A PROGRAM TO PROTECT MORTGAGE LENDERS
AGAINST INTEREST RATE INCREASES

(By Dr. James L. Pierce 1)

The wide fluctuations and protracted rise in interest rates over the last decade has brought into question the viability of thrift institutions (mutual savings banks and savings and loan associations) that are highly specialized in mortgage lending. These institutions have large portfolios of mortgage loans that carry interest rates far below current interest rates. When market interest rates rise, the low income from these mortgages has often made it difficult for the thrifts to offer interest rates on their deposits that are competitive with alternatives that are available to many savers. This situation has led to a Federal program of interest rate ceilings (Regulation Q) that gives thrifts an advantage relative to commercial banks and that often keeps interest rates on thrift accounts below market interest rates. When market interest rates are above Regulation Q ceiling rates, "Disintermediation" occurs in which depositors (or at least large and sophisticated depositors) withdraw their savings to purchase Treasury bills, money market mutual fund accounts and other higher yielding assets. Thus, when market interest rates are high, thrift institutions find themselves in an untenable situation because Regulation Q ceiling rates prevent them from competing with market interest rates. But even if there were not Regulation Q, the thrifts would still be in a difficult situation because the outstanding stock of low-yielding mortgage loans could prevent them from effectively competing for household savings.

It is this situation that has led to proposals that thrift institutions be allowed broader powers to invest in assets, such as consumer loans, that have shorter maturities than mortgage loans and to accept a wider range of deposits, including demand deposits. With these greater powers would come the lifting of Regulation Q ceiling rates. There appear to be two important concerns about the implications of these reforms. First, in light of the large portfolios of low-yielding mortgage loans currently held by the thrifts, could they survive in a competitive world without Regulation Q ceilings until these loans mature or are paid off? Second, can we be sure that the thrift institutions will continue to play a key role in mortgage finance if they have wider investment powers? This study addresses these concerns and recommends solutions.

Some observers have recommended that thrift institutions offer variable rate mortgage loans as a solution to their problems. With variable rate mortgages, if market interest rates rise, interest rates on both new and outstanding mortgage loans would also rise. It is argued that vari

1 Consultant and Director of the FINE Study, Committee on Banking, Currency and Housing, U.S. House of Representatives.

able rates would make mortgage lending more attractive and profitable to thrifts and that with widespread use of these mortgages, the thrifts would not require the protection of Regulation Q. The proposal has been criticized, however, on the grounds that the variable rate mortgage loan shifts the risk and burden of interest rate increases from the thrifts to mortgage borrowers who can ill afford such risk. The interest rate expense of a mortgage is too large a proportion of the income of most households to subject these mortgage borrowers to such risk.

There will always be interest rate risk and someone must bear it. However, it is possible to determine who bears the risk and design programs to shift it away from those least able to shoulder it. If thrift institutions are to hold large proportions of their total assets in longterm mortgage loans, a way must be found to protect them against the risk of unexpected interest rate increases. Since it has been deemed socially undesirable to shift the burden of risk to households, and since the thrifts cannot compete for funds if they bear all the risk, there is only one alternative left-the government must bear at least some of the risk. It can do so by offering a form of interest rate insurance in which the Federal Government insures the thrifts against unexpected interest rate increases.

Such an insurance scheme should be designed in such a way that the government bears the risk but does not subject itself to large and continuing expenditures. This turns out to be quite feasible. When a thrift institution originates a new mortgage loan (with a fixed interest rate over the life of the instrument), the government could offer to the thrift institution another instrument with the same principle but with interest payment based on short-term interest rates. The thrift would turn over to the government the fixed-interest income from the original mortgage loan. Thus, the thrift has, in essence, obtained the income from a variable-rate mortgage and has given up the income from a fixed-rate mortgage. The thrift would continue to hold the fixed-rate mortgage, would service it, and would bear all risks of costs of default, late payment, etc.

