fund exceeded current receipts, FHA would be required to redeem its
investments in Treasury securities to make the required payments. The
Treasury, then, would be required to either increase borrowing from the
public or use general tax revenues to meet its financial obligations to FHA.
In either case, any annual budget surplus would be lower or deficit higher.
Conclusions At the end of fiscal year 1999, the Fund had a capital ratio
that exceeded 2 percent of FHA's insurance- in- force- the minimum required
by law; however, whether the fund was actuarially sound is not so clear.
Neither
the statute nor HUD has established criteria to determine how severe of a
stress the Fund should be able to withstand, that is, what constitutes
actuarial soundness. Our results show that as of the end of fiscal year
1999,
only the most severe circumstances that we analyzed would cause the current
economic value of the Fund to fall below 0.
One method of determining actuarial soundness would be to estimate the value
of the Fund under various economic and other scenarios. In our analysis, the
required minimum capital ratio of 2 percent appears sufficient to cover most
of the adverse economic scenarios we tested, although it would not be
possible to maintain the minimum under all scenarios. Nonetheless, we urge
caution in concluding that the estimated value of the Fund today implies
that the Fund could withstand the specified economic scenarios regardless of
the future activities of FHA or the market. Our
estimates and those of others are valid only under a certain set of
conditions, including that loans FHA recently insured respond to economic
conditions similarly to those it insured in the more distant past, and that
cash inflows associated with future loans at least offset outflows
associated with those loans. However, HUD is changing several policies that
may affect the volume and quality of its future business. Further, adverse
economic events cannot be predicted with certainty; therefore, we cannot
attach a likelihood to any of the scenarios that we tested (even
though we recognize that it is less likely that a severe economic downturn
will affect the whole nation than one or two regions). It is instructive to
remember in considering the uncertainty of the future, that the Fund had an
even higher capital ratio in 1979 when the economic value of the Fund
equaled 5.3 percent of insurance- in- force, but in little more than a
decade-
after a national recession, the substitution of an up- front premium for
annual insurance premiums, and regional real estate declines- the economic
value of the Fund was negative. Thus, it is important to periodically
reevaluate the actuarial soundness of the Fund.
Today, FHA knows more about the condition of the Fund but could still
improve its evaluation of the impact that unexpected economic downturns and
policy changes may have on the Fund. HUD has already taken some action that
it estimates will lower the value of the Fund, including reducing up- front
insurance premiums on newly insured mortgages. HUD has done so without the
tools necessary to reliably measure the multiple impacts
that these policies are likely to have. While the direct impact of policies
that are likely to reduce the Fund's capital ratio can be estimated with the
models used in the actuarial reviews, those models cannot isolate the
indirect effects on the volume of loans insured by FHA and the riskiness of
those loans.
Matters for The Congress may want to consider taking action to amend the
laws Congressional
governing the Fund to specify criteria for determining when the Fund is
actuarially sound. Because we believe that actuarial soundness depends on
Consideration a variety of factors that could vary over time, setting a
minimum or target capital ratio will not guarantee that the Fund will be
actuarially sound over time. For example, if the Fund were comprised
primarily of seasoned loans with known characteristics, a capital ratio
below the current 2- percent minimum might be adequate, but under conditions
such as those that
prevail today, when the Fund is comprised of many new loans, a 2- percent
ratio might be inadequate if recent and future loans perform considerably
worse than expected. Thus, the Congress may want to consider defining the
types of economic conditions under which the Fund would be expected to meet
its commitments without borrowing from the Treasury. Recommendations If the
Congress decides that no further guidance is necessary, to better evaluate
the health of the Fund and determine the appropriate types and
timing of policy changes, we recommend that HUD develop criteria for
measuring the actuarial soundness of the Fund. These criteria should specify
the economic conditions that the Fund would be expected to withstand and may
specify capital ratios currently consistent with those
criteria.
Because many conditions affect the adequacy of a given capital ratio, we
recommend that the independent annual actuarial analysis give more attention
to tests of the Fund's ability to withstand appropriate stresses.
These tests should include more severe scenarios that capture worse-
thanexpected loan performance that may be due to economic conditions and
other factors, such as changes in policy and the conventional mortgage
market.
To more fully assess the impact of policy changes that are likely to
permanently affect the profitability of certain FHA- insured loans, we
recommend that the Secretary of HUD develop better tools for assessing the
impact these changes may have on the volume and riskiness of loans that FHA
insures. Such analysis is particularly important where the policy change
permanently affects certain loans, as in the case of underwriting and
premium changes. Without a better analytical framework to assess the full
impact of policy changes that permanently affect certain loans, we recommend
that such changes be made in small increments so that their
impact can be monitored and adjustments can be made over time. Agency
Comments and
We provided a draft of this report to the Secretary of HUD for his review
Our Evaluation
and comment. HUD agreed with the report's findings regarding the estimated
value of the fund, and the ability of the fund to withstand moderately
severe economic downturns that could lead to worse- thanexpected loan
performance. However, HUD expressed concern that the report did not note the
probability of the most stressful scenarios we tested and FHA's ability to
react to adverse developments. HUD also thought our reference to the
substantial decline in the capital ratio that occurred during
the 1980s left a false impression that the Fund is currently in jeopardy. In
addition, HUD expressed concern that the report did not fully recognize the
improvements it has made in analyzing policy changes and monitoring the
performance of the Fund and disagreed with our recommendations. HUD's letter
is reproduced in appendix IV.
In response, we clarified that scenarios in which we extend historical
adverse economic conditions more widely are less likely to occur. However,
we cannot attribute a probability to any scenario we used. We also
acknowledge that the annual actuarial reviews and the annual reestimates of
the Fund required under the housing and credit reforms of 1990 enable HUD to
better monitor the performance of the Fund and, therefore, react to adverse
developments. However, we remain concerned that HUD's analyses of policy
changes do not fully recognize the impact
that these policy changes may have on the volume of loans FHA will insure
and the riskiness of those loans. We also disagree that the reference to the
decline in the capital ratio experienced in the 1980s implies that the Fund
is in jeopardy today. In fact, this example serves to illustrate that
changes in the economy and HUD policy can have a dramatic impact on the
value of the Fund. With regard to our recommendation that HUD develop
criteria for measuring the actuarial soundness of the Fund, HUD seems to
infer that we believe a static capital ratio should be the criterion for
measuring actuarial soundness. We do not recommend a static capital ratio
for measuring actuarial soundness. Rather, we believe that it is important
to measure actuarial soundness under different economic and other scenarios;
therefore, we recommend that HUD specify the conditions that the Fund would
be expected to withstand. We revised this recommendation to make clear that
the definition of actuarial soundness should consider the economic
conditions that the Fund would be expected to withstand.
Regarding our recommendation that the independent annual actuarial analysis
give more attention to tests of the Fund's ability to withstand appropriate
stresses, HUD noted that it believed it was already complying with this
recommendation and asked that our report define more specifically what tests
are needed. In response, we clarified that the annual actuarial review
should include more severe scenarios that capture worsethan-
expected loan performance that may be due to economic conditions and other
factors, such as changes in HUD policy and the conventional mortgage market.
HUD's recent actuarial analysis included two scenarios- an interest rate
spike scenario and a lower house price appreciation scenario- for testing
the value of the Fund under a stressed economic state, and in neither
scenario do house prices decline or unemployment rates rise. 19
With regard to our recommendation concerning tools for assessing the impact
of policy changes, HUD disagreed that any tools are needed beyond those that
it already has. Specifically, HUD cites the annual analyses done in
compliance with the Federal Credit Reform Act of 1990 and its annual
actuarial reviews that already focus on policy changes. Further, HUD notes
that it has made its program data more accessible for policy analysis 19
These scenarios are in addition to the three forecasts of key economic
variables provided by Standard & Poor's DRI that Deloitte did not consider
to be stressed economic states.
through the creation of the Single Family Data Warehouse. However, we remain
concerned that HUD does not have adequate tools for assessing the full
impact that policy changes may have. Tools such as models for estimating the
change in demand and the risk characteristics of future loans would enable
HUD to better estimate the full impact that policy changes may have on the
value of the Fund. HUD also disagreed with the idea that any policy actions
it takes should be only incremental and reversible. We revised our
recommendation to make clear that incremental changes are appropriate where
a policy change permanently affects certain
loans. Copies of this report will be distributed to interested congressional
committees; the Honorable Mel Martinez, Secretary of the Department of
Housing and Urban Development; the Honorable Mitchell E. Daniels, Jr., the
Director of the Office of Management and Budget; and the Honorable Dan L.
Crippen, the Director of the Congressional Budget Office. We will also make
copies available to others on request. If you or your staff have any
questions about this report, please contact me at (202) 512- 8678. Key
contributors to this report are listed in appendix V. Thomas J. McCool
Managing Director, Financial Markets and Community Investment
Appendi Appendi xes x I
Scope and Methodology To estimate the economic value of the Federal Housing
Administration's (FHA) Fund as of September 30, 1999, and its resulting
capital ratio, we developed econometric and cash flow models. These models
were based on models that we developed several years ago for this purpose.
In developing the earlier models, we examined existing studies of the
singlefamily housing programs of both the Department of Housing and Urban
Development (HUD) and the Department of Veterans Affairs (VA); academic
literature on the modeling of mortgage foreclosures and
prepayments; and previous work that Price Waterhouse (now
PricewaterhouseCoopers), HUD, VA, ourselves, and others had performed on
modeling government mortgage programs. For our current analysis, we modified
our previous models on the basis of our examination of work performed
recently by PricewaterhouseCoopers, Deloitte & Touche, and
others; discussions we held with analysts familiar with modeling mortgage
foreclosures and prepayments; and program changes made by FHA since our
previous work was performed. For these models, we used data supplied by FHA
and Standard & Poor's DRI, a private economic forecasting company. We also
used information from FHA's independent actuarial reviews in our analysis.
Our econometric analysis estimated the historical relationships between the
probability of loan foreclosure and prepayment and key explanatory factors,
such as the borrower's equity and the interest rate. To estimate these
relationships, we used HUD's A- 43 data on the default and
prepayment experience of FHA- insured home mortgage loans that originated
from fiscal years 1975 through 1999. To test the validity of our econometric
models, we examined how well the models predicted the actual rates of FHA's
loan foreclosures and prepayments through fiscal year 1999. We found that
our predicted rates closely resembled the actual rates. Next, we used our
estimates of these relationships and forecasts of future economic conditions
provided by Standard & Poor's DRI to develop
a baseline forecast of future loan foreclosures and prepayments for loans
that were active at the end of fiscal year 1999.
To estimate the net present value of future cash flows of the Fund under
expected economic conditions, we used our forecast of future loan
foreclosures and prepayments in conjunction with a cash flow model that we
developed to measure the primary sources and uses of cash for loans
that originated from fiscal years 1975 through 1999. Our cash flow model was
constructed to estimate cash flows for each policy year through the life of
a mortgage. An important component of the model was the conversion of all
income and expense streams- regardless of the period in which they
are actually forecasted to occur- into their 1999 present value equivalents.
We then added the forecasted 1999 present values of the future cash flows to
the current cash available to the Fund, which we obtained from documents
used to prepare FHA's 1999 audited financial statements, to estimate the
Fund's economic value and resulting capital ratio. A detailed discussion of
our models and methodology for estimating the economic value and capital
ratio of the Fund appears in appendix II. To compare our estimates of the
Fund's economic value and capital ratio with the estimates prepared for FHA
by Deloitte & Touche, we reviewed Deloitte's report and met with its
analysts and HUD officials to learn more about that study's methodology,
data, and assumptions.
To determine the extent to which a capital ratio of 2 percent would allow
the Fund to withstand worse- than- expected loan performance, we developed
various scenarios for future economic conditions that we anticipated would
result in substantially worse loan performance than we forecasted in our
scenario using expected economic conditions. We based these scenarios on the
economic conditions that led to episodes of relatively high foreclosure
rates for FHA single- family loans in certain regions of the country at
different times during the 1975 through 1999
period and on those experienced nationally during the 1981- 82 recession. We
developed additional scenarios that extended the adverse regional economic
conditions to larger sections of the country to analyze how well the Fund
could withstand conditions even worse than what we had
experienced in the past 25 years. We also developed some additional
scenarios with even higher foreclosure rates to further analyze the Fund's
ability to withstand adverse conditions. Under each of the scenarios that we
developed, we used our estimated relationships between foreclosure and
prepayment rates and various explanatory factors, and the future economic
conditions implied by the scenarios, to forecast future foreclosures and
prepayments for loans that were active at the end of fiscal year 1999. We
then used these forecasts, in conjunction with our cash flow model, to
estimate the economic value and
capital ratio of the Fund under each scenario. The difference between these
estimates and our estimate under expected economic conditions shows whether
each scenario is likely to result in a reduction of the Fund's economic
value of more than 2 percent and, therefore, whether a 2- percent capital
ratio is likely to be sufficient to allow the Fund to withstand the
worse- than- expected loan performance associated with such a scenario.
