Printer Friendly

State foreclosure laws, risk shifting, and the private mortgage insurance industry.

State Foreclosure Laws, Risk Shifting, and the Private Mortgage Insurance Industry


The large losses suffered by private and government mortgage insurers during the last several year has renewed in identifying the risk factors associated with residential lending. This paper focuses on the effect of state foreclosure laws on that risk. Data on loss rates incurred by PMIs over the period 1980-1986 reveal that a judicial foreclosure requirement and a statutory right of redemption add significantly to mortgage risk. Anti-deficiency judgment statutes preclude the amelioration of such risk. Estimates of the losses due to such laws are included


One result of the mid-localized recessions in the agricultural and oil sectors of the economy has been an increase in residential loan defaults and foreciosures. Both government (FHA and VA) and private mortgage insurance (PMI) company claims have risen sharply since the early 1980s. (1) In 1986 the combined ratio of the loss and expense ratios for all PMI's was a record 168.9 percent. By comparison, the ratio was only 55 percent in 1979. Direct losses paid as a percent of the previous year's risk in force were .17 percent for 1980 and 1.35 percent for 1986. By 1986 the weighted average risk-to-capital ratio for the industry reached a record 25.8:1, several companies required capital infusions from their parent companies, and others had their ratings reduced by the rating agencies. (2)

Besides localized recessions, other frequently noted risk factors have been mentioned as deepening the financial difficulties of the PMI industry. They include the introduction of new loan types (negatively amortizing Arms and GPMs), an increase in insured investor (non-owner occupied) properties, and a greater proportion of over-90 percent loan-to-value ratio loans originated by lenders.

The consensus appears to be that the PMI industry failed both to adequately underwrite the risks associated with the characteristics of loans they insured and to charge premiums which reflected those risks. Throughout the early and mid-1980s, PMI companies charged uniform premiums regardless of the loan's characteristics, including location of origination. (3) It should not be surprising that in the fact of uniform premiums, differential amounts of risk were shifted to the insurers.

Among the risks that may have been differentially shifted, there is one that has received little attention in the literature. That is the risk associated with the legal structure of the default-foreclosure process of the state within which the mortgaged property is located. This structure varies dramatically. The laws in some states facilitate the foreclosure process and thereby ameliorate lender-insurer losses. The laws in others, in an attempt to provide greater protection to the mortgagor, cause the process to be more lengthy and costly.

This study demonstrates the extent to which different foreclosure laws create loan risk differentials which have been refelected in the losses of private mortgage insurers. The period of the interest is 1980 through 1986. (4)

The manner in which state foreclosure laws affect residential loan risk is briefly discussed. A model is then presented wherein this risk is shifted disproportionately to private mortgage insurers that charge uniform premiums. Empirical evidence then shows that such risk caused substantial losses for PMI's from 1980 through 1986. Findings are then summarized and suggestions for future research are offered.

State Foreclosure Laws and Loan Losses (5)

Aside from minor differences, there are three primary distinctions in state foreclosure laws. They all affect the size of residential loan losses. First, states are almost equally divided between those that employ a judicial foreclosure procedure and those that employ a power-of-sale procedure. (6) In a property procedure a court orders a foreclosure and supervises the sale of the property and the disbursement of the proceeds. Because they are more time consuming and necessitate the use of legal professionals, judicial procedures are more costly. Time consuming methods of foreclosure impose greater costs on lenders (foregone interest and real estate owned, REO, expenses such as hazard insurance, property taxes, maintenance, and so forth). A non-judicial procedure allows the foreclosure to be conducted expeditiously without a court proceeding, often through a trustee's sale.

Second, the laws in 29 states provide for a statutory right of redemption: a period of time subsequent to foreclosure (which varies by state with one year being the norm) during which the mortgagor can redeem the property by paying delinquent interest, penalties, and legal costs. (7) This provision can be particularly costly to lenders if the owner is granted the additional right to occupy the property during this time. The existence of the statutory right of redemption has the additional effect of reducing the interest of potential buyers of the property at a foreclosure sale, lowering the liquidation price that the lender would otherwise receive.

Third, although most states allows for deficiency judgments, six states (including California) do not. A deficiency judgemtn arises when the proceeds from the foreclosure sale are insufficient to satisfy the loan balance (which is usually the case). Equipped with a deficiently judgment, a lener (insurer) can proceed to recover directly against the mortgagor's personal assets. In some cases the cost of pursuing the judgment, combined with the paucity of personal assets of the mortgagor, may discourage the implementation of the judgment. As a results, deficiency judgments may not be pursued in every case where they are allowed. However, for some cases the threat of a judgment will often suggest a settlement. The lender, in order to avoid the cost of litigation and the borrower, to avoid a bad credit rating, may agree on a partial payment of the deficiency. Therefore, in those states where its use is universally prohibited, lender losses will be, ceteris paribus, greater. (8)

In summary, because a judicial procedure and a statutory right of redemption lengthen the foreclosure process and delay the liquidation of the property, losses are greater in states which require the former and grant the latter. A prohibition on deficiency judgments precludes any amelioration of these losses.