The key to this proposal is tying the government's payments (to thrifts) to current short-term interest rates and tying the government's income to the less volatile long-term mortgage rate. By this arrangement the government insures the thrifts against any unexpected increases in interest rates. If interest rates rise, then the income of thrifts will rise commensurately. But if interest rates fall, the income of thrifts will also fall. Thus, it is only "up-side" risk that thrifts are guarded against. The public is unaffected by the arrangement in the sense that households would continue to be offered a fixedinterest rate, fixed-maturity mortgage loan. The public would benefit however, because the thrifts would be able to pay higher deposit rates when interest rates rise. Because the interest income obtained from their portfolios would rise with short-term interest rates, the thrifts could pay interest rates on their deposits that would be competitive with short-term market interest rates. The rationale for Regulation Q would disappear! Furthermore, by being able to compete effectively for savings, the thrifts would be able to maintain a more constant flow of funds into housing.

2 Thrifts might try to circumvent this décline in earnings by attempting to swap back to the original fixed-rate income. A solution to this problem is discussed later.

As a practical matter, it would be only the difference in income between a short-term interest rate instrument and a fixed-rate mortgage that would change hands in the transactions between the thrifts and the government. For any given mortgage loan which a thrift has entered into the insurance program, the fixed interest income would be compared to the income based on short-term interest rates; if the income from the fixed-rate mortgage loan exceeded the income computed from short-term interest rates, the thrift would pay the difference to the government. If the interest income from the fixed-rate mortgage were less than the income based on short-term interest rates, the government would pay the difference to the thrift.

It might appear on first blush that this arrangement would be a losing proposition for the government because interest rates have been on an upward trend for years. Eventually, it would seem that shortterm rates would exceed the the original fixed rate on the mortgage loan. It must be recalled, however, that if interest rates are expected to rise in the future, current long-term interest rates must exceed current short-term rates sufficiently to compensate lenders of long-term funds. This occurs because in the market, individuals have the option of either borrowing or lending long term or of borrowing and lending in a succession of short-term contracts. Because of this flexibility, the long-term interest rate will be an average of expected future shortterm interest rates. Thus, if short-term interest rates are expected to rise in the future, the current long-term rate will exceed current shortterm interest rates.

Lenders of long-term funds lose out only if they underestimate the rise in short-term interest rates. Such underestimates can and do occur. For instance, there was no way for lenders of long-term funds in the early 1960's to have anticipated the inflationary pressures and special factors that produced high interest rates in the late 1960's and most of the 1970's. Thus, these lenders underestimated the interest rate risk and have been hurt as a result. It is precisely this kind of injury that the proposal would eliminate for the lender. It is also true, however, that historically, investors have been roughly correct, on average, concerning their expectations of the future course of interest rates. And it is because of this that the government, on average, will not have large expenditures for this program. This assertion will be demonstrated below.

But before turning to that, it is necessary to discuss three important problems. The first problem involves the outstanding stock of mortgage loans. The scheme described above would protect thrift institutions from the risk of unexpected increases in interest rates on new mortgage loans because they could swap the fixed-interest income from new mortgages for income based on short-term interest rates. There still remains the large stock of mortgage loans with relatively low interest income held by thrift institutions. So long as these mortgage loans are held, the thrifts will have difficulty in competing for deposits when short-term interest rates rise above the average yield on the mortgage portfolio.

This problem can be solved by applying the same kind of swap arrangement to old mortgage loans. Here, however, the swap would not occur dollar for dollar. To do so would involve great expense to the government because short-term interest rates are, and probably

will remain, above the interest rates on mortgage loans granted prior to the mid-1960's. It is proposed, therefore, that the income stream based on current short-term interest rates offered by the government be auctioned off to thrift institutions. Thus, a thrift might be willing to exchange a $10,000 fixed-interest rate mortgage for an $8,000 instrument whose interest return would be based on short-term interest rates. In essence, the thrifts would offer their fixed-interest income from old mortgage loans at a discount in order to earn a return based on shortterm interest rates. The discount on the old mortgage loan should be sufficient to equate the income on the fixed-rate mortgage with the expected income on the government's instrument. It is the expected income on the government's instrument (which is tied to short-term interest rate movements) that is relevant because short-term interest rates would vary over the life of the contract. Thus, the actual income from this instrument could not be known with certainty. However, once the swap is made, the thrift would have protection against unexpected increases in interest rates. If interest rates rise by more than was expected, the income from the short-term interest rate contract would rise commensurately. Thus, with this income protection thrifts could afford to compete for deposits when interest rates rise. This special treatment of the stock of outstanding mortgage loans is best viewed as a method of expediting the transition to a regime in which Regulation Q is completely abolished. It need only be used once; as the old mortgage loans are paid off, only new mortgage loans would be eligible for the income swap with the government.