Our analysis of the adequacy of FHA's capital ratio is limited to the
performance of loans in FHA's portfolio as of the end of fiscal year 1999.
That is, our analysis assesses the likelihood that an economic value of 2
percent of the unamortized insurance- in- force would be sufficient to cover
the excess of future payments over future cash inflows (on a net present
value basis) on those loans if they perform worse than expected. Our
analysis of the ability of the Fund to withstand various adverse economic
conditions requires making the assumption that the adverse conditions would
not also cause loans insured by FHA after fiscal year 1999 to be an economic
drain on the Fund. Since the 1990 reforms, the cash flows associated with
each year's loans have been estimated to have a positive economic value,
thereby adding to the economic value of the entire Fund. However, during
adverse economic times, new loans might perform worse than loans that were
insured by FHA during the 1990s. If the newly insured loans perform so
poorly that they have a negative economic value, then the loss to the Fund
in any of the adverse economic scenarios that we have considered would be
greater than what we have estimated. Alternatively, if the newly insured
loans have positive economic values, then the Fund
would continue to grow. To identify other factors, such as recent program
and market changes, that could cause worse- than- expected loan performance,
we reviewed the laws and regulations governing FHA's insurance program,
studied recent actuarial reviews of the Fund, and interviewed experts. We
considered
these other factors because the relationships estimated in our econometric
models are based on historical relationships since 1975. As a result, these
models might not capture the effects of recent changes in FHA programs or
the conventional mortgage market on the likelihood that loans insured in
the late 1990s will foreclose or prepay. In addition, our forecasts of
future cash flows assume that FHA's program and the private mortgage market
will not change over the 30- year forecast period in any way that would
affect FHA- insured loans originated through 1999. To identify options for
adjusting the size of the Fund and determining the impact that these options
might have, we reviewed the laws and regulations governing FHA's insurance
program and proposals to use the Fund's economic value or otherwise change
FHA's insurance program. Additionally, we interviewed experts both within
and outside the federal
government. When available, we collected HUD's estimates of the impact of
various options on the Fund and the estimates of other experts. To determine
the impact of these changes on the federal budget, we relied on
our own experts as well as those at the Office of Management and Budget and
the Congressional Budget Office.
We conducted our review from December 1999 to February 2001 in accordance
with generally accepted government auditing standards.
Models Used to Estimate the Economic Value
Appendi x II
of FHA's Mutual Mortgage Insurance Fund We built econometric and cash flow
models to estimate the economic value of HUD's FHA's Mutual Mortgage
Insurance Fund (Fund) as of the end of fiscal year 1999. The goal of the
econometric analysis was to forecast mortgage foreclosure and prepayment
activity, which affect the flow of cash into and out of the Fund. We
forecasted activity for all loans active at the end of fiscal year 1999 for
each year from fiscal years 2000 to 2028 on
the basis of assumptions stated in this appendix. We estimated equations
from data covering fiscal years 1975 through 1999 that included all 50
states and the District of Columbia, but excluded U. S. territories.
Our econometric models used observations on loan years- that is, information
on the characteristics and status of an insured loan during each year of its
life- to estimate conditional foreclosure and prepayment probabilities. 1
These probabilities were estimated using observed patterns of prepayments
and foreclosures in a large set of FHA- insured loans. More specifically,
our model used logistic equations to estimate the logarithm of the odds
ratio, 2 from which the probability of a loan's payment (or a loan's
prepayment) in a given year can be calculated. These equations are expressed
as a function of interest and unemployment rates, the borrower's equity
(computed using a house's price and current and contract interest rates as
well as a loan's duration), the loan- to- value (LTV)
ratio, the loan's size, the geographic location of the house, and the number
of years that the loan has been active. The results of the logistic
regressions were used to estimate the probabilities of a loan being
foreclosed or prepaid in each year.
FHA pays a claim on a foreclosed mortgage and sometimes, depending on the
age of the loan, refunds a portion of the up- front premium when a mortgage
prepays. These two actions contribute to cash outflows. Cash inflows are
generated when FHA sells foreclosed properties and when borrowers pay
mortgage insurance premiums. We forecasted the cash flows into and out of
the Fund on the basis of our foreclosure and
1 These probabilities are conditional because they are subject to the
condition that the loan has remained active until a given year. 2 If
“P” is the probability that an event will occur, the “odds
ratio” is defined as P/( 1P). The logistic transformation is the
natural logarithm of the odds ratio, or LN[ P/( 1- P)], of which the
logistic regression provides an estimate. See G. S.
Maddala, Limited Dependent Variables and Qualitative Variables in
Econometrics (Cambridge : Cambridge Univ. Press, 1983). Also see John H.
Aldrich and Forrest D. Nelson, Linear Probability, Logit, and Probit Models
(SAGE Publications: Beverly Hills, London, and New York, 1984), pp. 41- 44.
prepayment models and key economic variables. We then used the forecasted
cash flows, including an estimate of interest that would be earned, and the
Fund's capital resources to estimate the economic value of the Fund.
We prepared separate estimates for fixed- rate mortgages, adjustable rate
mortgages (ARMs), and investor loans. The fixed- rate mortgages with terms
of 25 years or more (long- term loans) were divided between those that
refinanced and those that were purchase money mortgages (mortgages
associated with home purchase). Separate estimates were prepared for each
group of long- term loans. Likewise, investor loans were divided between
mortgages that refinanced and the loans that were purchase money mortgages.
We prepared separate estimates for each group of investor loans (refinanced
and purchase money mortgages). A separate analysis was also prepared for
loans with terms that were less than 25 years
(short- term loans). A complete description of our models, the data that we
used, and the results that we obtained is presented in detail in the
following sections. In particular, this appendix describes (1) the sample
data that we used; (2) our
model specification and the independent variables in the regression models;
(3) the model results; (4) the cash flow model, with emphasis on key
economic variables; and (5) a sensitivity analysis that demonstrates the
sensitivity of our forecasts to the values of some key variables.
Data and Sample For our analysis, we selected from FHA's computerized files
a 10- percent Selection
sample of records of mortgages insured by FHA from fiscal years 1975 through
1999 (1,465,852 loans). For the econometric models related to long- term,
fixed- rate mortgages, we used 25 percent of the long- term loans in our
sample. From the FHA records, we obtained information on the initial
characteristics of each loan, such as the year of the loan's origination
and the state in which the loan originated; LTV ratio; loan amount; and
contract interest rates. We categorized the loans as foreclosed, prepaid, or
active as of the end of fiscal year 1999.
To describe macroeconomic conditions at the national and state levels, we
obtained data from Standard & Poor's DRI, 3 by state, on annual civilian 3
Formerly DRI/ McGraw- Hill, Standard & Poor's DRI is a leading economic
forecasting firm.
unemployment rates and data from the 2000 Economic Report of the President
on the implicit price deflator for personal consumption expenditures. We
used Standard & Poor's DRI data on quarterly interest rates for 30- year
mortgages on existing housing along with its forecast data,
at the state level, on median house prices and civilian unemployment rates,
and at the national level, on interest rates on 1- and 10- year U. S.
Treasury securities.
Specification of the People buy houses for consumption and investment
purposes. Normally,
Model people do not plan to default on loans. However, conditions that lead
to defaults do occur. Defaults may be triggered by a number of events,
including unemployment, divorce, or death. These events are not likely to
trigger defaults if the owner has positive equity in his/ her home because
the sale of the home with realization of a profit is better than the loss of
the home through foreclosure. However, if the property is worth less than
the mortgage, these events may trigger defaults.
Prepayments of home mortgages can also occur. These may be triggered by
events such as declining interest rates, which prompts refinancing, and
rising house prices, which prompts the take out of accumulated equity or the
sale of the residence. Because FHA mortgages are assumable, the sale of a
residence does not automatically trigger prepayment. For example, if
interest rates have risen substantially since the time that the mortgage was
originated, a new purchaser may prefer to assume the seller's mortgage. We
hypothesized that foreclosure behavior is influenced by, among other things,
the (1) level of unemployment, (2) size of the loan, (3) value of the home,
(4) current interest rates, (5) contract interest rates, (6) home equity,
and (7) region of the country within which the home is located. We
hypothesized that prepayment behavior is influenced by, among other
things, the (1) difference between the interest rate specified in the
mortgage contract and the mortgage rates generally prevailing in each
subsequent year, (2) amount of accumulated equity, (3) size of the loan, and
(4) region of the country in which the home is located.
Our first regression model estimated conditional mortgage foreclosure
probabilities as a function of a variety of explanatory variables. In this
regression, the dependent variable is a 0/ 1 indicator of whether a given
loan was foreclosed in a given year. The outstanding mortgage balance,
expressed in inflation- adjusted dollars, weighted each loan- year
observation.
Our foreclosure rates were conditional on whether the loan survives an
additional year. We estimated conditional foreclosures in a logistic
regression equation. Logistic regression is commonly used when the
variable to be estimated is the probability that an event, such as a loan's
foreclosure, will occur. We regressed the dependent variable (whose value is
1 if foreclosure occurs and 0 otherwise) on the explanatory variables
previously listed.
Our second regression model estimated conditional prepayment probabilities.
The independent variables included a measure that is based on the
relationship between the current mortgage interest rate and the contract
rate, the primary determinant of a mortgage's refinance activity. We further
separated this variable between ratios above and below 1 to allow for the
possibility of different marginal impacts in higher and lower
ranges. The variables that we used to predict foreclosures and prepayments
fall into two general categories: descriptions of states of the economy and
characteristics of the loan. In choosing explanatory variables, we relied on
the results of our own and others' previous efforts to model foreclosure and
prepayment probabilities and on implications drawn from economic
principles. We allowed for many of the same variables to affect both
foreclosure and prepayment. Equity The single most important determinant of
a loan's foreclosure is the borrower's equity in the property, which changes
over time because (1) payments reduce the amount owed on the mortgage and
(2) property values can increase or decrease. Equity is a measure of the
current value of a property compared with the current value of the mortgage
on that property. Previous research strongly indicates that borrowers with
small amounts of equity, or even negative equity, are more likely than other
borrowers to default. 4
We computed the percentage of equity as 1 minus the ratio of the present
value of the loan balance evaluated at the current mortgage interest rate,
to the current estimated house price. For example, if the current estimated
house price is $100, 000, and the value of the mortgage at the current 4
When we discuss the likely effects of one of our explanatory variables, we
are describing the marginal effects of that variable, while holding the
effects of other variables constant.
interest rate is $80,000, then equity is .2 (20 percent), or 1-( 80/ 100).
To measure equity, we calculated the value of the mortgage as the present
value of the remaining mortgage, evaluated at the current year's fixed- rate
mortgage interest rate. We calculated the value of a property by multiplying
the value of that property at the time of the loan's origination by the
change in the state's median nominal house price, adjusted for quality
changes, between the year of origination and the current year. 5 Because the
effects on foreclosure of small changes in equity may differ depending on
whether the level of equity is large or small, we used a pair of equity
variables, LAGEQHIGH and LAGEQLOW, 6 in our foreclosure regression.
The effect of equity is lagged 1 year, as we are predicting the time of
foreclosure, which usually occurs many months after a loan first defaults.
We anticipated that higher levels of equity would be associated with an
increased likelihood of prepayment. Borrowers with substantial equity in
their home may be more interested in prepaying their existing mortgage
and taking out a larger one to obtain cash for other purposes. Borrowers
with little or no equity may be less likely to prepay because they may have
to take money from other savings to pay off their loan and cover transaction
costs.