Some Initial Evidence

A study commissioned by the Federal Home Bank Board (FHLLB) (1975) confirmed the notion that state laws which retard the foreclosure process impose greater losses on lenders. The study examined the losses on 30 foreclosed loans of each of two thrifts in Illinois and two in Texas. At the time of the study, Illinois had a judicial procedure and a one-year statutory right of redemption. Texas employed a non-judicial procedure and granted no statutory right of redemption. In addition to the unpaid balance, the lenders' exposure on the loans included: foregone interest, attorneys' fees, court costs, property taxes, repairs, hazard insurance, and indirect costs. When the liquidation value of the property and rental revenue were deducted from these costs, the average loss in Illinois was 53 percent of the average loan balance. In Texas it was only 6 percent. Though the study was based on a relatively small sample, it nonetheless concluded that there was strong evidence that the magnitude of the financial impact of foreclosure is a function of state foreclosure law.

Private Mortgage Insurance and Risk Shifting

The typical mortgage insurance contract provides for the full payment of losses up to a minimum percent of the exposure. (9) For losses greater than the minimum percent, the contract takes on a coinsurance character. The loss is generally the exposure less the liquidation value of the property. The exposure is defined as:

E = U+F+T+Z+P+ Min[A,.03(U+R)] - R


U = unpaid loan balance,

F = foregone interest,

T,Z,P = property taxes, hazard insurance, and property preservation expenses (not major repairs),

A = attorneys' fees (limited to 3 percent of the unpaid balance plus forgone interest), and

R = rental revenue.

It is clear from the elements in equation (1) that the retardation of the liquidation process by arduous foreclosure proceedings increases the exposure.

Now the insurance indemnity is:

I = Min[E-H,[alpha]E] where H is the liquidation value of the houes and [alpha] is the co-insurance percentage. An increase in E via arduous foreclosure laws raises I. If the exposure is small such that the insurance payoff is E -- H, all of the lender's losses are covered. The lender may have little or no incentive to minimize the exposure. (10) However, if the exposure becomes greater, the coinsurance feature of the contract is more likely to be engaged, i.e. E -- H[is greater than][alpha]E. Since the lender bears some of these costs, it has an incentive to minimize them. it may be more likely to pursue or, in expectation of a settlement, threaten to pursue, a deficiency judgment. the result is that claims paid by private mortgage insurers, I, will be greater states require a judicial procedure, provide for a statutory right of redemption, and prohibit deficiency judgments.

Model and Empirical Results

In order to isolate the effect of state foreclosure laws on PMI losses, one must consider the remaining variables that can reasonably be expected to affect those losses. The following set of variables is proposed.

Residential Property Prices (Equity)

Studies by von Furstenberg (1969, 1970, 1974) have demonstrated an inverse relationship between the amount of equity in a property (as measured by the complement of the loan-to-value ratio) and the likelihood of default. The so-called equity theory holds that it is irrational for a mortgagor to default on a loan when there is positive equity in the property. States with appreciating property prices should experience fewer defaults and foreclosures than those with stagnant or declining values.

Ability-to-pay (Unemployment and Divorce Rates)

In contrast, the ability-to-pay theory suggests that defaults result from events such as involuntary unemployment and divorce which affect the ability or motivation of the borrower to meet the monthly payment. Theory would suggest, however, that there should be no default in the presence of such events if there is positive equity in the property. On the other hand, one could argue if one of the above events causes the borrower to fail to make payments for several months any positive equity may be eroded to the point where default becomes an optimal decision. In a test of the ability-to-pay and the equity theories, Jackson and Kaserman (1980) concluded that the equity theory is superior. However, Campbell and Dietrich (1983) found a statistical relationship between regional enemployment and default rates, and Foster and Van Order (1984) indicate marginal support for the effect of divorce rates on defaults. For completeness these variables are included.

