association, the committee, and the administration in title 6 of this bill, which has to do with revising the authority of the Home Loan Bank Board.

(The statement of Mr. Wellman is as follows:)

STATEMENT OF CHARLES A. WELLMAN

My name is Charles A. Wellman. I am executive vice president of the Glendale Federal Savings & Loan Association, a Federal savings and loan association with assets in excess of $90 million. We are engaged in making all types of loans, including FHA, title II, VA 501, title I, FHA property-improvement loans, our own unsecured property-improvement loans, and conventional mortgages with open-end provisions. My comments today with respect to the proposed housing bill of 1954 are confined to 3 sections, to title I, and more specifically to the changes in title I of the section 203 FHA for 1- to 4-family residences, on the provision of title III respecting the reorganization of the FNMA and the provisions of title VI respecting the Home Loan Bank System.

Title I of the proposed bill makes or authorizes the President to make substantial changes and extensions in the FHA loan programs. The effect of these changes will be a considerable expansion of the use of the old title II, section 203 FHA loan. You certainly could not extend the term of a loan plan up to a maximum of 30 years, lower the downpayment requirements, extend the maximum benefits to existing housing as well as new construction, relieve the previous statutory inhibition against the use of FHA for refinancing, and not have a substantial increase in the use of the old 203 loan. Determining the desirability of making these changes is a difficult task. Certainly, however, making it easier for individuals to buy houses, for builders to build houses, for realtors to sell houses, and for lenders to make loans on houses cannot be the sole objectives of any housing legislation. The costs that are involved must also be considered and there certainly is a price tag attached to the proposed extensions which must be critically examined. With respect to the changes in title I, there are two specific elements of price which need special attention. One element is the potential cost to the Federal Government of the proposed extensions; the second element is the possible effect of the extensions on the distribution of power over and responsibility for the mortgage credit structure of the country as between the Federal Government and the private home-financing industry.

The Federal Government has a direct liability for all insured losses suffered by private lenders under the FHA plan to the extent that the specific FHA insurance fund is inadequate. Under the mutual mortgage insurance plan, the Federal Government, in effect, occupies the role of backstop. How strenuous and taxing a job the role of backstop is depends on how good a player is the insurance fund. Measuring the performance capacity of the insurance fund to cover fully all insured losses that might occur is, of course, a very difficult and complex task. No matter how carefully and comprehensively it is done, you will never come up with a precise mathematical answer. Nevertheless, unless this element of price, namely the possible cost of the Federal Government, is going to be brushed aside as totally irrelevant, such a task of measurement, with all of its limitations, must be undertaken. To measure the performance capacity of the insurance fund requires an analysis of 3 elements or 3 parts. One is the estimated risk inherent in the total portfolio, the second is the amount in the insurance reserve, and the third is the relationship between these two. The risk inherent in the portfolio is a constantly changing thing. It is affected by many factors. Most important of all, however, it is affected by the rate of growth of the portfolio and the risks inherent in the individual new loans being made. For example, if you had a total portfolio of $100 million of fully amortized loans and you made no new loans, the risk in that portfolio would diminish, for the unpaid loan balances would be consequently reduced. On the other hand, if you had this same portfolio of $100 million and merely replaced the reductions in it by new loans, and the risks in the new loans were no greater than the reducing risks of the existing loans, the risks of the total portfolio would remain the same. These two illustrations, however, are not characteristic of a dynamic mortgage portfolio, nor are they characteristic of the FHA-insured operations in the past or as they are contemplated under this bill.

The bill contemplates an expansion of the total outstanding volume of mortgages insured and at the same time it changes the risk characteristics of the new

loans to be made. The stubborn fact is that increases in maximum loan amounts and extensions of the maximum term permitted materially influence the risk, even if you assume the same property and hold all other elements of risk identical.

Of course, if the insurance fund were now excessive, this would not be a critical problem. Unfortunately this is not the situation. The FHA prepared a study of the adequacy of its insurance reserves for the President's Advisory Committee on Housing. The study showed the reserves on the basis of the existing portfolio, as of June 30, 1953, to be short between $70 million and $100 million. Nor was the Committee satisfied with this study, excellent beginning though it was. In fact, one of the recommendations of the President's Committee was that an independent, objective, long-range study for prospective foreclosures and losses should be made. It is this evidence of a deficit in the existing reserves against the existing portfolio that makes the possible cost involed in the new proposals assume even greater significance.