For the prepayment regression, we used a variable that measures book equity-
the estimated property value less the amortized balance of the loan- instead
of market equity. It is book value, not market value, that the
borrower must pay to retire the debt. 7 Additionally, the important effect
of interest rate changes on prepayment is captured by two other equity
variables, RELEQHI and RELEQLO, which are sensitive to the difference
between a loan's contract rate and the interest rate on 30- year mortgages
available in the current year. These variables are described below. 5 The
estimated rate of appreciation in nominal median house prices, obtained from
Standard & Poor's DRI, was revised downward by 2 percentage points per year
to account for depreciation and the gradual improvement in the quality of
the existing housing stock over time. Also, to ensure that our estimates
were conservative, we subtracted an additional 1 percent annually from
Standard & Poor's DRI's forecasts. 6 Essentially, LAGEQHIGH takes the value
of equity minus .2 if equity is greater than 20 percent or 0 if equity is
less than or equal to 20 percent. LAGEQLOW takes the value of
equity if equity is 20 percent or less and .2 if equity is greater than 20
percent. 7 Similarly, for foreclosures within the ARM equations, we defined
equity as book equity (the estimated property value less the amortized
balance of the loan) and not market equity. The effects of interest rate
changes in the ARM equations were estimated using a separate variable.
LTV Ratio We included an additional set of variables in our regressions
related to equity: the initial LTV ratio. We entered LTV as a series of
dummy variables, depending on its size. Loans fit into eight discrete LTV
categories. In some years, FHA measured LTV as the loan amount less mortgage
insurance premium financed in the numerator of the ratio and appraised value
plus closing costs in the denominator. To reflect true economic LTV, we
adjusted FHA's measure by removing closing costs from the denominator and
including financed premiums in the numerator. A borrower's initial equity
can be expressed as a function of LTV, so we anticipated that if LTV was an
important predictor in an equation that also includes a variable measuring
current equity, it would probably be positively related to the probability
of foreclosure. One reason for including LTV is that it measures initial
equity accurately. Our measures of current equity are less accurate because
we do not have data on the actual rate of change in the mortgage loan
balance or the actual rate of house
price change for a specific house. Loans with higher LTVs are more likely to
foreclose. For the long- term nonrefinanced equation, the ARM equation, and
the short- term equation, we deleted the lower category of LTV loans. We
expected LTV to have a positive sign in the foreclosure equations at higher
levels of LTV. LTV in our foreclosure equations may capture the effects of
income constraints. We were unable to include borrowers' income or payment-
to- income ratio
directly because data on borrowers' income were not available. 8 However, it
seems likely that borrowers with little or no down payment (high LTV) are
more likely to be financially stretched in meeting their payments and,
therefore, more likely to default. The anticipated relationship between LTV
and the probability of prepayment is uncertain.
For some loan type categories, we used down payment information directly,
rather than the series of LTV variables. We defined down payment to ensure
that closing costs were included in the loan amount and excluded
from the house price. 8 We also did not know whether individual borrowers
had subsequently acquired a second mortgage or other obligations that would
affect prepayment or foreclosure probabilities.
Unemployment We used the annual unemployment rates for each state for the
period from fiscal years 1975 through 1999 to measure the relative condition
of the economy in the state where a loan was made. We anticipated that
foreclosures would be higher in years and states with higher unemployment
rates and that prepayments would be lower because property sales slow down
during recessions. The actual variable we used in our regressions, LAGUNEMP,
is defined as the logarithm of the
preceding year's unemployment rate in that state. Interest Rates We included
the logarithm of the interest rate on the mortgage as an explanatory
variable in the foreclosure equation. We expected a higher interest rate to
be associated with a higher probability of foreclosure
because high interest rates cause a higher monthly payment. However, in
explaining the likelihood of prepayment, our model uses information on the
level of current mortgage rates relative to the contract rate on the
borrower's mortgage. A borrower's incentive to prepay is high when the
interest rate on a loan is greater than the rate at which money can now be
borrowed, and it diminishes as current interest rates increase. In our
prepayment regression, we defined two variables, RELEQHI and RELEQLO.
RELEQHI is defined as the ratio of the market value of the mortgage to the
book value of the mortgage but is never smaller than 1. RELEQLO is also
defined as the ratio of the market value of the mortgage
to the book value but is never larger than 1. When currently available
mortgage rates are lower than the contract interest rate, market equity
exceeds book equity because the present value of the remaining payments
evaluated at the current rate exceeds the present value of the remaining
payments evaluated at the contract rate. Thus, RELEQHI captures a borrower's
incentive to refinance, and RELEQLO captures a new buyer's
incentive to assume the seller's mortgage. We created two 0/ 1 variables,
REFIN and REFIN2, that take on a value of 1 if a borrower had not taken
advantage of a refinancing opportunity in the past and 0 otherwise. We
defined a refinancing opportunity as having
occurred if the interest rate on fixed- rate mortgages in any previous year
in which a loan was active was at least 200 basis points 9 below the rate on
the mortgage in any year up through 1994 or 150 basis points below the rate
on 9 A basis point equals 1/ 100 of a percentage point.
the mortgage in any year after 1994. 10 REFIN takes a value of 1 if the
borrower had passed up a refinancing opportunity at least once in the past.
REFIN2 takes on a value of 1 if the borrower had passed up two or more
refinancing opportunities in the past.
Several reasons might explain why borrowers passed up apparently profitable
refinancing opportunities. For example, if they had been unemployed or their
property had fallen in value they might have had difficulty obtaining
refinancing. This reasoning suggests that REFIN and REFIN2 would be
positively related to the probability of foreclosure; that is, a borrower
unable to obtain refinancing previously because of poor
financial status might be more likely to default. Similar reasoning suggests
a negative relationship between REFIN and REFIN2 and the probability of
prepayment; a borrower unable to obtain refinancing previously might also be
unlikely to obtain refinancing currently. A negative relationship might also
exist if a borrower's passing up one profitable refinancing opportunity
reflected a lack of financial sophistication that, in turn, would be
associated with passing up additional opportunities. However, a borrower who
anticipated moving soon might pass up an apparently profitable refinancing
opportunity to avoid the
transaction costs associated with refinancing. In this case, there might be
a positive relationship, with the probability of prepayment being higher if
the borrower fulfilled his/ her anticipation and moved, thereby prepaying
the loan.
Another explanatory variable is the volatility of interest rates, INTVOL,
which is defined as the standard deviation of the monthly average of the
Federal Home Loan Mortgage Corporation's series of 30- year, fixed- rate
mortgage effective interest rates. We calculated the standard deviation
over the previous 12 months. Financial theory predicts that borrowers are
likely to refinance more slowly at times of volatile rates because there is
a larger incentive to wait for a still- lower interest rate.
We also included the slope of the yield curve, YC, in our prepayment
estimates, which we calculated as the difference between the 1- and 10year
Treasury rates of interest. We then subtracted 250 basis points from this
difference and set differences that were less than 0 to 0. This variable
10 Transaction costs associated with refinancing have fallen in recent
years, making it more profitable than before to refinance at a smaller
decrease in interest rates.
measured the relative attractiveness of ARMs versus fixed- rate mortgages;
the steeper the yield curve, the more attractive ARMs would be. When ARMs
have low rates, borrowers with fixed- rate mortgages may be induced into
refinancing into ARMs to lower their monthly payments.
For ARMs, we did not use relative equity variables as we did with fixed-
rate mortgages. Instead, we defined four variables, CHANGEPOS, CHANGENEG,
CAPPEDPOS, and CAPPEDNEG, to capture the relationship between current
interest rates and the interest rate paid on each mortgage. CHANGEPOS
measures how far the interest rate on the mortgage has increased since
origination, with a minimum of 0, while CHANGENEG measures how far the rate
has decreased, with a maximum
of 0. CAPPEDPOS measures how much farther the interest rate on the mortgage
would rise, if prevailing interest rates in the market did not change, while
CAPPEDNEG measures how much farther the mortgage's
rate would fall, if prevailing interest rates did not change. For example,
if an ARM was originated at 7 percent and interest rates increased by 250
basis points 1 year later, CHANGEPOS would equal 100 because FHA's ARMs can
increase by no more than 100 basis points in a year. CAPPEDPOS would equal
150 basis points, since the mortgage rate would eventually increase by
another 150 basis points if market interest rates did not change, and
CHANGENEG and CAPPEDNEG would equal 0. Because interest rates have generally
trended downwards since FHA introduced ARMs, there is very little experience
with ARMs in an increasing interest rate environment.
Geographic Regions We created nine 0/ 1 variables to reflect the geographic
distribution of FHA loans and included them in both regressions. Location
differences may capture the effects of differences in borrowers' income,
underwriting
standards by lenders, economic conditions not captured by the unemployment
rate, or other factors that may affect foreclosure and prepayment rates. We
assigned each loan to one of the nine Bureau of the Census (Census)
divisions on the basis of the state in which the borrower resided. The
Pacific division was the omitted category; that is, the
regression coefficients show how each of the regions was different from the
Pacific division. We also created a variable, JUDICIAL, to indicate states
that allowed judicial foreclosure procedures in place of nonjudicial
foreclosures. We anticipated that the probability of foreclosure would be
lower where judicial foreclosure procedures were allowed because of the
greater time and expense required for the lender to foreclose on a loan.
Loan Size To obtain an insight into the differential effect of relatively
larger loans on mortgage foreclosures and prepayments, we assigned each loan
to 1 of 10 loan- size categorical variables (LOAN1 to LOAN10). The omitted
category
in our regressions was loans between $80,000 and $90,000, and results on
loan size are relative to those loans between $80,000 and $90, 000. All
dollar amounts are inflation- adjusted and represent 1999 dollars. Number of
Units The number of units covered by a single mortgage was a key determinate
in deciding which loans were more likely to be investor loans. Loans were
noted as investor loans if the LTV ratio was between specific values,
depending on the year of the loan, or if there were two or more units
covered by the loan. Once a loan was identified as an investor loan, we
separated the refinanced loans from the purchase money mortgages and
performed foreclosure and payoff analyses on each. For each of the investor
equations, we used two dummy variables defined according to the number of
units in the dwelling. LIVUNT2 has the value of 1 when a property has two
dwelling units and a value of 0 otherwise. LIVUNT3 has a value of 1 when a
property has three or more dwelling units and a value of 0 otherwise. The
missing category in our regressions was investors with one
unit. Our database covers only loans with no more than four units. Policy
Year and Refinance
To capture the time pattern of foreclosures and prepayments (given the
Indicator
effects of equity and the other explanatory variables), we defined seven
variables on the basis of the number of years that had passed since the year
of the loan's origination. We refer to these variables as YEAR1 to YEAR7 and
set them equal to 1 during the corresponding policy year and 0 otherwise.
Finally, for those loan type categories for which we did not estimate
separate models for refinancing loans and nonrefinancing loans, we created a
variable called REFINANCE DUMMY to indicate whether a loan was a refinancing
loan.
Table 4 summarizes the variables that we used to predict foreclosures and
prepayments. Table 5 presents mean values for our predictor variables for
each mortgage type for which we ran a separate regression.
Table 4: Variable Names and Descriptions Variable name Variable description
Loan size dummy variables
LOAN1 1 if loan amount is less than $40,000, else 0 LOAN2 1 if loan amount
is $40, 000 or above but below $50,000, else 0 LOAN3 1 if loan amount is
$50, 000 or above but below $60,000, else 0 LOAN4 1 if loan amount is $60,
000 or above but below $70,000, else 0 LOAN5 1 if loan amount is $70, 000 or
above but below $80,000, else 0 LOAN6 1 if loan amount is $80, 000 or above
but below $90,000, else 0 LOAN7 1 if loan amount is $90, 000 or above but
below $100,000, else 0 LOAN8 1 if loan amount is $100,000 or above but below
$110, 000, else 0 LOAN9 1 if loan amount is $110,000 or above but below
$130, 000, else 0 LOAN10 1 if loan amount is at least $130, 000, else 0
Economic variables
LOGINT Log of the contract mortgage interest rate REFINANCE DUMMY 1 if the
loan is a refinancing loan, else 0 RELEQLO The ratio of the market value of
the mortgage to the book value if the market value is below the book value,
else 1
RELEQHI The ratio of the market value of the mortgage to the book value if
the market value is above the book value, else 1 REFIN 1 if, in at least 1
previous year, the mortgage interest rate had been at least
200 basis points below the contract rate in any year prior to 1995 or 150
basis points below the contract rate after 1994 and the borrower had not
refinanced, else 0
REFIN2 1 if, in at least 2 previous years, the above situation prevailed,
else 0 INTVOL The volatility of mortgage rates, defined as the standard
deviation of 30- year
fixed- rate mortgage interest rates over the previous 12 months YC The slope
of the yield curve, defined as the difference between 1- and 10year U. S.
Treasury interest rates minus 250 basis points, but not less than 0
LIVUNT2 1 if the property has two housing units, else 0 LIVUNT3 1 if the
property has three or more housing units, else 0 LAGUNEMP The log of the
previous year's unemployment rate in each state JUDICIAL 1 if state allowed
judicial foreclosure (list of states varies by year), else 0
Policy year dummy variables YEAR1 1 if in loan's first year, else 0 YEAR2 1
if in loan's second year, else 0 YEAR3 1 if in loan's third year, else 0
YEAR4 1 if in loan's fourth year, else 0 YEAR5 1 if in loan's fifth year,
else 0
(Continued From Previous Page)
Variable name Variable description
YEAR6 1 if in loan's sixth year, else 0 YEAR7 1 if in loan's seventh year,
else 0
Loan- to- value dummy variables
LTV0 1 if LTV equals 0, assumed missing data, else 0 LTV1 1 if LTV is above
0 and less than 60, else 0 LTV2 1 if LTV is greater than or equal to 60, but
less than 85, else 0 LTV3 1 if LTV is greater than or equal to 85, but less
than 92, else 0 LTV4 1 if LTV is greater than or equal to 92, but less than
96, else 0 LTV5 1 if LTV is greater than or equal to 96, but less than 98,
else 0 LTV6 1 if LTV is greater than or equal to 98, but less than 100, else
0 LTV7 1 if LTV is greater than or equal to 100, but less than 102, else 0
Equity variables
LAGEQLOW The lagged value of market equity (defined as 1 minus the ratio of
the present value of the loan balance, evaluated at the current mortgage
interest rate, to the current estimated house price) if equity is less than
or equal to 20 percent, else .2
LAGEQHIGH The lagged value of market equity (defined as 1 minus the ratio of
the present value of the loan balance, evaluated at the current mortgage
interest rate, to the current estimated house price minus .2) if equity is
greater than 20 percent, else 0
BOOKNEG The lagged value of book equity (defined as 1 minus the ratio of the
amortized loan balance to the current estimated house price) if equity is
less than or equal to 20 percent, else .2 BOOKPOS The lagged value of book
equity (defined as 1 minus the ratio of the amortized loan balance to the
current estimated house price minus .2), if
equity is greater than 20 percent, else 0 CHANGEPOS The amount by which the
interest rate of an ARM has increased since
origination, with a minimum of 0 CHANGENEG The amount by which the interest
rate of an ARM has decreased since
origination, with a maximum of 0 CAPPEDPOS The amount by which the interest
rate of an ARM could still rise, if prevailing
interest rates in the market did not change, with a minimum of 0 CAPPEDNEG
The amount by which the interest rate of an ARM could still decline, if
prevailing interest rates in the market did not change, with a maximum of 0
DOWNPAY The down payment, expressed as a percentage of the purchase price of
the house. Closing costs were excluded from the house price and included in
the loan amount
Census division dummy variables
DV_ A a 1 if the loan is in the Mid- Atlantic states (NY, PA, NJ), else 0
DV_ E 1 if the loan is in the east south central states (KY, TN, AL, MS),
else 0 DV_ G 1 if the loan is in the west north central states (MN, MO, IA,
NB, KS, SD, ND),
else 0
(Continued From Previous Page)
Variable name Variable description
DV_ M 1 if the loan is in the Mountain states (CO, UT, AZ, NM, NV, ID, WY,
MT), else 0 DV_ N 1 if the loan is in the New England states (MA, CT, RI,
NH, ME, VT), else 0 DV_ R 1 if the loan is in the east north central states
(IL, MI, OH, IN, WI), else 0 DV_ S 1 if the loan is in the South Atlantic
states (GA, NC, SC, VA, MD, DC, DE,
WV), else 0 DV_ W 1 if the loan is in the west south central states (TX, OK,
LA, AR), else 0
a DV = Division. Table 5: Means of Predictor Variables Loan type Predictor
Long- term Short- term
Investor variable name Long- term FRM a FRM refinance FRM ARM a Investor
refinance
Loan size dummy variables
LOAN1 0.0635 0.0160 0.1531 0. 0044 0. 0617 0.0267 LOAN2 0.0860 0.0355 0.1253
0. 0148 0. 0789 0.0504 LOAN3 0. 1197 0.0648 0.1369 0. 0336 0. 1038 0.0829
LOAN4 0.1338 0.1043 0.1391 0. 0577 0. 1165 0.1175 LOAN5 0.1292 0.1294 0.1216
0. 0849 0. 1253 0.1337 LOAN6 0. 1130 0.1434 0.1034 0. 1087 0. 1189 0.1293
LOAN7 0. 0980 0.1362 0.0774 0. 1231 0. 1102 0.1102 LOAN8 0. 0862 0.1171
0.0575 0. 1228 0. 0886 0.0858 LOAN9 0. 1033 0.1461 0.0582 0. 2032 0. 1070
0.1272 LOAN10 0.0673 0.1070 0.0276 0. 2468 0. 0891 0.1363
Economic variables
LOGINT -2.3773 -2.4890 -2.4165 -2. 6864 -2. 3025 -2.5168 REFINANCE DUMMY - -
0. 3650 0. 1088 - RELEQHI 1.0613 1.0535 1.0286 - 1. 0793 1. 0461 RELEQLO
0.9450 0.9855 0.9745 - 0. 9601 0. 9845 REFIN 0.1174 0.0506 0.1082 0. 0734 0.
1892 0.0453 REFIN2 0.0779 0.0237 0.0734 0. 0281 0. 1302 0.0240 LIVUNT2 - - -
- 0. 3085 0.2396 LIVUNT3 - - - - 0. 0859 0.0944
(Continued From Previous Page)
Loan type Predictor
Long- term Short- term
Investor variable name Long- term FRM a FRM refinance FRM ARM a Investor
refinance
LAGUNEMP -2.8063 -2.8870 -2.8379 -2. 9155 -2. 7788 -2.9044
Policy year dummy variables
YEAR1 0. 1514 0.2527 0.1634 0. 2453 0. 1363 0.2729 YEAR2 0. 1384 0.1993
0.1509 0. 2299 0. 1299 0.2048 YEAR3 0. 1227 0.1501 0.1354 0. 1820 0. 1185
0.1495 YEAR4 0. 1054 0.1285 0.1210 0. 1206 0. 1043 0.1228 YEAR5 0. 0873
0.1059 0.1047 0. 0850 0. 0900 0.0972 YEAR6 0. 0734 0.0903 0.0911 0. 0597 0.
0784 0.0817 YEAR7 0. 0603 0.0431 0.0628 0. 0337 0. 0672 0.0435
Loan- to- value dummy variables
LTV0 0.0327 0.7301 0.2551 0. 0555 0. 0099 0.2330 LTV1 0.0097 0.0067 0.0620
0. 0017 0. 0054 0.0064 LTV2 0.0842 0.0621 0.2366 0. 0517 0. 2321 0.6263 LTV3
0.0976 0.0892 0.0970 0. 1231 0. 5280 0.1027 LTV4 0.2038 0.0654 0.1146 0.
2898 0. 0689 0.0158 LTV5 0.1937 0.0326 0.0770 0. 3673 0. 0526 0.0093 LTV6
0.1683 0.0025 0.0855 0. 0464 0. 0463 0.0009 LTV7 0.1476 0.0081 0.0576 0.
0518 0. 0411 0.0036 LTV8 0.0624 0.0034 0.0146 0. 0126 0. 0157 0.0019
Equity variables
LAGEQLOW 0.1226 0.0718 0.1649 - 0. 1518 0. 1429 LAGEQHIGH 0. 1049 0.0131
0.1470 - 0. 1068 0. 0412 BOOKNEG 0. 1312 0.0943 0.1628 0. 1142 0. 1652
0.1628 BOOKPOS 0. 0903 0.0153 0.1438 0. 0112 0. 1063 0.0386 CHANGEPOS - - -
0. 8844 - CHANGENEG - - - -0. 4940 - CAPPEDPOS - - - 0. 1462 - CAPPEDNEG - -
- -0. 0860 - DOWNPAY - 0.0337 - - - 0. 1173
Census division dummy variables
DV_ A a 0.0725 0.0536 0.0757 0. 0578 0. 1579 0.1181 DV_ E 0.0737 0.0394
0.0890 0. 0454 0. 0499 0.0713 DV_ G 0.0874 0.1063 0.1242 0. 1036 0. 0700
0.0796 DV_ M 0.1344 0.1932 0.1286 0. 1377 0. 1418 0.1708 DV_ N 0.0098 0.0137
0.0077 0. 0320 0. 0274 0.0294
(Continued From Previous Page)
Loan type Predictor
Long- term Short- term
Investor variable name Long- term FRM a FRM refinance FRM ARM a Investor
refinance
DV_ P 0.1425 0.1589 0.0726 0. 2050 0. 1471 0.1305 DV_ R 0.1164 0.0822 0.1315
0. 1711 0. 1300 0.1362 DV_ S 0.2060 0.2152 0.1575 0. 2043 0. 1706 0.1792 DV_
W 0.1574 0.1375 0.2132 0. 0432 0. 1054 0.0848
a ARM = Adjustable rate mortgage; DV = Division; FRM = Fixed- rate mortgage.
Estimation Results As previously described, we used logistic regressions to
model loan foreclosures and prepayments as a function of a variety of
predictor variables. We estimated separate regressions for fixed- rate
purchase money mortgages (and refinanced loans) with terms over and under 25
years, ARMs, and investor loans. We used data on loan activity throughout
the life of the loans for loans originated from fiscal years 1975 through
1999. The outstanding loan balance of the observation weighted the
regressions. The logistic regressions estimated the probability of a loan
being
foreclosed or prepaid in each year. The standard errors of the regression
coefficients are biased downward because the errors in the regressions are
not independent. The observations are on loan years, and the error terms are
correlated because the same underlying loan can appear several times.
However, we did not view this downward bias as a problem because our purpose
was to forecast the dependent variables, not to test hypotheses concerning
the effects of independent variables.
In general, our results are consistent with the economic reasoning that
underlies our models. Most importantly, the probability of foreclosure
declines as equity increases, and the probability of prepayment increases as
the current mortgage interest rate falls below the contract mortgage
interest rate. As shown in tables 6 and 7, both of these effects occur in
each regression model and are very strong. These tables present the
estimated coefficients for all of the predictor variables for the
foreclosure and prepayment equations.
Table 6 shows our foreclosure regression results. As expected, the
unemployment rate is positively related to the probability of foreclosure
and negatively related to the probability of prepayment. Our results also
indicate that generally the probability of foreclosure is higher when LTV
and contract interest rate are higher. The overall goodness of fit was
satisfactory: Chi- Square statistics were significant on all regressions at
the 0.01- percent level.
Because the coefficients from a nonlinear regression can be difficult to
interpret, we transformed some of the coefficients for the long- term,
nonrefinanced, fixed- rate regressions into statements about changes in the
probabilities of foreclosure and prepayment. Overall conditional foreclosure
probabilities for this mortgage type are estimated to be about 0.5 percent.
11, 12 In other words, on average, there is a � of a 1- percent chance for a
loan of this type to result in a claim payment in any particular year. 13 By
holding other predictor variables at their mean values, we can describe the
effect on the conditional foreclosure probability of changes in the values
of predictor variables of interest. For example, if the average value of the
unemployment rate were to increase by 1 percentage point from its mean value
(in our sample) of about 6 percent to about 7 percent, the conditional
foreclosure probability would increase by about 20 percent (from 0. 5
percent to about 0.6 percent). Similarly, a 1- percentage- point increase in
the mortgage contract rate from its mean value of about 9.25 to about 10.25
would also raise the conditional foreclosure probability by 20
percent (from about 0. 5 percent to about 0.6 percent). Values of
homeowners' equity of 10 percent, 20 percent, 30 percent, and 40 percent
result in conditional foreclosure probabilities of 0.8 percent, 0. 7
percent, 0.5 percent, and 0. 3 percent, respectively, illustrating the
importance of
increased equity in reducing the probability of foreclosure. Table 7 shows
our prepayment regression results. The overall conditional prepayment
probability for long- term, fixed- rate mortgages is estimated to be 4.8
percent. This means that, in any particular year, about 5 percent of 11 The
conditional foreclosure probability is calculated as F( Z) = EXP( Z)/[ 1+
EXP( Z)], where Z = (X *B ), where X refers to the mean value of the i th
explanatory variable and the
i i i B i s are the estimated coefficients. 12 Conditional foreclosure
probabilities for the other mortgage types were estimated as follows: long-
term, fixed- rate, refinancing mortgages (0. 3); short- term, fixed- rate
mortgages (0. 2); ARMs (0. 3); investor, nonrefinancing mortgages (0. 6);
and investor, refinancing mortgages (0.2). 13 This average is for the dollar
worth of a loan, not the number of loans.
the loan dollars outstanding will prepay, on average. 14 Prepayment
probability is quite sensitive to the relationship between the contract
interest rate and the currently available mortgage rates. We modeled this
relationship using RELEQHI and RELEQLO. Holding other variables at their
mean values, if the spread between mortgage rates available in each
year and the contract interest rate widened by one percentage point, the
conditional prepayment probability would increase by about 80 percent to 8.6
percent. Table 6: Coefficients From Foreclosure Equations and Summary
Statistics Loan type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
INTERCEPT 2. 8424 8.4917 5.1918 5.7426 4.4798 6. 1160
Loan size dummy variables
LOAN1 0.4559 0.2015 0.6280 0.3699 0.2301 0. 1214 LOAN2 0.2089 0.3401 0.3577
0.5417 0.1367 0. 4806 LOAN3 0. 1238 0.1857 0.2741 0.4112 0.0957 0. 0825
LOAN4 0.0630 0.0670 0.1755 0.1728 0.0019 -0.1074 LOAN5 -0.0077 0.1196
-0.0346 0.0389 0.1093 0. 1611 LOAN7 0. 0533 0.0941 0.0829 -0.0570 -0.0522
-0.0088 LOAN8 -0.0116 0.2417 0.0100 -0.0213 -0.0226 0. 3838 LOAN9 0. 0499
0.3755 0.3107 -0.0142 -0.1076 0. 2515 LOAN10 0.1640 0.6596 0.2550 -0.0841
0.1166 0. 5903
Economic variables
LOGINT 1.7604 2.8065 3.3779 1.3074 2.5642 2. 3067 REFINANCE DUMMY - - -0.
1033 0. 0806 - REFIN 0. 2757 0.1427 -0.0690 - 0. 3024 0. 2010 REFIN2 -0.0467
- -0.1968 - 0. 1159 LIVUNT2 ---- 0. 2031- 0. 6954 LIVUNT3 ---- 0. 3245- 0.
5352 LAGUNEMP 1.1322 1.8867 1.2634 1.1528 1.1581 1. 3849
14 Conditional prepayment probabilities for the other mortgage types were
estimated as follows: long- term, fixed- rate, refinancing mortgages (7. 7);
short- term, fixed- rate mortgages (5. 9); ARMs (8. 2); investor,
nonrefinancing mortgages (5. 3); and investor, refinancing mortgages (6.6).
(Continued From Previous Page)
Loan type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
INTVOL 0.0661 -0.2651 0.2815 -1.8350 0.1054 -0.4593
Policy year dummy variables
YEAR1 -3.6549 -4.2892 -3.9423 -5.1043 -3.4559 -3.7991 YEAR2 -1.1583 -1.5378
-1.6170 -1.6688 -0.6838 -1.5670 YEAR3 -0.1734 -0.3399 -0.5061 -0.3611 0.2215
-0.5058 YEAR4 0. 1139 -0.0013 -0.2101 0.1190 0.3735 -0.1317 YEAR5 0. 2077
0.1734 -0.1072 0.3006 0.4455 -0.1305 YEAR6 0. 1868 0.0751 -0.0328 0.3101
0.4040 0. 1324 YEAR7 0. 1154 -0.0667 -0.0601 0.0950 0.2865 0. 0287
Loan- to- value dummy variables
LTV0 0.5864 -0.0349 1.3391 -0.7377 0.0208 0. 2826 LTV1 ---- 0. 2383 LTV2
-0.0521 - 0. 9851 -0.7269 0.3093 LTV3 0.2225 - 1. 2587 -1.0359 0.3037 LTV4
0.3585 - 1. 5278 -0.8733 0.2516 LTV5 0.4427 - 1. 7433 -0.8449 0.1997 LTV6
0.4746 - 1. 8241 -0.8075 0.2456 LTV7 0.4290 - 1. 7740 -1.0976 0.3345 LTV8
0.4634 - 1. 4832 -1.1230 - Equity
variables
DOWNPAY - 0.1062 - - - -0.3661 LAGEQLOW -1.5913 -1.7065 -1.3388 - -1.7707
-0.9796 LAGEQHIGH -3.9061 -4.8967 -3.4301 - -4.0435 -8.4771 BOOKNEG ---- 3.
3058- BOOKPOS ---- 7. 8496- CHANGENEG - - - -0. 1630 - CHANGEPOS - - - -0.
2325 - CAPPEDNEG - - - 0.4445 - CAPPEDPOS - - - 0.0120 - Census
division dummy variables
DV_ A a -0.0961 -0.6324 0.0873 -0.7830 -0.3807 0. 1808 DV_ E -0.2442 -0.6969
0.0282 -1.0519 0.1646 -0.3263 DV_ G 0.0559 -0.4792 0.1847 -1.1353 0.2876
-0.3948 DV_ M 0.3434 -0.4969 0.4345 -0.7531 0.6335 -0.2241 DV_ N 0.2225
0.1393 1.0141 -0.2386 0.4703 1. 0258
(Continued From Previous Page)
Loan type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
DV_ R -0.0896 -0.7246 -0.0257 -0.6189 -0.1482 -0.5150 DV_ S 0.1450 -0.3786
0.1830 -0.4121 0.3052 -0.1798 DV_ W 0.3436 -0.3450 0.5660 -0.8390 0.7350
-0.2109 JUDICIAL -0.1350 -0.2168 -0.0671 -0.3836 -0.1600 -0.6095
Summary statistics
Percentage of concordant pairs 76. 1 82. 6 78. 2 80. 0 78. 4 83. 0
Percentage of tied pairs 3. 6 3.3 9. 1 3.1 2. 7 4. 4 Number of unweighted
1,526,825 503,253 498,723 473,573 660, 253 52, 715 observations a ARM =
Adjustable rate mortgage; DV = Division; FRM = Fixed- rate mortgage.
Table 7: Coefficients From Prepayment Equations and Summary Statistics Loan
type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
INTERCEPT -16. 4264 -18.1645 -24.7363 -5.2631 -16.9717 -16.4240
Loan size dummy variables
LOAN1 -0.4966 -0.7217 -0.5502 -0.5199 -0.6714 -0.5439 LOAN2 -0.3867 -0.5551
-0.3690 -0.4400 -0.4162 -0.2566 LOAN3 -0.2690 -0.3367 -0.2361 -0.3412
-0.2607 -0.1642 LOAN4 -0.1513 -0.2250 -0.1493 -0.2173 -0.1544 -0.0420 LOAN5
-0.0965 -0.1058 -0.0535 -0.1171 -0.0624 -0.0518 LOAN7 0. 0786 0.0739 0.0717
0.1057 0.0866 0. 0984 LOAN8 0. 1645 0.1680 0.1333 0.1553 0.1433 0. 1179
LOAN9 0. 2682 0.2349 0.2028 0.2360 0.2433 0. 2469 LOAN10 0.3601 0.3576
0.2155 0.3823 0.3287 0. 3947
Economic variables
JUDICIAL - - - 0. 0834 - RELEQHI 8.6248 10. 6524 10. 8280 - 6. 5717 9. 7379
RELEQLO 4.0974 0.5224 9.9514 - 6. 8953 1. 1596 REFINANCE DUMMY - - 0.1667
0.0511 - -
(Continued From Previous Page)
Loan type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
REFIN -0.4126 -0.7985 -0.4159 - -0.2148 -0.6270 REFIN2 -0.9690 - -0.4800 -
-0.8893 LIVUNT2 ----- 0. 2852- 0. 3853 LIVUNT3 ----- 0. 3735- 0. 3973
LAGUNEMP -0.3093 -0.9917 -0.3465 -1.3883 -0.3207 -0.6892 INTVOL -0.1532
-0.9027 0.0953 -1.4628 0.2270 -0.4564 YC 0.6490 0.9190 1.0536 -1.9892 0.9197
0. 7492
Policy year dummy variables
YEAR1 -1.8664 0.0919 -1.3950 -1.7499 -1.5846 -0.0012 YEAR2 -0.3667 0.8526
-0.2351 -0.0641 -0.3511 0. 8935 YEAR3 0. 1074 0.9118 0.0214 0.2971 -0.0011
0. 9970 YEAR4 0. 2406 0.6924 0.0387 0.2847 0.1072 0. 7482 YEAR5 0. 1514
0.4692 0.0234 0.3413 0.0246 0. 7789 YEAR6 0. 0957 0.4168 -0.0102 0.2567
-0.0519 0. 6648 YEAR7 0. 2151 0.2470 0.0501 0.2376 0.2114 0. 4584
Loan- to- value dummy variables
LTV0 0.0623 0.8115 0.2864 0.4611 -0.3276 0. 0329 LTV1 -0.1218 - 0. 0824 -
-0.2461 LTV2 -0.0816 0.3427 0.2249 -0.2133 -0.3234 LTV3 -0.0370 0.4700
0.2172 -0.1507 -0.0841 LTV4 -0.0471 0.4921 0.2534 -0.1352 -0.2134 LTV5
-0.1297 0.4100 0.1665 -0.1460 -0.2729 LTV6 -0.1679 0.9448 0.1273 -0.4552
-0.3270 LTV7 -0.2416 0.7392 0.0910 -0.5858 -0.3709 LTV8 - 0. 7018 - -0.4581
- Equity
variables
BOOKNEG 1.3572 1.1104 0.8176 2.2369 1.2333 1. 1330 BOOKPOS 0. 7731 1.9607
1.3009 1.1137 0.6994 3. 3660 CHANGENEG - - - 0. 1549 - CHANGEPOS - - - -0.
1754 - CAPPEDNEG - - - -0.2042 - CAPPEDPOS - - - -0.1181 - DOWNPAY ------ 1.
7804
Census division dummy variables
DV_ A a -0.4397 -0.3413 -0.3437 -0.5644 -0.2616 -0.4401
(Continued From Previous Page)
Loan type Long- term
Long- term Short- term
Investor Predictor variable name FRM a FRM refinance FRM ARM a Investor
refinance
DV_ E -0.1071 0.2771 0.1027 -0.1597 -0.0949 0. 0914 DV_ G 0.1191 0.2663
0.0900 -0.4870 0.0932 0. 1025 DV_ M 0.1169 0.4774 0.2433 0.0334 0.0118 0.
3155 DV_ N -0.2030 -0.4385 0.0817 -0.6094 -0.0892 -0.0299 DV_ R 0.0951
0.3203 0.2295 -0.2247 0.1049 0. 2867 DV_ S -0.2746 -0.0977 -0.1176 -0.5520
-0.2545 -0.1784 DV_ W -0.3377 0.0109 -0.2565 -0.4196 -0.3003 -0.0208
Summary statistics
Percentage of concordant pairs 78. 5 72. 5 73. 7 74. 1 76. 3 73. 3
Percentage of tied pairs 0. 5 0.8 0. 7 0.5 0. 6 0. 8 Number of unweighted
1,526,825 503,253 498,723 473,573 660, 253 52, 715 observations a ARM =
Adjustable rate mortgage; DV = Division; FRM = Fixed- rate mortgage.
Model Predictions for To test the validity of our model, we examined how
well the model Historical Period
predicted actual patterns of FHA's foreclosure and prepayment rates through
fiscal year 1999. Using a sample of 10 percent of FHA's loans made from
fiscal years 1975 to 1999, we found that our predicted rates closely
resembled actual rates. To predict the probabilities of foreclosure and
prepayment in the historical
period, we combined the model's coefficients with the information on a
loan's characteristics and information on economic conditions described by
our predictor variables in each year from a loan's origination through
fiscal year 1999. If our model predicted foreclosure or prepayment in any
year, we determined the loan's balance during that year to indicate the
dollar amount associated with the foreclosure or prepayment. We estimated
cumulative foreclosure and prepayment rates by summing the
predicted claim and prepayment dollar amounts for all loans originated in
each of the fiscal years 1975 through 1999. We compared these predictions
with the actual cumulative (through fiscal year 1999) foreclosure and
prepayment rates for the loans in our sample. Figure 4 compares actual and
predicted cumulative foreclosure rates, and figure 5 compares actual and
predicted cumulative prepayment rates.