Market Interest Rates

The relationship between the current market rate of interest and the loan's contract rate can affect the likelihood of default and foreclosure. The effect is not unambiguous, however. Mulherin and Muller (1987) suggest that if the market rate is substantially above the contract rate on the delinquent loan, the lender has an incentive to foreclose (rather than seek a non-forclosure alternative such as a forbearance) because it is able to lend the proceeds of the foreclosure, including insurance payments, at the higher market rate. Clauretie (1987) found support for this hypothesis in a model of foreclosure behavior. On the other hand, mortgagors have an incentive to make an effort to preserve the low-rate loan. In the situation where the market rate is above the contract rate, mortgagors have less incentive to default. The effect of changing market rates is an empirical one.

Time Lags

Studies by United States agencies (1962, 1963, 1966) indicate that defaults on residential properties peak between two and six years subsequent to origination. The available data set on mortgage insurer's losses is aggregate and does not include the date of origination of individual loans. The current study allows the property price appreciation and interest rate variables to take on a lag between two and six years. That lag is selected for which the t-value of the coefficient is the largest. It is expected that voluntary unemployment anddivorce would affect the likelihood of default within a year or so. These variables are entered with shorter lags.


The model to be tested with annual data is:

[L.sub.j,t] = [alpha] + [[beta].sub.1.P.sub.j,t - x] + [[beta].sub.2.R.sub.t - x] + [[beta].sub.3.[Delta]U.sub.j,t - x] + [[beta].sub.4.D.sub.j,t - x] + [[beta].sub.5.PS.sub.j] + [[beta].sub.6.SRR.sub.j] + [[beta].sub.7.DJ.sub.j] + [[epsilon].sub.t]


[L.sub.j,t] = the log of the loss rate for all PMI's in state j in period t, (11)

[P.sub.j,t _ x] = the percent change in residential property prices in state j from period t - x to period t: x is allowed to range from two through six,

[R.sub.t - x] = the level of market mortgage rates (national) in period t relative to period t - x: x is allowed to range from two through six: entered as [r.sub.t. - r.sub.t - x][/r.sub.t - x]

[[Delta]U.sub.j,t - x] = change in the unemployment rate in state j in period t - x: x varies from one to two years,

[D.sub.j,t - x] = divorce rate in state j in period t - x: x varies from one to two years,

[PS.sub.j] = dummy variable equal to one if a state has a power-of-sale foreclosure process and zero otherwise,

[SRR.sub.j] = dummy variable equal to one if a state has a statutory right of redemption and zero otherwise (alternatively, the length of the statutory right of redemption is included),

[DJ.sub.j] = dummy variable equal to one if a state prohibits a deficiency judgment and zero if otherwise, and

[[epsilon].sub.t] = error term.

The signs on [beta.sub.1] and [beta.sub.5] are expected to be positive while those on [beta.sub.4], [beta.sub.6], and [beta.sub.7] are expected to be negative.

The model is tested using annual data for the dependent variable for 51 jurisdictions from 1980 through 1986 (357 observations). Data for the explanatory variables are described in the appendix.

Empirical Results

Equation 3 was tested employing, alternatively, a dummy variable for the presence of a statutory right of redemption and the length of the statutory right to redemption (in months) as one of the three legal variables of interest. The results are displayed in Table 1.

With the exception of the incorrect sign on the unemployment variable, the explanatory variables behaved as hypothesized. The important non-legal variable is the measure of property price appreciation. This result supports the equity theory of default. The legal variables appear important in explaining insurers' losses over this period. Given that the average length of a statutory right of redemption where it exists is ten to 12 months, the coefficients on the alternative forms of this variable are consistent. Loan losses are also less where lenders can pursue a deficiency judgment or employ a power-of-sale foreclosure.

Estimate of losses imposed on lenders by variations in the legal structure of the foreclosure process can be made from these results. The effect of the various legal distinctions on the log of the loan loss rate (at the mean) and, in turn, the loan loss rate can be determined from the coefficients included in Table 1. These effects are summarized in Table 2. Given the amount of risk-in force for private mortgage insurers, loss rates imply certain dollar losses.

During the period of analysis, private mortgage insurers had a total risk in force that averaged approximately $40 billion. In 1986 approximately 30 percent of insurer's risk-in-force was allocated to states with a statutory right of redemption. (12) Furthermore, the existence of this redemption period adds approximately 21 basis points to insurers' loss rates. With an assumed $40 billion national risk exposure, insurers' losses can be estimated to be $25 million annually as a result of this provision ($40 billion X .3 X .0021).

Similarly, 44 percent of the national risk-in-force has been allocated to states which employ a judicial foreclosure. Losses as a result of this provision are estimated at $26 million annually ($40 billion X .44 X .001463).

Finally, the prohibition of the use of a deficiency judgment by six stats that represent 18 percent of the risk-in-force add 36 basis points to the loss rate or approximately $26 million annually ($40 billion X .18 X .0036).

