United States GeEral Accounting Office Testimony GAO Mr. Chairman and Members of the Subcommittee: I am pleased to appear here today to discuss the results of our analyses of the impact of three proposed policy changes on the cash position of the Federal Housing Administration's (FHA) Mutual Mortgage Insurance (MMI) Fund. This discussion will serve to update the interim results and data that we presented in November 1989 before the Subcommittee on Housing and Community Development, House Committee on Banking, Finance and Urban Affairs.1 Our work, which was performed at the request of Congressman Gerald Kleczka, focused on the impact, over the next 10 years, on the Fund's cash balance from (1) increasing the FHA mortgage ceiling limits, (2) reducing downpaycent requirements, and (3) allowing FHA tc insure adjustable rate mortgages (APJ‘I) With higher interest rate caps. To analyze the financial impacts on the MM1 Fund, we developed a model for performing economic estimations of the Fund's cash flow over a lo-year period covering fiscal years 1989 through 1998. Before presenting the results of our work, however, I would like to stress that our analysis focused on the cash position of the Fund and is therefore not an assessment of the actuarial soundness of the Fund. The difference is that a 'cash analysis focuses on the revenue and expenses of the Fund on an annual basis rather than on the Fund's ability. to support potential losses over the entire life of the insured mcrtqaqes. An actuarial analysis, on the other hand, would focus on whether the Fund has enough reserves to cover future losses from insurance currently in force and nest insurance written. The important distinction here is that the cash balance alone cannot indicate the Fund's actuarial soundness. The volume of loans insured and anticipated future losses must also be considered to determine if pclicy changes (such as raising the mortgage ceiling that result in higher cash balances, but higher potential l"'Impact of FHA Loan Policy Chanqes" (GAO/T-RCED-90-17, Nov. 16, 1989). liability, are preferable from an actuarial standpoint. In this regard, the President's fiscal year 1991 budget states that the Fund is not actuarially sound and changes, such as increasing the FHA premium, will be needed. FHA has contracted for an independent actuarial study to determine what reforms are needed to make the Fund actuarially sound. Having said all that, let's take a look at what our model shows are the possible impacts of each policy alternative on the Fund's cash position. SUmARY The overall cash position of the MM1 Fund during the 1990s and the effect of the various policy options will depend heavily on actual econcmir conditions durinr; the next decade. If house prices appreciate at a rate of 5 to 9 percent per year, and overall economic conditicns remain generally favorable, the Fund's cash balance will likely be adequate to cover anticipated losses during the next 10 years. However, the Fund will not perform nearly as well if the recent trend of lower rates of house appreciation continues through the 1990s. For example, if the rate of house price appreciation is less than 4 percent, the Fund will likely not be able to survive without U.S. Treasury assistance even if overall economic conditions remain generally favorable. Of the prcposals we analyzed, increasing the mortgage ceiling would have the greatest effect on the Fund's cash balance. Besides the overall cash position of the Fund and the effect of the various policy options, of equal importance is the question of the Fund's actuarial soundness. Increasing the mortgage ceiling will have a positive effect on cash balances: however, it will also increase the government's financial risk in the form of additional insurance-in-fcrce. This raises serious concerns, particularly if the Fund is not actuarially sound. We analyzed the cash position of the Fund under four different economic scenarios. In three scenarios, we assumed economic conditions were generally favorable, while applying various rates of annual house price appreciation to each scenario--7 to 9 percent, 5 to 7 percent, and 2 to 4 percent. In the fourth scenario, economic conditions in the 1990s were assumed to approximate those experienced in the 1980s. Assuming generally favorable economic conditions, that house prices increase at about 7 to 9 percent annually, and that the mortgage ceiling is allowed to increase with the annual increase in house prices - assumptions which constitute our base case - the Fund's cash balance will grow from1 $6.2 billion at the beginning of 19E9 to an estimated $8.8 billicn in 1998 (see exhibit I). At the same time, however, the Fund's insurance-in-force will more than double from $271 billion in 1989 to $685 billion by the end of 1998. If our other base