How much, then, do the proposed changes in individual loan plans for the FHA title I, section 203 loan increase the possible risk? A simple illustration using the same assumptions as were employed by the FHA in its study will mathematically indicate the possible extent of the increased risk. Assume a singlefamily residence valued by the FHA at $12,000. Assume a loss of the magnitude used by the FHA in its study at the end of the first 3 years of the loan. At a 42 percent rate on a 20-year term, a loan in the amount of the maximum permitted under the now current regulations will result in a possible gross loss of approximately $379. If you retain the same interest rate and hold the maturity of the loan still to 20 years but increase the amount of the loan from the present maximum to the maximum possible under the proposed will, which is $1,000, you raise the possible loss to $560. In other words, with an increase in the loan amount of slightly over 10 percent, you increase the extent of the possible loss 47.75 percent. If simultaneously with the increase in the loan amount you extend the term for only an additional 5 years, you raise the possible gross loss to $609, an increase over the base figure of 60.69 percent.

If the objective of the FHA operation is to make it a self-supporting operation to the fullest extent possible, these costs assume critical importance. How, then, can we deal with this problem?

I would like to respectfully submit for the Congress' consideration two possible amendments which at least might help in clarifying this problem. One would be to instruct the Housing and Home Finance Agency to carry out the recommendations of the President's Advisory Committee on Housing to initiate an independent study of the possible long-term foreclosure and loan experiences and report back to the Congress. This would, of course, give a factual basis for the contingent liability of the Government on the existing portfolio. The other possible change would be to instruct the Commissioner of the FHA to increase the mutual mortgage insurance premium rate to compensate for the additional risk when the President authorizes the use by the Commissioner of the maximum increases permitted by this bill. There is certainly no magic in a set annual premium rate of one-half of 1 percent. If one-half of 1 percent is an adequate annual premium rate for loans now being insured under current regulations, it is obviously inadequate to insure loans that would be made under the maximums or even short of the maximums permitted by this bill. We should create in the FHA insurance plan a flexible premium rate dependent upon the estimated risk if the insurance of mortgages is to conform to the most elementary actuary principles.

This whole problem of potential liability, of course, is present in an even more acute form in the VA program. Here, however, the problem is complicated by the fact that the veterans' loan program is so intertwined with the whole issue of veterans' welfare. Here, too, however, a similar study should be made to enable the Congress to more properly affiliate its contingent liability.

The second element of cost is the effect the proposed bill would have upon the distribution of power over and responsibility for the Nation's mortgage credit structure. The great issue which is raised by the FHA insurance plan again for title I, section 203, loans is that it practically relieves the individual lender of any individual loss. As a result of the assumption of this burden by the FHA, it has been forced to take over the basic-lenders function of discriminating among potential buyers. This is proper. The FHA certainly cannot guarantee individual lenders against loans and simultaneously permit these same private lenders to perform the necessary underwriting functions of

measuring the desirability of each loan. When you couple this necessary assumption of power with a financing plan considerably more liberal than that permitted private lenders in the conventional-loan field, you have put together the necessary ingredients for a highly centralized politically directed mortgage system. The private lender no longer performs the classic function of a lender. One of the greatest virtues of a privately operated mortgage credit system is the real and legitimate differences that exist between private lenders as to what constitutes a good borrower, what is good security, and what is a proper loan. When you centralize all decisions on these and similar matters into a single political agency, you are creating a highly brittle and rigid mortgage structure. You are, in effect, depriving that structure of the freedom and mobility it needs and requires. Many of the complaints both Government and private lenders have received about the low FHA valuations, unrealistic minimum property requirements, and cumbersome processing procedures are an inevitable consequence of that centralized mortgage structure. Are we confronted with an either/or kind of choice? Must we have a highly centralized mortgage structure in order to have Government insurance? Personally, I do not believe we are. I think our major problem has been that we have just gone along the path of the original FHA of 1934 when in the face of the general economic climate 100-percent insurance was necessary. I do not believe that, in the run-of-mill single-family residence loan, you need 100-percent insurance. For other types of loans, for certain of the types considered in this bill, on which I am making no comment, you undoubtedly do need 100 percent. If we can eliminate 100-percent insurance, if we can more properly distribute risk of loss between the insurance fund and the private lender, we can have an insurance plan and at the same time a flexible mortgage structure. However, this is a difficult task. One possibility that certainly, to my knowledge, has never been adequately explored, is the use of the insurance plan employed by the FHA in its property-improvement loan where a portion of the portfolio is insured in each individual loan. The result is that the FHA title I propertyimprovement operations are not involved in determining the desirability of each individual loan. Certainly you could not convert the present FHA setup overnight from 100-percent insurance to partial insurance of the total portfolio. This bill, however, seeks to extend to existing housing the benefits previously restricted to new construction. Here is a fruitful area in which the concept of pooled insurance might properly be used.