Policy Implications

There are a couple of implications of this study. first, since the foreclosure laws of the state within which the mortgaged property lies affect the risk of the loan, private mortgage insurers should, and perhaps will, consider pricing the premium to rflect that risk. If this becomes a standard practice among mortgage insurers, then decidedly more of the risk will be shifted bck to the borrowers that pay the premium. Differential premiums which reflect the legally induced risk will then lead consumers (borrowers) to reflect on the laws of their state which may presently favor them by impeding the foreclosure process.

Second, an awareness of the impact on loan losses from differential foreclosure laws may lead to greater state acceptance of the Uniform Land Transactions Act. This Act, would cause foreclosure laws to become more uniform while still affording protection to the borrower. The Act encourages the use of the non-judicial foreclosure procedures, eliminates the statutory right of redemption, and makes uniform the use of a deficiency judgment.


This study demonstrates that the variation in state laws regulating the foreclosure process causes differential rates of losses on residential loans. The evidence comes from the loss rate (claims as a percent of risk-in-force) for private mortgage insurers from 1980 through 1986. Laws which benefit the mortgagor through slowing the foreclosure process or limiting remedies through deficiency judgments result in greater losses for insurers and, because of the coinsurance nature of the contracts, lenders. The three main legal distinctions discussed likely added approximately $ 75 million annually to private mortgage insurers' losses over the period.

Future research sshould extend the investigation of this issue by analyzing loss data from lenders or losses imposed on government insurance agencies. There may be some structural differences between PMI and government mortgage insurance. Since the FHA rules for the administration of home mortgages provide for the termination of a statutory right of redemption for delinquent loans where a deed-in-lieu of foreclosure is obtained, FHA losses may be ameliorated. Analysis of FHA claims should reveal any structural differences.


1. Legal Variables: Foreclosure-Redemption Summary, in C. F. Sirmans, Real Estate Finance, pp. 83-85 or in D. Sirota, Essentials of Real Estate Finance, 4th edition (Chicago: Real Estate Education Company, 1986), p. 252. Entered as dummy variables or years for length of time variables.

2. Mortgage Rate: Average Effective Interest on Loans Closed, FHLBB Series, Federal Reserve Bulletin, various issues. Entered as [(r.sub.t -- r.sub.o)/r.sub.O.]

3. Residential Property Prices: Two data sources: (a) Value of New, Privately Owned, Single Family Residents by State From Housing Units Authorized by Building Permits and Public Contract, Annual Issues, U. S. Dept. of Commerce Bureau of the Census; (b) Monthly Report, Existing Home Sales, National Association of Realtors, various issues (data here are on a regional, not a state basis). Entered as percent increase in period t over t -- x.

4. Unemployment: entered as change in rate of unemployment as reported by the BLS, various issues and divorce rate entered the rate as reported in Vital Statistics, various issues.

5. Loss Rate: Direct Losses paid as a percent of previous End of Year Risk, 1980-1986, Moody's Annual Report of PMI Industry, Table 5, P. 18.

(1) Losses for private mortgage insurance companies have been greatest for loans made in states where mineral (including oil and gas) extraction is a dominant industry. Six states -- Alaska, Texas, Louisiana, Oklahoma, Colorado, and Wyoming -- accounted for 50.8 percent of all direct losses incurred in 1986 although they represented only 18 percent of the industry's risk.

(2) The loss ratio is the dollar amount of claims incurred as a percentage of earned premiums and the expense ratio represents the costs of obtaining new business as a percentage of premium written. The combined ratio can be used to indicate the claim-paying ability of an insurer. The risk-to-capital ratio is the risk-in-force (defined below) divided by the company's capital. These values are from Moody's Investors Service (1987).

(3) In a recent issue of Mortgage Banking, two articles suggested that the era of uniform pricing by PMIs may be passing. Blood (1987) indicated that PMI losses in the 1980s have caused insurers to reconsider the manner in which they underwrite loans in the various states. Kureera (1987) indicated that at least one major company had begun to rate loans by state and that territorial pricing was the next logical step.

(4) Although modern PMI (in the post-war era) began in 1957, the volume written was negligible until 1975. The most dramatic increase occurred in the early to mid-1980s.

(5) Only a sketch of foreclosure laws is presented here. The interested reader is referred to Durham (1986) for a more detailed presentation of the subtle distinctions of state foreclosure laws.

(6) Technically, the judicial procedure is available in all jurisdictions but is the only method allowed in 23 states. In the remaining states lenders may use a power-of-sale procedure and, since it is less costly, generally provide for its use in the mortgage instrument.

(7) This period of time is termed a statutory right of redemption because it is created by state statutes. All states provide for an equitable right of redemption which allows the mortgagor to redeem the property prior to foreclosure by paying all delinquent costs.