The second problem involves potential attempts by thrift institutions to avoid declines in income when short-term interest rates fall below mortgage rates on those instruments which were entered into the swap program. This avoidance could be accomplished by the thrifts if they were to institute an automatic refinancing of those outstanding mortgage loans in the income-swap program whose fixed income exceeded the income on the instrument based on short-term interest rates. The thrifts could offer a lower fixed rate to mortgage borrowers which would, under appropriate circumstances, still exceed short-term interest rates. The thrifts could then keep these loans out of the income-swap program and earn a higher income. If short-term interest rates subsequently rose, the outstanding mortgage loans outside of the program could be refinanced or rolled over again and the "new" instrument placed in the program. If this practice became widespread, the government would incur costs even when short-term interest rates fall below the rates received on outstanding mortgage loans. The government would not receive the income from the original fixed rate mortgage loan but from a loan with a lower fixed rate.

There appear to be two solutions to this problem. First, the government could prohibit the thrifts from engaging in the practices of refinancing or rolling over mortgage loans when these activities serve to circumvent the income-swapping program. Such prohibitions obviously would involve some cost of enforcement.

The second solution involves a different approach to the program. Rather than swapping income, the government could simply recognize the asymmetry of the situation and actually issue insurance against interest rate increases by paying to thrift institutions the difference between the rate on mortgage loans and short-term interest rates when the latter exceeds the former. No money would change hands when the

mortgage rate exceeded short-term interest rates. The government would receive in return an insurance fee that would be sufficient to make the program self-financing on average. The fee could be calculated from the historical average differential between short-term and long-term interest rates. This differential is appropriate because it represents the premium that lenders of funds at long-term have required to protect them against the risk of unexpected interest rate movements and of illiquidity. For example, the yield on long-term corporate bonds exceeded yields on 4-6 month commercial paper by an average of about 20 basis points over the period 1950-74. This differential. combined with the average maturity of mortgage loans, could be used as the basis for assessing insurance premiums. The total premium would have to be paid in advance to avoid the refinancing problems encountered in the income-swap program. Thus, the thrift institutions would have to pay, say, the discounted present value of 20 percent times the average maturity of mortgage loans on every dollar of mortgage loans insured. If the difference between short-term interest rates and long-term rates remains a good average predictor of future shortterm rates, the program should be self-financing on average.

The third problem has equal relevance for both the income-swap and the pure insurance program. Mortgage loans granted in states that have relatively low usury ceiling interest rates will complicate the situation. In those states where usury ceilings are binding, the fixed interest rate on new mortgage loans will be lower than in states in which such ceilings are not binding. The income-swap program would go into deficit sooner or the insurance payments would occur sooner if mortgage loan rates are kept artificially low by usury ceiling rates.

A promising solution to this troublesome problem would lie in using an index of the national average mortgage rate, rather than the actual mortgage loan rate in those states with binding ceiling rates, as the fixed-interest rate for either program. Thus, thrift institutions in states with binding usury ceiling rates would receive additional income under either the income-swap or the pure insurance programs only if short-term interest rates rose above the index of national average mortgage loan rates. This solution might appear to discriminate against institutions in those states. However, to do otherwise would give states a great incentive to set usury ceiling interest rates or mortgage loans arbitrarily low on the grounds that the federal government would make up the income differences to thrifts in the state. While it is difficult to calculate accurately what the future cost to the government would be of either the income-swap or the pure insurance program because the calculations would depend upon the assumed future course of both short-term interest rates and mortgage rates— it is possible to provide estimates of what the costs would have been to the government if the program had been introduced in 1950 and allowed to continue to the present. For purposes of this calculation, it is assumed that all income from mortgage loans outstanding at the beginning of 1950 were swapped dollar for dollar for income from the government's short-term interest rate instrument and all income on mortgage loans made after that date was also swapped. In practice, it is unlikely that all mortgage loan holders would have fully exercised

*Only the costs of the income-swap program are shown here. The calculations for the pure insurance program would have been similar.