The second major facet of the proposed housing program is contained in title III of the bill relating to the reorganization of the Federal National Mortgage Association.

Actually, there is no essential economic difference between the current FNMA operations on military, defense, and disaster housing mortgages and the proposed special assistance functions assigned the reorganized facility by section 305. With respect to the other function of management and liquidation of the existing FNMA portfolio, the only essential change the bill makes in section 306 is the provision of subsection B for private financing as a substitute for Treasury borrowings.

I have two objections to this proposed change. First, replacement of Government debt by long-term private debt at this time would have a depressing effect on mortgage credit. The bill itself expresses official concern about an overhasty liquidation of the existing mortgage portfolio. Pushing on to the long-term capital market obligations to finance the holding of mortgage loans already made would, in my opinion, have practically the same economic effects as would the direct sale of the mortgages themselves if the conversion of the obligations into private hands was successful to any extent.

Second, to give the reorganized corporation's capital the double duty of supporting not only private financing of its secondary operation but also of what the President himself has termed "frozen investments" is risking inadequate performance by the corporation of both functions.

With the single exception of conversion of public debt into private obligations, there is certainly no reason for such a reorganization of the existing FNMA as is proposed by the bill to deal with these two major functions. Indeed, if it were not for section 304, there would certainly be no necessity to convert the present stock of the FNMA, its paid-in surplus, surplus reserves and undistributed earnings into a new stock issue to be delivered to the Secretary of the Treasury, since at present all these amounts already belong to the Government. In fact, there might well be a disadvantage in such a conversion for by converting reserves

and undistributed earnings into capital stock the Government is depriving itself of a possible cushion for the absorption of future losses which might arise in the further liquidation of the Government's existing mortgage holdings. The basic changes proposed in title III, therefore, stand or fall on the practicality of the proposed secondary market operations outlined in section 304 of the bill.

Section 304 is an attempt by the administration to create a so-called secondary market, about which there has been so much discussion in the past few years. Most of the basic difficulties created by the proposals in this section arise because there is little incentive for membership in the facility and because the facility is being called upon to purchase outright assets offered it instead of lending funds on the security of those assets. These two essential changes separate the proposals from the experience of all secondary market operations previously developed to deal with real-estate mortgages. Because of the lack of incentive for membership in the facility, it is necessary to have Government ownership of the stock at the outset, and to resort to an elaborate device to convert eventually Government ownership into private ownership.

This procedure as outlined in the bill requires the purchase of certificates by users of the facility equal to 3 percent of the dollar amount of mortgages sold the corporation. At this rate of certificate purchases, it will take the sale of $2,333,000,000 in mortgages to retire the Government stock. Pending this retirement of Government ownership, the certificates will not receive any earnings. To attach a price tag of 3 percent to the users of the facility and, simultaneously, to require, as does section A, that the price to be paid by the association for any mortgage purchased by it under this section should be established at or below the market price for that particular class of mortgage involved is, in effect, to discount all loans purchased at the very best 3 percent below the current market price for such mortgages.

To make this discount more palatable, the 3 percent requirement could be reduced. A drop to 2 percent, however, raises the dollar amount of mortgages to be purchased by $770 million. If the certificate-purchase requirement were set at only 1 percent of the loans sold to the facility, the corporation would have to buy $7 billion worth of mortgages at a price still 1 percent below, at best, the current market price. To expect investors to utilize a device of this kind under these circumstances is a forlorn hope. While the investor ultimately would recover his investment in that the certificate would be convertible into stock ownership when, if, and as the Government investment is retired, this will certainly take a considerable length of time.

I personally would not place a value on the present worth of the future benefits which might accrue to an investor out of the conversion of these certificates to stock ownership.

One of the real but almost totally ignored problems of creating a genuine secondary market facility is the constant insistence that such a ficility purchase outright mortgages offered to it by its member institutions. A facility which operates on the purchase principle must confine its borrowings from the money markets entirely to the long-term market. It must depend solely on its own credit position. It must assume fully the risk of return of principal, of the servicing costs of the mortgages it buys and of the abrupt changes that can and have occurred in the long-term interest rate structure. In short, once you impose upon such a facility the obligation that it operate on a purchase basis, while at the same time it must obtain funds by borrowing in the capital markets, you create the credit dilemma which has haunted all of the debates on a secondary market facility.

The dilemma can be briefly stated. Seemingly everyone wants some kind of a central mortgage facility. On the other hand, everyone is equally insistent that the proposed facility not operate as a primary lender. Unfortunately, no concise definition of the term "primary lender" is offered. However, what is obviously intended is avoidance of construction of a one-way street whereby mortgages enter the portfolio of the facility and never again emerge into the normal, everyday secondary market except as costly losses to the facility. How can this be prevented? The answer is simple but harsh.

There is only one way such a ficility can operate without being a primary lender. It must operate in response to the same pecuniary incentives that motivate the investment program of any other large lender. In essence, it must operate like a private bank, a private insurance company, or a private savings and loan association. It must mintain a margin between the cost of its money

and the yields it obtains on the mortgages it purchases wide enough for it to be a genuine private institution. This fundamental operating principle does not mean that the facility must invest its funds in exactly the same way as private lenders. It does mean it must maintain a proper margin between cost and income.

For a secondary facility not to operate as a primary lender, it must consummate its purchases on the same terms and conditions as those imposed by the vast network of banks and insurance companies which today constitute the private secondary market for mortgages. The postwar FNMA purchased loans of a type and at a price higher than the private market offered; hence, it became loaded with what the President termed "frozen investments." The reason it could purchase at terms and conditions other than those offered by the private market was because its operations were animated by motives other than the pecuniary incentives which dominated the investment decisions of the private secondary market. It raised its investable funds under the protection of the Treasury at rates below the rates private investors paid for their funds. A successful Government-sponsored mortgage facility must therefore be reasonably held to the same profit and loss bookkeeping as is a large insurance company or any other purchaser of mortgages. This means there can be no subsidy, direct or indirect, by the Government to such a corporation. Such a subsidy can only operate to weaken, if not to completely destroy, the pecuniary motives which must dominate the facility's operations and, consequently, compel the facility to operate as a primary lender.

Even with such private incentive, the facility could still get locked in through abrupt changes in the yield structure.

The critical point, however, is that secondary operations of the facility must be conducted without subsidy. Whether the proposed facility will in fact be subsidized or not depends entirely on the meaning of the separate accountability requirements of section 307 of the bill. Does the separate accountability apply to income and expenses as well as to assets and liability? What happens to any net income that may accrue to the corporation from performance of its special assistance functions and the management and liquidation of the existing FNMA portfolio? How are the expenses of operation to be allocated?

These are critical questions, for if the existing earning assets of FNMA are turned over to the new corporation and if the new corporation is charged that rate of interest the present FNMA is currently paying the Treasury, a substantial net income will be made available to the corporation. Even if this net income is not allocable to reserve and surplus accounts, it could easily be dissipated by absorbing expenses which would otherwise not be possible.

The same can be said of net income arising from performance of the special assistance function. Here again a subsidized money cost is available through the Treasury. If this is directly or indirectly made available to the corporation in its secondary operations, we would again have an indirect but nonetheless substantial subsidy to the secondary operations. As a matter of fact, in view of the bookkeeping and accounting difficulties inherent in the consolidation into one corporation of three diverse functions, two of which are subsidized, a serious question can be raised as to the desirability of a single corporation.

It would be a much cleaner, clearer operation if the secondary market operations were the sole concern of one corporation, and the subsidized operations of the proposed facility were handled by a second corporation.

One further special comment needs to be made about the operation of the special assistance function where the facility will operate in effect as a primary lender to encourage the use of certain types of mortgage plans.

Some type of incentive should be placed on lenders using the facility to absorb mortgages of the types designated so that the facility does not become a dumping ground. Subsidized insurance schemes assure builders of a ready and pressing market for houses financed under these specialized mortgage devices. Every effort should be made to require builders and lenders to assume a proportion of the risk involved in the long-term performance of such loans. It might be that a 3 percent or 4 percent deposit should be required to accompany the submission of the loans to the facility. Such a deposit could be returnable after the expiration of a period of time depending on the performance of the particular mortgage submitted. This requirement would make the original lender a partial guarantor of the soundness of its own underwriting processes and hence would partially inhibit improper use of the facility.