(8) The value to insurers of a deficiency judgment is indicated by the terms of the typical PMI contract. Generally, a lender's claim is limited insofar as the amount of attorney's fees that may be recovered. A larger amount will be paid, however, if required to pursue a deficiency judgment.

(9) For an example of a typical contract, see Mortgage Insurance Companies of America (1988).

(10) The lender may, for example, fail to vigorously pursue the borrower for delinquent payments. Muller and Mulherin (1988) demonstrate that under these circumstances there is an incentive conflict. Lenders may fail to collect delinquent payments or make needed repairs on the property. Their model of incentive conflicts lends support to the risk shifting behavior of lenders.

(11) The loss is defined to be the total claims divided by the risk-in-force at the end of the previous year. The risk-in-force is the product of the loan amount and the coverage ratio. The log of this rate is taken because the rate is bounded by zero on the downside and is positively skewed Inspection of the data indicate that the loss rate is log-normally distributed. A normal distribution of the dependent variable is required for application of an OLS procedure.

(12) Data for 1986 from Moody's Investors Service (1987). Data for previous years are unavailable. However, this is no reason to expect that the proportion of risk-in-force allocated to states with various legal provisions changed substantially over the period.


[1.] Blood, Roger, 1987, "Which Way Will the Pendulum Swing?" Mortgage Banking, 47: 8-19.

[2.] Campbell, T. S. and J. K. Dietrich, 1983, "The Determinants of Default on Insured Conventional Residential Loans," Journal of Finance, 38: 1569-81.

[3.] Clauretie, Terrence, 1987, "The Impact of Interstate Foreclosure Cost Differences and the Value of Mortgages on Default Rates," Journal of the American Real Estate and Urban Economics Association, 3: 152-67.

[4.] Durham, James, 1986, "In Defense of Strict Foreclosure: A Legal and Economic Analysis of Mortgage Foreclosure," South Carolina Law Review, Spring: 461-510.

[5.] Evans, R. D., et. al., 1985, "Expected Loss and Mortgage Default Risk," Quarterly Journal of Economics and Business, 24: 75-92.

[6.] Foster, Chester and Robert, Van Order, 1984, "An Option Based Model of Mortgage Default," Housing Finance Review, 8: 351-72.

[7.] Jackson, J. R. and D. L., Kaserman, 1980, "Default Risk on Home Mortgage Loans: A Test of Competing Hypotheses," Journal of Risk and Insurance, 47: 678-90.

[8.] Kurcero, Jeffrey, 1987, "Pricing Mortgage Insurance for Adequate Return," Mortgage Banking, 73-6.

[9.] Moody's Investors Service, 1987, Special Report: Moody's Annual Report of the Private Mortgage Insurance Industry, New York, N.Y.

[10.] Mortgage Insurance Companies of America. Fact Book and Directory, 1987-1988, Washington, D.C.

[11.] Mulherin, J. H. and W. J., Muller 1987, "Volatile Interest Rates and the Divergence of Incentives in Mortgage Contracts," Journal of Law, Economics and Organization, 30: 99-115.

[12.] Mulherin, J. H. and W. J., Muller, 1988, "Resolution of Incentive Conflicts in the Mortgage Industry," Journal of Real Estate Finance and Economics, 1: 35-46.

[13.] Touche, Ross and Company, 1975, The Cost of Mortgage Loan Foreclosure: Case Studies of Six Savings and Loan Associations.

[14.] United States Government. Veterans Administration. Report of Loan Service and Claims Studies, 1962, Washington, D.C.

[15.] United States Government. Federal Housing Administration, FHA Experience with Mortgage Foreclosures and Property Acquisitions, 1963, Washington, D.C.

[16.] United States Government. Housing and Home Finance Agency, Mortgage foreclosure in Six Metropolitan Areas, 1963, Washington, D.C.

[17.] United States Government, U.S. Bureau of Labor Statistics, Geographical Profile of Employment and Unemployment, 1985, Washington, D.C.

[18.] United States Government, U.S. National Center for Health Statistics, Vital Statistics of the United States, Washington, D.C.

Terrence M. Clauretie is Professor of Finance at the University of Nevada, Las Vegas. He thanks Robert Chatfield for helpful comments on an earlier version of this article.
COPYRIGHT 1989 American Risk and Insurance Association, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Clauretie, Terrence M.
Publication:Journal of Risk and Insurance
Date:Sep 1, 1989
Previous Article:Conflict and channel management in property-liability distribution systems.
Next Article:Estimating the present value of future income losses: an historical simulation 1900-1982.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters