PMI Reform: Good Intentions Gone Awry
by Stanley D. Longhofer
March 15, 1997

Introduction
What Is PMI?
So What's the Problem?
Congress to the Rescue
The Proverbial "But..."
What about Notification?
Conclusion
Footnotes

Last April, the House of Representatives overwhelmingly passed the Homeowners Insurance Protection Act (H.R. 607) with the noblest of intentions—relieving the burden of as many as 250,000 home owners who currently pay for private mortgage insurance (PMI), even though their loan agreements may stipulate that such insurance is no longer necessary.1 Similar legislation, the Homeowners Protection Act of 1997 (S. 318), is now being considered by the Senate Committee on Banking, Housing, and Urban Affairs (Banking Committee).

Unfortunately, like many well-intentioned interventions into private markets, this legislation could actually hurt the very borrowers it is intended to help, making mortgage loans more costly and reducing their availability for all borrowers. In this Economic Commentary, I look at the role of PMI in the mortgage industry and consider the fundamental economic problem this legislation is intended to solve. I also discuss some of the unintended consequences that passage would likely entail.

Return to Index

What Is PMI?

PMI plays a crucial role in helping millions of Americans realize the dream of home ownership. According to industry sources, PMI companies insured nearly $127 billion in new loans in 1996, ending the year with more than $513 billion of insurance in force (see table 1). Despite the fundamental importance and prevalence of PMI, however, many current and potential homeowners have little understanding of its purpose or the way it works.

TABLE 1   PRIMARY INSURANCE ACTIVITY (NET)

Year Number of Applications Number of Certificates New Insurance Writtena Total Insurance in Forcea
1991 681,508 494,259 53,967 255,917
1992 1,235,033 907,511 101,047 284,552
1993 1,614,513 1,198,307 136,767 337,708
1994 1,373,502 1,148,696 131,402 406,250
1995 1,281,491 960,756 109,625 460,817
1996 1,371,927 1,068,707 126,972 513,240
a. In millions of dollars.
SOURCE: Mortgage Insurance Companies of America.

Mortgage lending involves a variety of risks. At its core, however, are two fundamental questions: How likely is the applicant to default on his or her loan, and how large will the lender’s loss be in the event of default? It is well understood that a borrower’s loan-to-value (LTV) ratio is intimately related to both of these risks.2,3 Because borrowers with low LTV ratios have substantial equity stakes in their homes, they are more likely to take every step possible to avoid default. In addition, this equity position provides a cushion between the value of the loan at risk and the value of the house on which the mortgage is held, thereby maximizing the lender’s recovery should the homeowner default. As a result, lenders have historically been leery of making loans to borrowers with LTV ratios above 80 percent.

Unfortunately, the greatest single barrier to homeownership for most potential home buyers is lack of a sizable down payment. PMI can help fill this gap by covering much of a lender’s losses on a property if the borrower defaults.4 As a result, millions of home buyers are able to obtain mortgages with LTV ratios of 85, 90, or even 95 percent. Although the premiums for PMI increase the borrower’s monthly mortgage payment, this insurance fills a vital need in the market; without it, many home buyers would be unable to obtain mortgages.

Return to Index

So What’s the Problem?

Although having PMI can be enormously beneficial for borrowers when they first obtain their mortgages, its value diminishes over the life of the loan. This is because as a borrower’s equity position in his home increases, the risk of default declines. In recognition of this fact, lenders generally allow this insurance to be canceled once the borrower’s equity in the house exceeds 20 percent of its current market value.5 Nonetheless, many borrowers who could cancel their PMI fail to do so, presumably because they are unaware of this option. As a result, less sophisticated borrowers may be paying for insurance they no longer need. Worse yet, they could actually be cross-subsidizing those more sophisticated borrowers who cancel their insurance optimally.

How does this occur? For any given set of risks they insure, PMI companies set their premium rates so that the present value of these premiums gives them a competitive return on their investment. Because some borrowers fail to cancel their insurance as soon as they are eligible, PMI companies are able to charge all policyholders a lower rate than they would if everyone canceled their policies optimally. As a result, the total cost of PMI is lower for those who cancel their policies earlier. In other words, less sophisticated borrowers cross-subsidize those with more financial savvy.

Return to Index

Congress to the Rescue

The Homeowners Insurance Protection Act attempts to remedy this problem through two primary provisions. First, it would require loan servicers (those who collect a borrower’s payments for the mortgage holder) to notify borrowers—both at the time the loan is originated and annually thereafter—of their right to cancel PMI once they have achieved sufficient equity in their homes. More important, it would require lenders to cancel a borrower’s PMI automatically once the principal balance falls below 75 percent of the home’s original value.6

The good intentions behind this legislation are obvious. After all, much of federal housing policy is designed to help low-income and less financially astute families achieve the dream of home ownership, and it is precisely these families that are less likely to be aware of their option to cancel their mortgage insurance. By reminding borrowers of their right to cancel PMI, and by automatically doing so at a specified time, proponents of these bills hope to eliminate this undesirable cross-subsidy, making the mortgage market "more fair" and reducing the costs of homeownership for those borrowers that most federal programs are intended to help.

Return to Index

The Proverbial "Butů"

Unfortunately, governmental intrusions into the workings of private contracts can often have perverse, unintended consequences, regardless of their original objectives. The automatic cancellation requirement of this legislation provides a clear example of this problem. Because cancellation would be based on a government-determined equity level (as opposed to one determined competitively in the market), it would alter the terms and conditions under which creditors and servicers willingly participate in private mortgage contracting. By reducing the set of options available to consumers, this legislation would increase the costs of mortgage insurance and reduce its availability.

Even worse, cancellation would be based on the original value of the mortgaged property (rather than on its current market value), creating dramatic new risks for mortgage holders. Property values are not static over time. Although home prices generally rise, in some markets they can decline substantially over the life of a loan. As a result, the proposed legislation could mandate cancellation even for borrowers whose true LTV ratio is higher than it was when the mortgage was originated. In other words, mandatory cancellation could occur at a time when the borrower’s risk of default is greater than it was when the loan was obtained.


FIGURE 1   LTV RATIOS OVER TIME

SOURCE: Author's calculations

Figure 1 demonstrates this problem. Suppose a borrower were to take out a $100,000, 30-year fixed-rate loan with a 7.5 percent interest rate. Assuming he made a 5 percent down payment, his initial LTV ratio would be 95 percent, and most lenders would require him to purchase PMI. According to the House version of the bill, this insurance would be canceled automatically after 13 years and 8 months, assuming the homeowner made only the required monthly payment. If his property value declined as little as 2 percent per year, however, his true LTV ratio at the time of cancellation would be 98.52 percent, higher than when the loan was originated.7,8 Mandating PMI cancellation for such a borrower would impose stark new risks on lenders, the costs of which would inevitably be borne by the borrowers themselves.

Even if the law based automatic cancellation on the borrower’s true equity position, lenders might still want to continue PMI for borrowers whose payment histories have been less than ideal. Indeed, some borrowers are required to purchase PMI even when their initial LTV ratio is less than 80 percent, presumably because lenders recognize that they pose a greater default risk than most.9

Unfortunately, this legislation makes little allowance for other factors that affect the riskiness of a loan. Although the House version does exempt delinquent loans from automatic cancellation, this provision applies only to loans that are currently delinquent; once the loan is no longer past due, cancellation is immediate. In other words, a chronically delinquent borrower could have his PMI canceled even though he still represented a serious default risk. Forced cancellation of insurance when it is still warranted would unnecessarily increase the initial costs of obtaining a mortgage and could potentially reduce credit availability for all home buyers.

It is also worth asking why Federal Housing Administration (FHA) insurance is exempted from the mandatory notification and cancellation required by this legislation. Currently, FHA insurance is expressly non-cancelable, perhaps in recognition of the argument that this translates into lower premiums for low-income FHA borrowers.10 Why should the private market be prohibited from offering the very contracts the government itself values?

Return to Index

What about Notification?

On the surface, the notification requirement seems much more reasonable. After all, what can be wrong with telling consumers about the options available to them? The answer is that doing so could be unnecessarily burdensome. In the modern mortgage market, loans are often originated, serviced, and held by different firms, and large servicers typically process payments for literally hundreds of investors, each with their own cancellation requirements. Complying with the annual notification provision of this legislation would force these servicers to create computer systems that could track each borrower’s eligibility to cancel PMI, regardless of which investor owns his loan. Such an endeavor would obviously be nontrivial. In addition, lenders would be required to notify borrowers of the conditions under which PMI could be canceled when the loan is originated. 11 Unfortunately, at that time it is often uncertain who the ultimate holder of the mortgage will be, making compliance with this aspect of the legislation nearly impossible. Thus, it is clear that implementing the necessary changes in industry practice would impose new costs on the mortgage industry—costs that would likely be larger than most suspect and that would ultimately be passed on to borrowers in the form of higher mortgage interest rates and fees.

The fact that compliance is costly does not, in and of itself, imply that mandatory notification is a bad idea. Unfortunately, the benefits from this legislation would be much smaller than proponents believe. First of all, industry statistics demonstrate that the magnitude of the cross-subsidy is well below what supporters of this legislation suggest. According to the Mortgage Insurance Companies of America, of the roughly 5 million policies in effect at the end of 1996, only about 5 percent are sufficiently seasoned to be eligible for automatic cancellation.12 In other words, fewer than 250,000 home owners would currently benefit from this legislation, not the millions that proponents often cite.13 Furthermore, premium rates for many PMI policies are automatically reduced after 10 years, limiting the total number of dollars collected from borrowers who may be eligible to cancel their insurance.14

Finally, even before this legislation was introduced, Fannie Mae (the largest secondary market agency) was already in the process of revising its cancellation guidelines to ensure that consumers are aware of their option to cancel their PMI. These guidelines, negotiated by agents of mortgage investors and other market participants, would likely be better targeted at the source of the problem and more flexible to changing market circumstances than would a statutory solution. Given the relatively small benefits that mandatory notification would provide, it is apparent that even minor compliance costs would outweigh them.

Return to Index

Conclusion

Making the mortgage market more fair and accessible is an important goal in federal housing policy. Although current borrowers who no longer need PMI would certainly find notification and automatic cancellation a welcome surprise, it is nonetheless essential to consider all the ramifications this legislation would have on future, as well as current, borrowers.

Too often, consumer protections hurt the very people they are supposed to help. Mandatory notification and cancellation of PMI would impose large compliance costs and new risks on the mortgage industry—costs that would be passed on to all future homeowners, including those less sophisticated borrowers the regulations were intended to benefit.

Return to Index

Footnotes

1. This article went to press on July 14, 1997.
Back to commentary

2. The LTV ratio is calculated by dividing the amount of the loan by the lesser of the sale price of the property being mortgaged and its appraised value. A borrower with a 20 percent down payment is often said to have an 80 percent LTV ratio, although this may not always be accurate.
Back to commentary

3. For a review of the literature on mortgage default, see Roberto G. Quercia and Michael A. Stegman, "Residential Mortgage Default: A Review of the Literature," Journal of Housing Research, vol. 3, no. 2 (1992), pp. 341–79. See also Robert Van Order, "The Hazards of Default," Secondary Mortgage Markets, Fall 1990, pp. 29–31.
Back to commentary

4. Federal Housing Administration insurance and Veterans Administration guarantees provide essentially the same benefits, but at a much higher cost than private insurance.
Back to commentary

5. The two secondary market agencies, Fannie Mae and Freddie Mac, will not purchase mortgages whose PMI is not cancelable, presumably to minimize the prepayment risk that mortgages with non-cancelable PMI would entail.
Back to commentary

6. The Senate version of the bill uses an 80 percent threshold.
Back to commentary

7. Under the Senate bill, cancellation would occur after 11 years and four months, and the borrower’s true LTV ratio at that time (assuming 2 percent depreciation in property values) would be more than 100 percent.
Back to commentary

8. Although housing prices rarely decline by small amounts over long periods (as assumed in this example), large drops are not uncommon. For example, median housing prices in the Boston market fell 18.3 percent between 1988 and 1991, and the Dallas–Fort Worth area suffered a 23.7 percent decline from 1989 to 1994. Given that these are changes in median house prices over an entire metropolitan statistical area, it is easy to see how prices in individual neighborhoods could be even more volatile.
Back to commentary

9. For evidence of this fact, see Tim S. Campbell and J. Kimball Dietrich, "The Determinants of Default on Insured Conventional Residential Mortgage Loans," Journal of Finance, vol. 38, no. 5 (December 1983), pp. 1569–81.
Back to commentary

10. Again, since they are paid over a longer expected horizon, the annual premium payments for FHA insurance are lower than they would be if the insurance were cancelable.
Back to commentary

11. This requirement is in the Senate version of the bill.
Back to commentary

12. This figure is based on the 80 percent LTV threshold in the Senate bill; the number of loans affected by the House bill would be much smaller. This estimate was provided by William H. Lacy (chairman and CEO of the Mortgage Guarantee Insurance Corporation) during his testimony before the Senate Banking Committee on behalf of the Mortgage Insurance Companies of America.

It is also worth noting that these remaining policyholders are among the riskiest. Since most PMI policies are canceled because the borrower refinances, those who keep their insurance are more likely to have suffered declines in income and/or housing value that make them ineligible to refinance their mortgages.
Back to commentary

13. For example, in his testimony before the Senate Banking Committee, R. Layne Morrill, 1997 president-elect of the National Association of Realtors, claimed, "It is estimated that millions of American homeowners are paying for unneeded private mortgage insurance."
Back to commentary

14. Ironically, if the magnitude of this cross-subsidy were as large as proponents of the legislation suggest, notification could actually hurt the borrowers it intends to help by raising PMI premium rates. Why? If notification caused additional borrowers to cancel their insurance once they became eligible, PMI companies would be underwriting essentially the same risks, but would receive premium payments for a shorter period. As a result, PMI premium rates would likely rise. Although the total cost (present value) of this insurance would be lower for less sophisticated borrowers, if they were liquidity constrained (with little income available for housing expenses), the higher premium payments required each month could make them settle for a less expensive house than they might otherwise choose, and some could be forced out of the market altogether.
Back to commentary

Return to Index


Stanley D. Longhofer is an economist at the Federal Reserve Bank of Cleveland. Many of the ideas in this Economic Commentary were originally presented in an article (co-authored with Charles Calomiris) that appeared in Investor’s Business Daily, April 18, 1997. The author would like to thank Glenn Canner, Steve Moore, Basil Petrou, and Mark Sniderman for helpful comments.

The views stated herein are those of the author and not necessarily those of the Federal Reserve Bank of Cleveland or of the Board of Governors of the Federal Reserve System.


[Abstracts] [FRB ClevelandHomepage] [Corporate Communications Homepage] [Research Homepage]

http://www.clev.frb.org/research/com97/