Figure 4: Cumulative Foreclosure Rates by Book of Business for 30- Year,
Fixed- Rate, Nonrefinanced, Mortgages; Actual and Predicted, Fiscal Years
1975- 99 25
Percentage terminating in foreclosure 20 15 10
5 0
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Fiscal year
of origination
Predicted Actual
Figure 5: Cumulative Prepayment Rates by Book of Business, for 30- Year,
Fixed- Rate, Nonrefinanced, Mortgages: Actual and Predicted, Fiscal Years
1975- 99
90 Percentage terminating in prepayment
80 70 60 50 40 30 20 10
0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Fiscal
year of origination
Predicted Actual
Estimation of the The economic value of the Fund is defined in the Omnibus
Budget Economic Value of the
Reconciliation Act of 1990 as the “current cash available to the Fund,
plus the net present value of all future cash inflows and outflows expected
to Fund result from the outstanding mortgages in the Fund.” We
obtained information on the capital resources of the Fund from documents
used to prepare FHA's audited financial statements. These capital resources
were reported to be $14.3 billion.
To estimate the net present value of future cash flows of the Fund, we
constructed a cash flow model to estimate the five primary future outflows
and inflows of cash through 2028 resulting from the books of business
written from fiscal years 1975 through 1999. Cash flows out of the fund
from payments associated with claims on foreclosed properties, refunds of
up- front premiums on mortgages that are prepaid, and administrative
expenses for management of the program. Cash flows into the fund from income
from mortgagees' insurance premiums and from the net proceeds from the sale
of foreclosed properties. To estimate the Fund's cash flow, we first
forecasted, for active loans at the end of 1999, the dollar value of loans
predicted to foreclose or prepay in any year through 2028. From those
estimates, we derived estimates of the outstanding principal balances for
the loans remaining active for each year in the forecast period. Our cash
flow model used these estimates of foreclosure and prepayment dollars and
outstanding principal balances to derive estimates of each of the primary
cash flows.
We forecasted future loan activity (foreclosures and prepayments) on the
basis of the regression results described above and forecasts of the key
economic and housing market variables made by Standard & Poor's DRI.
Standard & Poor's DRI forecasts the median sales price of existing housing,
by state and year, through fiscal year 2005. We assumed that after 2005
those prices would rise at 3 percent per year. In creating the borrower's
equity variable, we used DRI forecasts of existing housing prices by state
and subtracted 2 percentage points per year to adjust for improvements in
the quality of housing over time and the depreciation of individual housing
units. We also subtracted another 1 percentage point per year from the
company's forecasts, to be conservative. We made similar adjustments to our
assumed value of median house price change for the years beyond the range of
these forecasts. We used DRI forecasts of each state's unemployment rate and
assumed that rates from fiscal year 2026 on would equal the rates in 2025.
We also used Standard & Poor's DRI forecasts of interest rates on 30- year
mortgages and 1- and 10- year Treasury securities.
Using the results of the econometric model, the cash flow model estimates
cash flows for each policy year through the life of a mortgage. An important
component of the model is converting all income and expense streams-
regardless of the period in which they actually occurred- into 1999 present
value dollars. We applied discount rates to match as closely as
possible the rate of return FHA likely earned in the past or would earn in
the future from its investment in Treasury securities. 15 As an 15 Actual
rates vary by the specific date on which the investment is made and the
length of maturity of the note. Precise data on the length of maturity of
FHA's investments were unavailable, but we estimated the average to be
approximately 7 years and used this estimate as the basis for our selection
of discount rates.
approximation of what FHA earned for each book of business, 16 we used a
rate of return comparable to the yield on 7- year Treasury securities
prevailing when that book was written to discount all cash flows occurring
in the first 7 years of that book's existence. We assumed that after 7
years, the Fund's investment was rolled over into new Treasury securities at
the interest rate prevailing at that time and used that rate to discount
cash
flows to the rollover date. For rollover dates occurring in fiscal year 1999
and beyond, we used 6 percent as the new discount rate. As an example, cash
flows associated with the fiscal year 1992 book of business and occurring
from fiscal years 1992 through 1998 (i. e., the first 7 policy years) were
discounted at the 7- year Treasury rate prevailing in fiscal year 1992. Cash
flows associated with the fiscal year 1992 book of business but occurring in
fiscal year 1999 and beyond are discounted at a rate of 6 percent. Our
methodology for estimating each of the five principal cash flows is
described below.
Premium Income Because FHA's premium policy has changed over time, our
calculations of premium income to the Fund change depending on the date of
the mortgage's origination. We describe all premium income, including
upfront premiums, even though they play no role in estimating the future
cash flows for the Fund at the end of fiscal year 1999.
For loans originating from fiscal years 1975 through 1983, premiums equal
the annual outstanding principal balance times 0.5 percent. For loans
originating from fiscal years 1984 through June 30, 1991, premiums equal the
original loan amount times the mortgage insurance premium. The mortgage
insurance premium during this period was equal to 3. 8 percent for 30- year
mortgages and 2.4 percent for 15- year mortgages. Because
there are no annual premiums for this group of loans, the future cash flows
would include no premium income. For the purposes of this analysis,
mortgages of other lengths of time are grouped with those they most closely
approximate. Effective July 1, 1991, FHA added an annual premium of 0.5
percent of the outstanding principal balance to its up- front
premiums. The number of years for which a borrower would be liable for
making premium payments depended on the LTV ratio at the time of
origination. (See tables 8 and 9.) For loans originating from July 1, 1991,
16 New mortgage loans insured by FHA in a given fiscal year.
through the time of our review, premiums equal the original loan amount
times the respective up- front premium plus the product of the annual
outstanding principal balance times the respective annual premium rate for
as many years as annual premiums were required.
Table 8: Premium Schedule for 30- Year Non- Streamline Mortgages, by Date of
Mortgage Origination Annual premium Up- front premium
rates and rates for 30- year
durations for non- streamline
30- year nonstreamline Applicable LTV Applicable dates
loans loans ratio
10/ 1/ 74 - 9/ 30/ 83 None 0. 5% for 30 years All loans 10/ 1/ 83 - 6/ 30/
91 3. 80% None 7/ 1/ 91 - 9/ 30/ 92 3.80% 0. 5% for 5 years
Under 90% 0. 5% for 8 years
90% through 95% 0. 5% for 10 years
Over 95% 10/ 1/ 92 - 4/ 16/ 94 3. 00% 0. 5% for 7 years
Under 90% 0. 5% for 12 years
90% through 95% 0. 5% for 30 years
Over 95% 4/ 17/ 94 - present 2. 25% a 0. 5% for 7 years
Under 90% 0. 5% for 12 years
90% through 95% 0. 5% for 30 years
Over 95% a From September 3, 1996, to January 1, 2000, new homeowners who
received financial counseling before buying an FHA- insured home were
eligible for reduced up- front premiums.
Table 9: Premium Schedule for 15- Year Non- Streamline Mortgages, by Date of
Mortgage Origination Annual premium Up- front premium
rates and rates for 15- year
durations for non- streamline
15- year nonstreamline Applicable LTV Applicable dates
loans loans ratio
10/ 1/ 74 - 9/ 30/ 83 None 0. 50% for 15 years All loans 10/ 1/ 83 - 6/ 30/
91 2. 40% None 7/ 1/ 91 - 9/ 30/ 92 3. 80% 0. 5% for 5 years
Under 90% 0. 5% for 8 years
90% through 95% 0.5% for 10 years
Over 95% 10/ 1/ 92 - 12/ 25/ 92 3.00% 0. 5% for 7 years
Under 90% 0.5% for 12 years
90% through 95% 0.5% for 15 years
Over 95% 12/ 26/ 92 present 2. 00% . None
Under 90% 0.25% for 4 years
90% through 95% 0.25% for 8 years
Over 95%
Some loans that originated in the 1990s are streamline refinanced mortgages
that are subject to different premium rates. Since streamline refinances do
not require an appraisal, we decided that mortgages coded in
FHA's database with an LTV of 0 could reasonably be assumed to represent
streamline refinance business. For streamline refinance mortgages that
originated before July 1, 1991, we applied the premium rates from table 10.
Table 10: Premium Schedule for 15- and 30- Year Streamline Refinanced
Mortgages That Originated Before July 1, 1991
Up- front Annual
Up- front Annual
premium rates premium
premium rates premium
Applicable for 30- year
rates for 30 for 15- year
rates for 15 refinancing dates
loans year loans
loans year loans
Before 7/ 1/ 91 3.80% None 2. 40% None 7/ 1/ 91 - 4/ 24/ 92 3. 80% 0. 5% for
5
3.80% 0. 5% for 5 years years
4/ 24/ 92 12/ 25/ 92 3.80% None 3. 80% None 12/ 26/ 92 present 3.80% None 2.
40% None
For all streamline refinance mortgages that originated after July 1, 1991,
we applied the premium rates for non- streamline loans. That is, for up-
front premium rates, we followed the 15- year or 30- year non- streamline
premium schedule for loans of those maturities. For annual premium rates and
number of years that annual premiums are paid, we applied the rates for
loans with an LTV of less than 90 percent.
Claim Payments Claim payments equal the outstanding principal balance on
foreclosed mortgages times the acquisition cost ratio. We defined the
acquisition cost ratio as being equal to the total amount paid by FHA to
settle a claim and acquire a property (i. e., FHA's "acquisition cost" as
reported in its database) divided by the outstanding principal balance on
the mortgage at the time of foreclosure. For the purposes of our analysis,
we calculated an average acquisition cost ratio for each year's book of
business using actual data for fiscal years 1975 through 1999. Acquisition
cost ratios generally decreased over time from a high of 1. 51 for loans
originating in 1975 to a low of 1. 09 for loans originating in 1999.
Net Proceeds FHA's net proceeds from the sale of foreclosed properties
depend on both the lag rate- the proportion of a year that passes between
the time of a foreclosure and the time the proceeds are received- and the
loss rate- the proportion of the cost of the property acquired that is not
recovered when the property is sold. These are calculated as follows:
Net Proceeds = Lag rate x claim payments from previous period x (1 - loss
rate) + (12- lag rate) x claim payments from the current period x (1 - loss
rate). The lag, which is the number of months between the payment of a claim
and the receipt of proceeds from the disposition of the property, varied as
follows. Before 1995, the lag was 5.9 months; in 1995, 5.35 months; in 1996,
4.7 months; and in 1997, 5. 26 months. For the years after 1997, we used a
lag of 5.26 months. To calculate the lag rate for each period, we divided
the lag by 12.
We defined the loss rate as equal to FHA's reported dollar loss after the
disposition of property divided by the reported acquisition cost over the
historical period. We determined a loss rate for each year per book of
business for years 1 through 25. We used an auto- regressive model to
forecast future loss rates. In addition to past values of loss rates, we
used
the origination year and policy year of the loan as independent variables in
this model. Using the results of this model, we forecast loss rates over the
period from fiscal years 2000 through 2023. For fiscal years 2024 through
2028, we used the estimated rate for 2023. Our loss rates averaged 37
percent over the forecast period. Refunded Premiums The amount of premium
refunds paid by FHA depends on the policy year in
which the mortgage is prepaid and the type of mortgage. For mortgages
prepaid between October 1, 1983, and December 31, 1993, refunds were equal
to the original loan amount times the refund rate. However, we converted
these rates to express them as a percentage of the up- front
premium. In 1993, FHA changed its refund policy to affect mortgages prepaid
on or after January 1, 1994. For loans prepaying on or after January 1,
1994, refunds are equal to the up- front mortgage insurance premium times
the refund rate. (See table 11.)
Table 11: Premium Refund Rates for Loans That Were Terminated After October
1, 1983 Refund rates on loans prepaid between
10/ 1/ 83 and 12/ 31/ 93 Refund rates
on all loans Year 15- year loans 30- year loans
prepaid after 1/ 1/ 94
1 99. 0% 99. 0% 95. 0% 2 93. 0% 94. 0% 85. 0% 3 81. 0% 82. 0% 70. 1% 4 66.
0% 67. 0% 49. 4% 5 51. 0% 54. 0% 30. 2% 6 39. 0% 43. 0% 15. 1% 7 29. 0% 35.