case assumptions remain unchanged but house prices are assumed to increase at rates of 5 to 7 percent per year, as some experts are now predicting, the Fund's cash balance will remain positive but will decline to an estimated $3.9 billion by the end of 1998. If economic conditions instead approximate those experienced during the 1980s (which included a period early in the decade of high interest rates and high unesployment rates) and the mortgage ceiling is still allowed to increase with the annual rate of house price appreciation, the Fund's cash balance will also remain positive but will decline to an estimated $3.8 billion by the end of 1998. Under economic scenarios having generally favorable economic conditions but lower house price appreciation rates, the Fund would likely insure fewer loans and lower loan amounts. It would thus receive less income from premiums. As a result, its cash balance would decline--in some cases becoming negative by the end of fiscal year 1998. For example, at an annual 2-4 percent house price appreciation rate, the Fund will be unable to maintain a positive cash balance without Treasury assistance if no policy change is made or if any of the proposals we reviewed are adopted, even if economic conditions remain generally favorable. If the Congress decides not to raise FHA's mortgage ceiling above the $101,250 loan limit, the Fund's cash balance would fall to an estimated $3 billion in 1998 even under favorable economic conditions. This would occur because FHA's share of the mortgage market, under a constant loan limit, decreases as house prices appreciate above the FHA maximum limit. The decreased market share, in turn, would add less premium income to the Fund to cover future potential claims. Without increases in the maximum allowable loan limit, both the Fund's cash balance and home purchase options for homebuyers iii11 decrease. Of the three proposals GAO analyzed, increasing the mortgage ceiling to 95 percent of a state's median house price would have the greatest effect on the Fund's balance and would also assist in generating the most new business when compared with GAO's estimated base case. The reduced downpayment and ARM proposal are estimated to have a relatively small effect on the Fund's cash position. Under generally favorable economic conditions and annual house price appreciation rates of 7 to 9 percent, the 95-percent mortgage ceiling proposal, we estimated, will result in a cash balance of $14.4 billion. This would be an increase in the Fund's cash balance of $8.2 billion by the end of 1998, and is $5.7 billion higher than the balance we estimated if the mortgage ceiling rises only at the annual rate of house price appreciation. Because of the substantial increase in the FHA mortgage ceiling, however, the amount of insurance-in-force will more than triple to $886 billicn b,y 1998. Under the other economic scenarios, the 4 effects of raising the mortgage ceiling to 95-percent of a state's median house price are generally similar. What does this te 1 us? It tells us that while increases in the FHA's mortgage ceiling to account for house price increases are necessary to prevent the deterioration of the Fund's balance and to allow FHA to maintain its current share of the housing market, there is a need to proceed with caution on how hiuh to raise the mortqaqe ceilinq. If the Fund is not actuarially sound, we believe the Congress should not raise the mortgage ceiling to the 95 percent level because this action may subject the federal government to enormous costs over the life of the new insurance that will be created. We believe that a decision to raise the mortgage ceiling, under these conditions, should be made in conjunction with a decision on how to resolve the problem of actuarial soundness so that the potential financial risks assumed by the federal government in the long run are adequately considered. If, on the other hand, the Fund is actuarially sound, we believe the Congress should consider raising the mortgage ceiling to 95 percent of a state's median house price because doing so will have a positive effect on the Fund's cash position and will al low homebuyers in high priced states to participate more act ively in the FHA program. In the statement which follows, I provide a more detailed discussion of our assumptions, methodology, and results. In addition, I will comment briefly on the role of FHA in the national mortgage market and summarize the comments we received from academia, government, and housing industry representatives on FHA's role. We expect to issue a report containing a more complete discussion of the policy changes' financial impacts and related FHA management issues during the summer. BACKGROUND FHA was established in 1934 under authority granted to the President by the National Housing Act (P.L. 73-479). In 1948, FHA became a wholly owned government corporation subject to the Government Corporation Control Act, as amended. FHA and its functions were transferred to the Department of Housing and Urban Development (HUD) in 1965. After the transfer, FHA's staff and facilities were merged with those of other housing activities. The basic purpose of FHA programs is to encourage improvement in housing standards and conditions, provide an adequate nome financing system through mortgage insurance, and exert a stabilizing influence on the mortgage market. To carry out this purpose, the Secretary of HUD administers FHA through four separate Funds for its various mortgage insurance programs--the Mutual Mortgage Insurance (MMI) Fund, the Cooperative Management Housing Insurance (CMHI) Fund, the General Insurance (GI) Fund, and the Special Risk Insurance (SRI) Fund. Our work dealt only with the MM1 Fund, which basically insures single-family homes. GAO'S ANALYSIS OF MM1 FUND POSITIOJ The MM1 Fund is FHA's largest fund with $271 billion of insurance-in-force as of September 30, 1989. As a result of a full financial audit, we determined that the MM1 Fund had a loss of $1.4 billion in fiscal year 1988. This loss caused the government's equity in this Fund to fall to $1.8 billion at the end of the fiscal year. The MM1 Fund provides basic single-family mortgage insurance and is intended to be self-sustaining through charging the homebuyer a premium o f 3.8 percent of the mortgage amount. Let's look at why the MM1 Fund is losing money. The $1.4 billion loss in the MM1 Fund for fiscal year 1988 is mainly attributable to a $1.2 billion increase in its loss 0 reserves. These reserves are necessary to account for losses on foreclosed loans that will eventually lead to claims. Two major factors contribute to the MM1 Fund's increase in loss reserves. First, the record high single-family mortgage insurance endorsements in 1986 and 1987 are entering the period in which historical evidence suggests that high claim rates could occur. Thus, foreclosures may remain at a high level. Like many private mortgage insurers, the NM1 Fund generally experiences its highest rate of claims in the second and third year after the insurance is written. The claim rate usually decreases gradually after the third year and levels off after the tenth year IIf the policy. Given the significant level of insurance written by- the MM1 Fund in 15E6 and 1987 ($44.5 and $82.6 billion, respectively), foreclosures are likely to continue at a relatively high level, at least in the near term. The other factor contributing to the increase in loss reserves in the IVDKI Fund is the persistently high default and foreclosure rates in economically stressed regions, particularly the Rocky Mountain and Southwest regions. While the percentage of total MM1 insurance-in-force written in these regions has remained relatively stable, claim rates, and thus losses, have been substantial in these stressed regions. FHA POLICY OFTIONS Our analysis focused on the cash position of the Fund at the end of each fiscal year, during the period 1989 to 1998. It shows that the cash pcsition of the Fund is influenced by FHA loan insurance policies and economic conditions. To conduct this analysis, we developed econometric models based on an analysis of historical trends in FHA mortgages originated during fiscal years 1579 through 1988. These 7 econometric models identify the relationships between claim and nonclaim terminations and a variety of explanatory variables, including loan-to-value ratios, loan amounts, the rate of house price appreciation, and other economic variables. The results from these models were then combined with a cash-flow model to provide projections of the cash position of the Fund over fiscal years 1989 through 1998. Our analysis of claim rates developed from FHA's data base is consistent with prior studies and conventional economic reasoning. For example, -- Claim rates tend to pea): in the second and third year after loan crigination and then decline in subsequent years. -- Claim rates are higher for loans with higher loan-to-value ratios. -- Claim rates are lower for higher valued mortgages (within the 19EC FHA loan lim<t of $101,250). -- Claim rates decline as a homeowner's equity increases through repayment of the mortgage balance and through home price appreciation. We projected the cash positlon of the Fund for several polic) options. These included the following: a base case in which the mortgage ceiling is raised in accordance with house price increases to enable FHA to maintain its market share; a more substantial increase in the mortgage ceilin,g: a reduction in downpayment requirements; and allowing FHA to guarantee ARMS with higher caps on the annual and lifetime interest rate increases. To make these projections, we used forecasted values of economic var iables developed blV Data Resources, Incorporated (DRI ; 3 in the fall of 1989. DRI provided forecasts of unemployment rates, interest rates, housing prices, and loan volumes and values. The DRI "trend" economic forecast we used predicts that the economy will perform reasonably well over the next 10 years--mortgage rates average from 9.4 to 10.3 percent: the unemployment rate does not exceed 5.5 percent: and housing prices, except for fiscal year 1989 which showed an annual increase of about 4 percent, increase at 7 to 9 percent annually over the 1989-1998 period.2 The forecast values that were used are shown in exhibit III. Base Case Analvsis In our base case analysis, we used the loan ceiling cf $101,250 for 1989 and changed this limit each year according to an index of housing prices so that, using DRI's trend economic forecast, it reaches slightly more than $206,000 in 1998. The base case thereby assumes that FHA's market share is not eroded because of properties increasing in price so that they can no longer qualify for FHA insurance. The base case further assumes that the proportion of ARMS will remain at its current level in the FHA portfolio and that current downpayment requirements will remain unchanged. Under authority granted by Congress, HUD raised the maximum loan limit to $124,875 for the period January 12, 1990, to September 30, 1990. Barring further congressional action, the loan limit will revert to $101, 250 at the end of that period. Because of the higher loan limit's introduction late in the our review period and its possible short duration, we have used the $101,250 limit in our base case calculations throughout this statement. 2 DRI's most recent projections show housing prices for the 1989-98 period increasing at about 6 percent annually. According to the National Association of Realtors, median house prices rose 7.4 percent annually over the 21-year period from 1968-86, although they rose by only 4.6 percent per year from 1979-88. 9 Under this base case scenario and using DRI's trend economic forecast, we projected the MM1 Fund will have positive cash flows in 8 of the 10 forecast years. These results are displayed in exhibit I. The projections show the cash position of the Fund increasing by the end of 1998 to an estimated $8.8 billion. Under these economic conditions, the volume of insurance-in-force for FHA is estimated to increase from $271 billion in fiscal year 1989 to $685 billion in fiscal year 1998. Thus, the Fund in 1998 will have greater potential exposure to loss. This base case scenario can be compared to an analysis of the Fund's projected cash balance if the Congress decides not to raise FHA's mortgage ceiling above the $101,250 loan limir. This projection showed that if the ceiling remained at $101,250, the Fund's cash balance would, based on our projections, fall to $3 billion by 1998. This projection again uses DRI's trend economic forecast. This decrease would occur because a constant mortgage ceiling would over time reduce the volume of newly-insured FHA loans in housing markets where home prices are increasing. This would cause a drop in premium revenues while claims continue to be paid. As a result, the Fund's cash balance would begin to decline by 1995. Raisins FHA Loan Ceilinq to 95 Percent of State Median House Price In our analysis of the effect of increasing the mortgage ceiling, we allow the ceiling to increase in the first year to 95 percent of each state's median house price and thereafter increase each year at the same rate at which house prices appreciate. Setting a higher ceiling in this manner expands FHA's business in very high price states, such as California and Connecticut (see exhibit IV). FHA would be able to write substantially more insurance in these states, although it still would be limited within certain metropolitan areas. By increasing FHA's volume of 10 business, this change would increase FHA's premium income and cash position. The increased ceiling would provide higher cash balances for the Fund for two reasons. First, the Fund would receive greater premium income because it would insure more mortgages. On the other hand, accompanying the higher premium income would be a greater volume of insurance-in-force, meaning that the Fund would have higher future liabilities. Second, our analysis of the loans insured by FHA showed that the foreclosure and loss rates would be slightly lower for higher valued loans. According to FHA's experience over the last 10 years, larger loans tend to show slightly lower foreclosure rates and experience lower per_entaqe losses when they are foreclosed. Using DRI's trend economic forecast, we projected that end- of-year cash balances would grow from $6.2 billion in 1988 to $14.4 billion by 1998, an increase of SE.2 billion. This compares with a projected 1998 cash balance of $8.8 billion in our base case, indicating that with this policy change the 1998 cash balance would be expected to be $5.7 billion higher than in the base case.3 At the same time, the amount of insurance-in-force would reach $886 billion by fiscal year 1598 under this proposal. While raising the loan ceiling would lead to a large growth in FHA business and cash balances, lower claim rates and losses would result, given the generally favcrable economic conditions and sizeable increases in house prices projected under the DRI trend forecast. (See exhibit I.) However, several factors that might reduce the positive impact of this policy change on the Fund include the following: 3The numbers presented here do not necessarily add due to rounding. (See exhibit I). 11 -- Analysis of FHA's data base indicates that higher value loans within present FHA limits have lower claim rates, partly because these loans have had higher downpayments associated with them. However, as the ceiling amount of the loan would be raised in high cost areas, many new borrowers either might not desire or might not be able to make correspondingly higher downpayments. Therefore, to the extent this happens, the potential risk associated with these loans would increase. -- When higher value loans are foreclosed, dollar losses might be higher than they are on lower value loans. Some private mortgage insurers contend that claim rates rise with loan size. High-value mortgages would also result in higher dollar losses, should a foreclosure occur. As part of our study, we obtained the sometimes conflicting views of officials knowledgeable about housing issues, representing government agencies, academia, and mortgage industry organizations. Examples of the commer,ts received in support of raising the limit to a percentage of the median hoJse price follow: -- Raising the limit is the only way to deal with the housing affordability problem, because it would increase homeownership opportunities while reducing claim rates and losses. -- Raising the limit would increase FHA's volume and market share and improve the geographical distribution of FHA loans, better insulating it from sectional risks. -- Raising the limit would have only a minimal impact on private mortgage insurers. 12 On the other hand, examples of comments from those opposed to the increase in the loan ceiling differ: -- Losses on high loan-to-value loans will be compounded for FHA and risks will be more geographically concentrated in places like California and some eastern states. -- As builders use the FHA ceiling as their benchmark, house prices will increase. -- The market share of private insurers will be reduced substantially. -- Linking FHA limits to area median house prices will (1) significantly increase FHA's market and risk exposure without benefittinq moderate- or low-income households and (2) inevitably make FHA insurance more available to households in areas with higher real incomes than in areas with lower real incomes. Revised Downpayment Requirement FHA currently requires a downpayment of 3 percent on the first $25,000 and 5 percent on the amount above $25,000 unless the appraised value of the home is $50,000 or less, in which case the required downpayment is 3 percent. We evaluated the proposal that FHA reduce its downpayment requirements by requiring 3 percent down for amounts at or below $50,000 and 5 percent down for amounts over $50,000. For example, under this proposal, a $100,000 mortgage would require a minimum downpayment equaling 4 percent--3 percent on the first $50,000 and 5 percent on the second $50,000. If this alternative is adopted and the mortgage ceiling is still increased annually according to the annual increase in house prices, under 3-l DRI's trend economic forecast, the Fund balance would increase by $1.7 billion from the end of 1568 to 1998--a reduction of $0.9 billion in the 1998 cash balance from the base case. (See exhibit I.1 Adiustable Rate Mortqaqes The FHA currently insures ARMS having a l-percent annual cap and a 5-percent lifetime cap. The ARM most frequently offered by private lenders has a 2-percent annual cap and a 6-percent lifetime cap. Under current policy FHA cannot insure the preferred instrument. Therefore, very little of its portfolio is in ARMS. The third policy change we considered in our analysis was to allow FHA to insure "two-six" AFGls but to limit them to 30 percent of the FHA portfolio. Under DRI's trend economic forecast, adoption of this policy would have very little effect on the cash balances of the Fund, increasing it by less than $200 million compared with the base case by the end of the forecast period. This occurs for two reasons. First, we have assumed that ARMS will not represent new business but simply transfers of fixed mortgages into ARMS. To the extent that ARMS represent new business, the Fund will receive additional premium income--but experience corresponding growth in loan exposure. Second, the forecast of economic conditions includes no significant increases in interest rates. ARMS, unlike fixed-rate mortgages, increase the risk of foreclosure during periods of rising interest rates and reduce risks during periods of declining rates. With forecasts of stable rates, losses associated with ARI"I business would not differ substantially from those of fixed-rate loans. IMPACT UNDER LESS OPTIMISTIC ECONOMIC FORECASTS DRI's trend economic forecast reflects generally favorable economic conditions and housing price appreciation rates that exceed the inflation rate--factors that are very favorable to the cash position of the Fund. To test the sensitivity of our results to DRI's forecast, we considered alternative economic scenarios. For two alternative economic scenarios, we assumed lower rates of house price appreciation while keeping DRI's other forecast values ur,changed. We used a medium housing price appreciation scenario in which house prices rise at 2 percent per year less than DRI's fall 1989 forecast. This produces price increases in the range of 5 to 7 percent annually, which are consistent with short-term forecasts produced by the National Association of Realtors and DRI's more recent lo-year forecast. We alsc constructed a low housing price appreciation scenario in which house prices rise at 5 percent per year less than forecast by DRI last fall. At this level, the housing price appreciation rate would be less than the overall rate of inflation. A widely publicized academic study has suggested that long-term housing price increases may be at a level below the inflation rate. We also considered a third scenario that assumed that the country would experience a repeat of the economic conditions of the 1980s. During the early 198Os, interest rates rose to 15 percent and unemployment levels reached 10 percent. House price appreciation stayed below 3 percent per year until 1986. Under the base policy case (loan ceiling increases along with housing prices, while ARMS and downpayment requirements remain unchanged), the Fund would fare substantially worse with alternative economic scenarios than under DRI's trend economic 15 forecast. Under the low housing price appreciation scenario, the Fund's cash balance would be depleted by 1996. Under the other two alternative economic scenarios, the Fund balance would shrink to less than $4 billion. Medium house price appreciation would result in a $4.9 billion reduction in the 1998 cash balance relative to the trend economics case; the 1980s economic conditions scenario would result in a $5.0 billion reduction relative to that same base. The effects of the alternative policy options under the alternative economic scenarios are generally similar to their effects under the trend economic forecast. As shown in exhibit I, under the 1980s economic conditions scenario, increasing the loan ceiling would produce a larger 1998 cash balance than would be obtained under the base case with these economic conditions, although the difference is fairly small. At house price appreciation rates ranging from 2 to 4 percent a year, the Fund would have an estimated $2.9 billion deficit by 1998 compared to an estimated $5.0 billion deficit for the base case. That same policy change, in contrast, would increase the Fund's projected balance to $14.4 billion under DRI's trend economic condition, compared to a projected $8. 8 billion for the base case. Under all economic scenarios, revising the downpayment requirement would lower the Fund's balance relative to the base case, while the increased use of ARMS shows results nearly equal to the base case for all economic conditions. VIEWS ON WHAT FHA'S ROLE IN THE HOUSING MARKET SHOULD BE From the time of the Great Depression through the 196Os, FHA was the nation's primary insurer of mortgage credit for the purchase of single-family homes. With the subsequent growth of the private mortgage insurance industry, policy makers began to ask 1c what role FHA's programs should play in the housing market and how its responsibilities should differ from those of private mortgage insurers (PNIs). In response to conditions in the housing market and the economy in general, major changes in FHA's single-family insurance program since the 1970s included -- removing FHA's ceiling interest rate, -- increasing the maximum mortgage amount, -- encouraging the direct endorsement of FHA-insured loans by private lenders, -- collecting the full premiun at loan closing and allowing the premium to be added to the mortgage amount and financed over the life of the mortgage, -- liberalizing underwriting standards to enable more people to participate in FHA's program, -- liberalizing loan-to-value ratios, and -- allowing the use of adjustable rate mortgages. The extent to which FHA duplicates private sector activity was considered by the 1982 President's Commission on Housing. The Commission recommended that FHA should increasingly complement, rather than compete with, the private market. In the Commission's view, FHA should maintain its historic role in assisting low- and moderate-income families to achieve homeownership, while allowing the private insurance companies to take all home loans that they can and will insure. As part of our work, we solicited views on the impact of the proposed changes from knowledgeabie representatives of government agencies, academia, and industry organizations. Examples of the favorable views expressed to us follow: -- FHA should perform the same fundamental role in the 1990s that it has performed throughout its history--to provide homeownership opportunities for low- and moderate-income families. With higher home prices, it will be essential to help first-time, lower income families. But, providing this help should be done without a government subsidy. -- There is little evidence that PMIs will provide insurance to those families who can afford a downpayment of only 5 percent or less to purchase a home. Therefore, the need continues for FHA to assist these families. To do so, FHA must be able to maintain its participation in healthier markets. Without the pcsitive effects of cross subsidization, FHA would be unable to provide assistance tc riskier households. On the other hand, we also heard from those with serious misgivings about the proposed changes. For example, we were told that -- Raising the loan limit tc 95 percent of the area median house price would significantly increase FHA's market and risk exposure without proportionate benefits to moderate- or low-income households. -- It is time for FHA to refocus on its mission of serving people most in need. The best way to do this is to target FHA's assistance based on income because income is direct11 related to the house one can afford. Although the answers to policy questions on FHA's future role are not easy, we believe that the bases exist for formulating policies that take into account the potential impacts on the financial viability of the MM1 Fund. FHA'S MANAGEMENT PROBLEMS In November 1989 we testified that a number of financial management problems exist that HUD and FHA top management need to address if future losses are to be kept under control no matter what changes are made to the ceiling limits, downpayment requirements, or adjustable rate mortgages. GAO and HUD's Inspector General have been reporting on these management problems since the early 1980s. Among these problems were the need for (1) more effective monitoring of existing underwriting standards and procedures, (2) improved inter%1 controls, (3) more diligent fol low-up of audit findings, and (4) a HUD Chief Financ ial Officer and FHA Controller. The new management at HUD under Secretary Kemp has taken specific steps to address and to strengthen FHA's financial position. These include -- establishing a Chief Financial Officer for HUD and a controller within FHA, -- stepping up monitoring and enforcement activities, -- performing an independent actuarial analysis of the MMI and GI Funds, -- publishing annual audited financial statements, -- improving the efficiency and effectiveness of HUD audit resoluticn and follow:-uy' process, 12 -- establishing a task force to ensure full compliance with the Federal Managers' Financial Integrity Act, and -- reviewing lender requirements to ensure that only responsible and soundly capitalized firms participate in FHA programs. We are currently reviewing HUD's progress in implementing these management initiatives and plan to issue a report later this year. In conclusion, Mr. Chairman, if overall economic conditions remain general1 y favorable and house prices appreciate at a rate of 5 to 9 percent annually, the Fund's cash balance will remain positive under our base case and will be higher if the mortgage ceiling is raised. However, the additional insurance-in-force that this pclicy change will generate may expose the federal government to potentially greater financial risks. If the Fund is not actuarially sound, these risks may subject the federal government to enormous costs over the life cf the new insurance that will be created, unless the problem of actuarial soundness is addressed. Lower house price appreciation rates like those we are experiencing today will have a negative impact on the Fund if the> continue. For example, if during the 1990s the annual rate is less than 4 percent, the Fund will likely not be able to