0% 4. 2% 8 21. 0% 29. 0% 0. 0% 9 15. 0% 24. 0% 0. 0% 10 11. 0% 21. 0% 0. 0%
11 8. 0% 18. 0% 0. 0% 12 6. 0% 16. 0% 0. 0% 13 4. 0% 15. 0% 0. 0% 14 3. 0%
13. 0% 0. 0% 15 2. 0% 12. 0% 0. 0%
Refund rates on loans prepaid between 10/ 1/ 83 and 12/ 31/ 93
Refund rates on all loans Year 15- year loans 30- year loans
prepaid after 1/ 1/ 94
16 0. 0% 11. 0% 0. 0% 17 0. 0% 10. 0% 0. 0% 18 0. 0% 9. 0% 0. 0% 19 0. 0% 9.
0% 0. 0% 20 0. 0% 8. 0% 0. 0% 21 0. 0% 7. 0% 0. 0% 22 0. 0% 7. 0% 0. 0% 23
0. 0% 6. 0% 0. 0% 24 0. 0% 5. 0% 0. 0% 25 0. 0% 5. 0% 0. 0% 26 0. 0% 4. 0%
0. 0% 27 0. 0% 4. 0% 0. 0% 28 0. 0% 4. 0% 0. 0% 29 0. 0% 4. 0% 0. 0% 30 0.
0% 0. 0% 0. 0%
Administrative Expenses Administrative expenses equal the outstanding
principal balance times the administrative expense rate. The estimates of
the administrative expense rates were 0.098 percent for the years before
1995, 0.113 percent for 1995, 0.097 percent for 1996, 0. 102 percent for
1997, and 0.103 percent for 1998 and all future years. Sensitivity Analysis
We conducted additional analyses to determine the sensitivity of our
forecasts to the values of certain key variables. Because we found that
projected losses from foreclosures are sensitive to the rates of
unemployment and house price appreciation, we adjusted the forecasts of
unemployment and price appreciation to provide a range of estimates of the
Fund's economic value under alternative economic scenarios. Our starting
points for forecasts of the key economic variables were forecasts
made by Standard & Poor's DRI, as previously described. For our low case
scenario, we made these forecasts more pessimistic by subtracting 2
percentage points per year from the forecasts of house price appreciation
rates and adding 1 percentage point per year to the
unemployment rate forecasts. For our high case scenario, we added 2
percentage points per year to our base case forecast of house price
appreciation rates. Under these alternatives, we estimated economic values
of about $13.6 billion and about $16.4 billion, respectively, for the low
and high cases, compared with about $15.8 billion for our base case. These
estimates correspond to estimates of the capital ratio of about 2.75
percent and 3.32 percent, respectively, for the low and high cases, compared
with our base case estimate for the capital ratio of 3.20 percent. These
estimates are shown in table 12.
Table 12: Alternative Estimates of Capital Ratios for FHA's Mutual Mortgage
Insurance Fund
Dollars in billions
Capital ratio Scenario Economic value (percent)
Most likely economic $15.8 3. 20 conditions Low case scenario 13. 6 2.75
High case scenario 16. 4 3. 32 Source: GAO analysis.
To assess the impact of our assumptions about the loss and discount rates on
the economic value of the Fund, we operated our cash flow model with
alternative values for these variables. We found that for the economic
scenario of our base case, a 1- percentage- point increase in the forecasted
loss rate resulted in a 0.7- percent decline in our estimate of the economic
value of the Fund. Conversely, each percentage point decrease in the loss
rate resulted in a 0.7- percent increase in our estimate of economic value.
With respect to the discount rate, we found that for our base case economic
scenario, a 1- percentage- point increase in the interest rate applied to
most periods' future cash flow resulted in a 0. 3- percent increase in our
estimate of economic value. Conversely, each percentage point decrease in
the discount rate resulted in a 0. 4- percent decrease in our estimate of
economic value.
Development of Scenarios of Adverse Economic Conditions Used to Estimate the
Appendi x II I Economic Value of the Fund This appendix describes the
scenarios that we used to estimate the ability of the Fund to withstand
adverse future economic conditions. Each scenario specifies values of key
economic variables, which our models indicate are associated with mortgage
claims and prepayments, during the
forecast period. We used these values with the forecasting models presented
in appendix II to estimate future mortgage claims and prepayments. We then
used these forecasted values of claim and prepayment dollars in our cash
flow model to estimate the economic value of the Fund and the capital ratio
under each scenario.
We developed two types of scenarios- historical and judgmental. We designed
the historical scenarios to test the ability of the Fund to withstand
adverse economic conditions similar to those that adversely affected the
Fund in the 1980s and 1990s. Because some of these adverse conditions
affected only certain regions, in some scenarios we expanded our analysis to
include estimates of the capital ratio when the historical conditions were
assumed to affect a larger share of FHA's business, including when they were
assumed to affect the entire nation. In contrast, the judgmental scenarios
that we developed are not based on historical experience. Instead, they
represent conditions that we believe might place stress on the
Fund. The key economic variables for which we forecast different values in
the different scenarios are the rate of house price appreciation; the
unemployment rate; and, in some instances, certain interest rates,
especially the mortgage interest rate. In addition, we assumed that FHA's
loss per claim (the loss rate), expressed as a percentage of the claim
amount, was greater than the loss rate that we used in our base case
analysis under expected economic conditions. We assumed that FHA would
experience higher loss rates when foreclosures were substantially
higher because of the difficulty of managing and disposing of a large number
of properties at the same time. In addition, the demand for housing would be
likely to fall during an economic downturn, making it more difficult to
dispose of properties than in the base case.
Historical Scenarios Three regional economic downturns and the 1981- 82
national recession form the bases of our historical scenarios. Each regional
downturn was associated with a regional decline in house prices. Declining
house prices
represent a particularly adverse condition for the Fund because of the
strong negative relationship between borrowers' equity and the probability
of defaults leading to foreclosures. The three regional economic
downturns, and associated housing price declines, that we used were (1) the
late 1980s' decline in the oil- producing states of the west south central
region; (2) the late 1980s' and early 1990s' decline in New England; and (3)
the early to mid- 1990s' decline in the Pacific region, particularly in
California. 1 For each scenario that is based on a regional downturn, we
assumed that for 4 years the rate of house price change for the part of the
nation assumed to be affected by the downturn equaled the rate of house
price change in the state in that region that we selected to represent the
regional experience. We selected the experiences of (1) Louisiana, beginning
in 1986, to represent the oil price downturn; (2) Massachusetts, beginning
in
1988, to represent the New England economic downturn; and (3) California,
beginning in 1991, to represent the California housing market downturn.
Table 13 shows the median house prices for existing houses in
these states during their economic downturns. In calculating homeowner's
equity, we made the same adjustment to annual changes in median house prices
that we did in our base case, as described in appendix II. Similarly, in our
scenarios that are based on regional downturns, we assumed that unemployment
rates would change in the affected area for 4 years by the
same percentages as those rates changed in Louisiana; Massachusetts; and
California, respectively.
1 The west south central region is comprised of Arkansas, Louisiana,
Oklahoma, and Texas. The Pacific region is comprised of Alaska, California,
Hawaii, Oregon, and Washington. The New England region is comprised of
Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont.
Table 13: Historical Median House Price and Unemployment Experience in
States Representing Regional Downturns House price Unemployment rate Ordinal
Percentage Percentage State Year year Amount of year 0 Percent of year 0
Louisiana 1986 0 $75.3 100. 0 11. 2 a 100.0 1987 1 78. 7 104. 5 12. 8 114.1
1988 2 73.5 97.6 12.5 113. 4 1989 3 69.6 92.4 11.2 100. 1 1990 4 67.3 89.4 -
Massachusetts 1988 0 142. 3 100. 0 3. 1 100. 0 1989 1 137. 5 96. 6 3. 7
118.7 1990 2 135. 4 95. 1 5. 4 171.2 1991 3 128. 0 90. 0 8. 6 274.1 1992 4
133. 1 93. 5 8. 7 277.7 California 1991 0 199.9 100. 0 7.4 100. 0
1992 1 198. 9 99. 5 8. 9 120.1 1993 2 191. 6 95. 8 9. 5 128.6 1994 3 183. 7
91. 9 8. 9 120.9 1995 4 183. 0 91. 5 7. 9 106.6 a For Unemployment rates in
Louisiana, we used 1985 as the base year (year 0) and used changes in the
unemployment rate for 3 years (1986 through 1988) only.
We developed six separate scenarios that are based on each regional
downturn, by varying the scope (i. e., the number of states assumed to be
affected) and timing of the adverse economic conditions in the forecast
period. Specifically, we used three different scopes. In the narrowest
scope, we assumed that only the particular region was affected. That is, for
the scenario based on the downturn in the west south central region in the
late 1980s, we assumed that during 4 years of the forecast period, all of
the states in the west south central region experienced the same changes in
key economic variables as Louisiana experienced from 1987 through 1990. We
then expanded the scope by assuming that two regions in which FHA has a lot
of borrowers, the west south central and Pacific regions, were
affected. 2 Finally, we then expanded the scope to the entire nation, by
assuming that all states were affected.
Regarding timing, for each scope we developed two scenarios, one in which
the downturn began in 2000 and one in which it began in 2001. Although we
know that an economic downturn did not begin in 2000, we developed scenarios
starting then to test the ability of the Fund to withstand an economic
downturn that occurs when the portfolio contains many recent loans.
Scenarios in which the downturn does not begin until 2001 would be expected
to be less adverse because most of the large number of borrowers who took
out mortgages in 1998 and 1999 would have seen substantial price
appreciation in 2000, thereby reducing the likelihood of default. We
developed two historical scenarios that are based on the 1981- 82 recession
and subsequent recovery. In those scenarios, we assumed that in each state,
the rates of change in house price appreciation and unemployment for 5 years
during the forecast period are the same as they were from 1981 through 1985.
In one scenario, we assumed that these
adverse conditions replicating 1981 through 1985 began in 2000; in the other
scenario, we assumed that they began in 2001. Under these scenarios, some
states fared better than in the base case scenario.
Because it will be more difficult to manage and dispose of foreclosed
properties during an economic downturn, we increased the loss rates on the
proportion of mortgages affected by a given scenario during the years the
scenario runs. We assumed that losses on affected foreclosed properties
would rise to 45 percent of the property's value. Without this, loss rates
average about 37 percent. Our estimates of the economic value of the Fund
and the capital ratio for the historical scenarios are presented in table
14.
2 We used these two regions for each set of scenarios, even the set that is
based on the New England decline, because FHA does relatively little
business in New England, and we wanted to test the ability of the Fund to
withstand a downturn like the one a decade ago in New England, if it
occurred simultaneously in two regions in which FHA does a lot of business.
Judgmental Scenarios We developed several judgmental scenarios to test the
ability of the Fund to withstand various types of economic conditions that
might adversely affect the Fund without regard to their relationship to
historical experience. In one scenario, we assumed that median existing
house prices declined by 5 percent per year for 3 consecutive years- an
extremely steep rate of decline 3 -and that unemployment increased compared
with the base case, with both changes beginning in 2001. Specifically, we
increased the
unemployment rates in each state from forecasted levels by 2 percentage
points in 2001; 5 percentage points in 2002, 2003, and 2004; and 2
percentage points in 2005. In a second scenario, we allowed the mortgage
interest rate to decline in 2000- by 2 percentage points from its forecasted
level- and then to return to forecasted levels. We did this to precipitate a
wave of refinancing. We also assumed declining house prices and rising
unemployment beginning in 2001, as in the previous judgmental scenario.
We used this scenario to test what might happen if premium income turns out
to be substantially less than expected and premium refunds substantially
more than expected because of rapid prepayment of loans, most of which would
not default. In our third scenario, we added 1 percentage point to the base
case forecasts of the mortgage interest rate, and 1- and 10- year Treasury
rates for the year 2000, 3 percentage points to the forecasts of these
interest rates between 2001 and 2003, and 1 percentage point in 2004. We
used this scenario to test what might happen
if interest rates were to rise more than anticipated. In a fourth scenario,
we used the same rising interest rates as in the third scenario and also
added one percentage point to the forecasts of median existing house prices
over that period. Our estimates of the economic value of the Fund and the
capital ratio for the judgmental scenarios are also presented in table 14.
3 Since we adjusted growth in median house prices for quality improvements,
as described in appendix II, the decline in house prices facing individual
borrowers is even greater.
Table 14: National Results of Alternative Scenarios Average unemployment
Average
Economic Starting
Number of Average
rate equity
value Capital ratio Geographic extent year loan years house price
(percent) (percent)
(millions) (percent) Base Case
National NA 3,974,076 $124, 076 4.3 49.9 $15,810 3. 20
Historical scenarios that were based on the Louisiana experience of 1986- 90
One region (WSC) 2000 4, 010,385 $116, 739 4.3 45.6 $15,130 3. 06 Two
regions (WSC and PAC) 2000 4, 058,778 111, 618 4.4 43.2 13,861 2. 81 Nation
2000 4, 308,066 85, 005 4. 4 29.1 11,430 2. 31 One region (WSC) 2001 4,
003,327 117, 360 4.3 46.2 15,320 3. 10 Two regions (WSC and PAC) 2001 4,
040,790 112, 900 4.3 44.2 14,606 2. 96 Nation 2001 4, 233,621 89, 016 4. 3
32.2 12,848 2. 60
Historical scenarios that were based on the Massachusetts experience of
1988- 92
One region (NE) 2000 3, 991,873 122, 936 4.3 49.3 15,510 3. 14 Two regions
(WSC and PAC) 2000 4, 024,416 112, 829 4.9 43.9 10,566 2. 14 Nation 2000 4,
261,430 88, 058 5. 8 31.2 4, 020 0.81 One region (NE) 2001 3, 985,822 123,
108 4.3 49.5 15,623 3. 16 Two regions (WSC and PAC) 2001 4, 020,651 113, 934
4.8 44.8 12,291 2. 49 Nation 2001 4, 211,557 91, 832 5. 5 33.9 7, 759 1.57
Historical scenarios that were based on the California Experience of 1991-
95
One region (PAC) 2000 4, 026,719 118, 921 4.4 47.5 14,281 2. 89 Two regions
(WSC and PAC) 2000 4, 057,387 112, 031 4.4 43.5 12,815 2. 59 Nation 2000 4,
304,672 86, 298 4. 5 30.1 10,683 2. 16 One region (PAC) 2001 4, 013,887 119,
695 4.3 48.0 14,912 3. 02 Two regions (WSC and PAC) 2001 4, 041,526 113, 280
4.4 44.4 14,353 2. 91 Nation 2001 4, 231,923 90, 223 4. 4 33.0 12,319 2. 49
Historical scenarios that were based on state- by- state experiences during
the 1981- 82 recession
Nation 2000 4, 025,522 123, 503 4.7 49.1 13,876 2. 81 Nation 2001 3, 997,043
127, 824 4.5 50.2 14,673 2. 97
Judgmental scenario: mortgage rate decline in 2000 followed by 3 years of 5
percent declines in house prices and increased unemployment starting in 2001
Nation 2000 3, 795,878 86, 731 5. 4 29.5 6, 779 1.37
Judgmental scenario: 3 years of 5 percent declines in house prices and
increased unemployment starting in 2001
Nation 2001 4, 285,284 87, 005 5. 4 30.3 6, 918 1.40
Judgmental scenario: higher mortgage rates early in forecast period
Nation 2000 4, 594,178 125, 287 4.3 53.9 16,608 3. 36
(Continued From Previous Page)
Average unemployment Average
Economic Starting
Number of Average
rate equity
value Capital ratio Geographic extent year loan years house price
(percent) (percent)
(millions) (percent) Judgmental scenario: higher mortgage rates early in
forecast period, accompanied by faster house price growth
Nation 2000 4, 538,531 139, 330 4.3 58.1 16,818 3. 40
In another type of judgmental scenario, we did not forecast the economic
variables and then use the forecasted claims and prepayments from our
econometric model, as we did with all of our other scenarios, both
judgmental and historical. Instead, because none of our other scenarios
produced foreclosure rates nearly as high as FHA experienced in the 1980s,
we developed two scenarios in which we directly assumed higher foreclosure
rates. First, we assumed that in 2000 through 2004, the proportion of loans
insured in each region experienced for the 1989 through 1999 books of
business the same foreclosure rates that the 1975 through 1985 books of
business experienced in that region in 1986 through 1990. This scenario
produced a capital ratio of 0.92 percent. Second, we assumed that in 2000
through 2004, varying proportions of FHA's portfolio
experienced for the 1989 through 1999 books of business the same foreclosure
rates that the 1975 through 1985 books of business experienced in the west
south central states in 1986 through 1990. Because streamline refinanced
mortgages and ARMs did not exist or were minimal parts of FHA's portfolio
from 1975 through 1985, foreclosure rates were not adjusted for these types
of loans. For the other products- 30- year fixedrate, 15- year fixed- rate,
investor, and graduated payment mortgages- foreclosure rates were adjusted
accordingly for each type of product. For this scenario, we found that if
36.5 percent of FHA- insured mortgages experienced these high default rates,
the estimated capital ratio for fiscal year 1999 would fall by 2 percentage
points, and if about 55 percent of FHA's portfolio experienced these
conditions, the economic value would
be depleted.
Comments From the Department of Housing
Appendi x V I and Urban Development
Appendi x V
GAO Contacts and Staff Acknowledgments GAO Contacts Thomas J. McCool, (202)
512- 8678 Mathew Scir�, (202) 512- 6794 Acknowledgments In addition to those
named above, Nancy Barry, Elaine Boudreau, Steve
Brown, Jay Cherlow, Kimberly Granger, DuEwa Kamara, John McDonough,
Salvatore F. Sorebllo Jr., Mark Stover, and Patrick Valentine made key
contributions to this report.
Related GAO Products Financial Health of the Federal Housing
Administration's Mutual Mortgage Insurance Fund (GAO/ T- RCED- 00- 287,
Sept. 12, 2000). Level of Annual Premiums That Place a Ceiling on
Distributions to FHA Policyholders (GAO/ RCED- 00- 280R, Sept. 8, 2000).
Single- Family Housing: Stronger Measures Needed to Encourage Better
Performance by Management and Marketing Contractors (GAO/ T- RCED00- 180,
May 16, 2000, and GAO/ RCED- 00- 117, May 12, 2000).
Single- Family Housing: Stronger Oversight of FHA Lenders Could Reduce HUD's
Insurance Risk (GAO/ RCED- 00- 112, Apr. 28, 2000).
Homeownership: Results of and Challenges Faced by FHA's Single- Family
Mortgage Insurance Program (GAO/ T- RCED- 99- 133, Mar. 25, 1999).
Homeownership: Achievements of and Challenges Faced by FHA's SingleFamily
Mortgage Insurance Program (GAO/ T- RCED- 98- 217, June 2, 1998).
Homeownership: Management Challenges Facing FHA's Single- Family Housing
Operations (GAO/ T- RCED- 98- 121, Apr. 1, 1998).
Homeownership: Mixed Results and High Costs Raise Concerns about HUD's
Mortgage Assignment Program ( GAO/ RCED- 96- 2, Oct. 18, 1995).
Homeownership: Information on Single Family Loans Sold by HUD (GAO/ RCED-
99- 145, June 15, 1999). Homeownership: Information on Foreclosed FHA-
Insured Loans and HUDOwned Properties in Six Cities (GAO/ RCED- 98- 2, Oct.
8, 1997).
Homeownership: Potential Effects of Reducing FHA's Insurance Coverage for
Home Mortgages (GAO/ RCED- 97- 93, May 1, 1997).
Homeownership: FHA's Role in Helping People Obtain Home Mortgages (GAO/
RCED- 96- 123, Aug. 13, 1996). Mortgage Financing: FHA Has Achieved Its Home
Mortgage Capital Reserve Target (GAO/ RCED- 96- 50, Apr. 12, 1996).
(385822) Lett er
Ordering Information The first copy of each GAO report is free. Additional
copies of reports are $2 each. A check or money order should be made out to
the Superintendent of Documents. VISA and MasterCard credit cards are
accepted, also.
Orders for 100 or more copies to be mailed to a single address are
discounted 25 percent.
Orders by mail: U. S. General Accounting Office P. O. Box 37050 Washington,
DC 20013
Orders by visiting: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW)
U. S. General Accounting Office Washington, DC
Orders by phone: (202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537
Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.
Orders by Internet: For information on how to access GAO reports on the
Internet, send an e- mail message with “info” in the body to:
info@ www. gao. gov or visit GAO's World Wide Web home page at: http:// www.
gao. gov
To Report Fraud,
Contact one:
Waste, or Abuse in
? Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm
Federal Programs
? e- mail: fraudnet@ gao. gov
? 1- 800- 424- 5454 (automated answering system)
GAO United States General Accounting Office
Page 1 GAO- 01- 460 Mortgage Financing
Contents
Contents Page 2 GAO- 01- 460 Mortgage Financing
Page 3 GAO- 01- 460 Mortgage Financing United States General Accounting
Office
Washington, D. C. 20548 Page 3 GAO- 01- 460 Mortgage Financing
Page 4 GAO- 01- 460 Mortgage Financing
Page 5 GAO- 01- 460 Mortgage Financing
Page 6 GAO- 01- 460 Mortgage Financing
Page 7 GAO- 01- 460 Mortgage Financing
Page 8 GAO- 01- 460 Mortgage Financing
Page 9 GAO- 01- 460 Mortgage Financing
Page 10 GAO- 01- 460 Mortgage Financing
Page 11 GAO- 01- 460 Mortgage Financing
Page 12 GAO- 01- 460 Mortgage Financing
Page 13 GAO- 01- 460 Mortgage Financing
Page 14 GAO- 01- 460 Mortgage Financing
Page 15 GAO- 01- 460 Mortgage Financing
Page 16 GAO- 01- 460 Mortgage Financing
Page 17 GAO- 01- 460 Mortgage Financing
Page 18 GAO- 01- 460 Mortgage Financing
Page 19 GAO- 01- 460 Mortgage Financing
Page 20 GAO- 01- 460 Mortgage Financing
Page 21 GAO- 01- 460 Mortgage Financing
Page 22 GAO- 01- 460 Mortgage Financing
Page 23 GAO- 01- 460 Mortgage Financing
Page 24 GAO- 01- 460 Mortgage Financing
Page 25 GAO- 01- 460 Mortgage Financing
Page 26 GAO- 01- 460 Mortgage Financing
Page 27 GAO- 01- 460 Mortgage Financing
Page 28 GAO- 01- 460 Mortgage Financing
Page 29 GAO- 01- 460 Mortgage Financing
Page 30 GAO- 01- 460 Mortgage Financing
Page 31 GAO- 01- 460 Mortgage Financing
Page 32 GAO- 01- 460 Mortgage Financing
Page 33 GAO- 01- 460 Mortgage Financing
Page 34 GAO- 01- 460 Mortgage Financing
Appendix I
Appendix I Scope and Methodology
Page 35 GAO- 01- 460 Mortgage Financing
Appendix I Scope and Methodology
Page 36 GAO- 01- 460 Mortgage Financing
Appendix I Scope and Methodology
Page 37 GAO- 01- 460 Mortgage Financing
Page 38 GAO- 01- 460 Mortgage Financing
Appendix II
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 39 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 40 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 41 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 42 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 43 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 44 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 45 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 46 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 47 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 48 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 49 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 50 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 51 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 52 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 53 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 54 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 55 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 56 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 57 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 58 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 59 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 60 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 61 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 62 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 63 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 64 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 65 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 66 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 67 GAO- 01- 460 Mortgage Financing
Appendix II Models Used to Estimate the Economic Value of FHA's Mutual
Mortgage Insurance Fund
Page 68 GAO- 01- 460 Mortgage Financing
Page 69 GAO- 01- 460 Mortgage Financing
Appendix III
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 70 GAO- 01- 460 Mortgage Financing
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 71 GAO- 01- 460 Mortgage Financing
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 72 GAO- 01- 460 Mortgage Financing
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 73 GAO- 01- 460 Mortgage Financing
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 74 GAO- 01- 460 Mortgage Financing
Appendix III Development of Scenarios of Adverse Economic Conditions Used to
Estimate the Economic Value of the Fund
Page 75 GAO- 01- 460 Mortgage Financing
Page 76 GAO- 01- 460 Mortgage Financing
Appendix IV
Appendix IV Comments From the Department of Housing and Urban Development
Page 77 GAO- 01- 460 Mortgage Financing
Appendix IV Comments From the Department of Housing and Urban Development
Page 78 GAO- 01- 460 Mortgage Financing
Appendix IV Comments From the Department of Housing and Urban Development
Page 79 GAO- 01- 460 Mortgage Financing
Page 80 GAO- 01- 460 Mortgage Financing
Appendix V
Page 81 GAO- 01- 460 Mortgage Financing
United States General Accounting Office Washington, D. C. 20548- 0001
Official Business Penalty for Private Use $300
Address Correction Requested Presorted Standard
Postage & Fees Paid GAO Permit No. GI00
*** End of document. ***