survive without U.S. Treasury assistance even if economic conditions remain generally favorable. On the basis of our analysis, we believe that while increases in the FHA's mortgage ceiling tc account for house price increases are necessary to prevent the deterioration of the Fund's balance 20 and to allow FHA to maintain its current share of the housing market, there is a need to proceed with caution on how hish to raise the mortsaoe ceilins. If the Fund is not actuarially sound, we believe the Congress should not raise the mortgage ceiling to the 95 percent level because this action may subject the federal government to enormous costs over the life of the new insurance that will be created. We believe that a decision to raise the mortgage ceiling, under these conditions, should be made in conjunction with a decision on how to resolve the problem of actuarial soundness so that the potential financial risks assumed by the federal government in the long run are adequately considered. If, on the other hand, the Fund is actuarially sound, we believe the Congress should consider raising the mortgage ceiling to 95 percent of a state's median house price because doing so will have a positive effect on the Fund's cash position and will allow homebuyers in high priced states to participate more actively in the FHA program. These issues will be addressed further in our upcoming report. This concludes my statement, Mr. Chairman. We will be pleased to respond to any questions you or members of the Subcommittee may have. 21 SII~IWI-~ of I’olicy Oplions hy Economic Scenario I<ntl of I+%cal Year Cash Ilalancc ($Million) IWiry Swnnrio 1vxv IVVfl 1991 IVVZ lW.3 lW4 tws 1996 1997 19R(k Fxnnom in Pljh Maintninz IVRR Marker Shnrr S5.SR6 S?.<T s6,osx $7,070 SR.oQS SA.06 S7.2vl S7,lRl S6.527 FHA lnc~aws Loan rcil!nR SS.MW %h.l52 67.07H 68.41X SV.hW %lO.?fiV 611.729 S8.727 67.999 FliA Rcdluccs Dounpnymcnc Rcqmrcmrnr 65,5Xfl sq.u4 Sh.07 I S7,003 S7.999 sx.3r I S6.917 S6,7R6 5.975 FHA l’rwic!c< 2/h AKMx $5,WV J5,Ydl S6.009 s6.Xoo S7.4 17 Sl.~SS 56.126 S6.079 ss.s60 Medium House Rirr Apprrriation FIIA Mammnc 19XX Markrt Shnrc S5.5Rfi x5.779 55.179 S4.719 SS.324 SS.27S S4.775 S4,3R2 WlS9 WA Incruws Loan C’ciling SS.N)R WI.49 I $6,573 vi.714 S7.RS9 W.277 SU.134 SR.123 SX.321 RIA Rrducc~ Downpnvmcnt Rcquirrmcnt $5,.7X6 55,769 S5.0X.S S4.671 fS,140 S‘l,QW s4.356 S3.RO7 53.406 WI/\ l’rov~drs 2/h ARhlr SS.SX’J 55,7H5 $5.1~1’) S44,X24 $5,253 55,177 S4.740 s4,.416 M,256 ha Itauw Price Appreciation %IA Mamtam I’JXX hlarkcr Show TS,~Xfi $5,74X $2,705 53.674 $3.479 52.376 MS6 ($1,510) (f3230) i3iA Incruccc Loan CcilmX Ss,fax b6.440 $6.032 SS.489 S5.76R S4.956 S3.13H $1.179 RIA Rrti~Jcr~ Downpaymrnt Rcqtrlrcmcnt S5,5Xh 55,737 S4.649 SL54R S3.279 %2.(K3 (‘6100) ($2,19X) (S44.062) T-‘11A F’mvitlcs ‘L/O ARM\ SS.SXV SS,7Sl S4,747 S3.71 I $3.194 y,,‘.IL $191 (%1,49X) ($1.1 NIUC, I.ml d I.I‘;c;II Yc~:lr IWS c .1x1111:1l~m~cII:IC nppr[l*tm:ltc.lv $0 ? hllltrrn FXHTRTT TT c C (sJelloa UO!l1!8) UO!l!SOd WeA Jea;r, J” PW 8667 Summary of Economic Variables Fcoeonir Vwiabkn 1989 I990 1991 1992 I993 I994 1995 1996 9.M 9.44 PM 10.26 IO.02 9.14 9.66 9.61 597 $104 $121 $130 $141 5153 Slfh 5 23 5.M 5.40 s.22 !I.39 5 17 5.28 5.17 574 w $85 $9 I sY9 sm 5117 5127 9.88 1092 12.95 15.12 15.38 12 85 12.49 11.74 $97 Slcn $117 $125 $128 SIW $139 $145 5.23 sm 7 16 7 78 9 87 9.33 7.4 I 7 20 374 $81 $89 $95 597 SW2 SIM SllO 9.88 9.44 9.66 10.26 IOSL? 9.74 9.66 9.61 597 $103 $107 $113 $120 $128 $137 SlUl 5.23 5.50 5.40 5.22 5.39 5.37 5.28 5.17 574 578 $82 SH6 SW $98 SUM $111 9.88 9.44 9.66 IO.26 IO.02 9.74 9.66 9.61 $97 $100 SlOl s104 Slrn Slll $115 $119 5.23 J.50 5.40 5.22 5.39 5.37 5.28 5.17 Ave. Nnmmal I mm Vhc $74 376 577 579 $82 $84 $88 WI NC&: Rurs UC exprcwcd in pcrccnts. Prices ud vrluo arc eaprrssed in SCUIO. Median Home Prices Al:lh~~l:l $K!.YX) Kentucky $59,700 North Dakota Al;ukn $100. loo I,oui\innna $8 1,200 Ohio hri7cm FIno,7(XJ hlainc $98,300 Oklahoma Arkan\;ts 67 1.200 * hlmland $111,600 Oreg0ll *Califomin % Ih2.7UO * hlass;ichusctt~ $lS9,WO Pennsylvania COlOGldO $89,800 Michipn $78,800 *Rhode Island *Connecticut 6171,1(x) Minnesota 595.800 South Carolina DdZlWCllY fA3,SW Mississippi $71,000 South Dakota *District of Columbia $133,000 Missouri $74,600 Tennessee Florida $90,900 Mont‘ana $73,900 Texas Georgia $92,800 Nebraska $65,500 Utah *Hawaii $170,700 Nevada $92,900 Vermont Idaho $79,800 * New Hampshire s I 34,500 * Virginia Illinois $91 ,RW * New Jersey $lSS,lW Washington Indiana $59,200 New Mexico $70,300 West Virgina Iowa $63,200 * New York $149,200 Wisconsin KnnsnT $7 1,100 North C:m~lin:t $X6,WO Wyoming
Impact of FHA Loan Policy Changes on Its Cash Position
Published by the Government Accountability Office on 1990-06-06.
Below is a raw (and likely hideous) rendition of the original report. (PDF)