Ron J. Feldman - Financial Specialist
Published September 1, 1996 | September 1996 issue
"... Freddie Mac benefits home buyers and renters at no cost to the government."
Chairman and CEO
Federal Home Loan Mortgage Corp. (Freddie Mac)
Congressional testimony, March 1996
"Fannie Mae ... is efficiently and effectively delivering billions of dollars each year in benefits to the low- and moderate-income home buyers Congress intends to help, at absolutely no cost to the taxpayer."
Chairman and CEO
Federal National Mortgage Association (Fannie Mae)
Congressional testimony, April 1996
"Using [government-sponsored enterprise] status to enhance the credit quality of the enterprises provides Fannie Mae and Freddie Mac with savings in funding costs worth billions of dollars each year ... economic benefits are being transferred that are equivalent to those provided by writing Treasury checks."
Congressional Budget Office Study May 1996
How can one viewpoint hold that a federal program is costing the taxpayers nothing, and another insist that the cost reaches in the billions of dollars per year? Obviously, each party is interpreting the same set of facts in different ways. In this case, Fannie Mae and Freddie Mactwo agencies created by the federal government to make funds available for home loanshave an incentive to convince Congress that their efforts have no cost to the taxpayer. However, the agencies are getting a valuable taxpayer resource for free, namely: the federal government's implicit credit standing, something for which competing firms would gladly pay.
The agenciesone with roots in the 1930shave evolved into stockholder-owned and privately managed companies. But Fannie and Freddie, which purchase mortgages from lenders, are not completely privatized: Taxpayers, not investors, would likely bear the burden of Fannie and Freddie's debts if the two firms could not repay them, even though taxpayers are not legally liable. Together, the potential financial responsibility of taxpayers is substantial. Fannie and Freddie financed about 36 percent of all one- to four-family housing debt outstanding as of March 1996more than the entire commercial banking system.
There are questions about whether the current arrangement with Fannie Mae and Freddie Mac is the most effective method to aid potential homeowners. But these questions do not have obvious answers. Unfortunately, there is much uncertainty surrounding the performance of the two agencies. Analysis is difficult because the picture is often not clear.
This uncertainty is not uncommon with federal programs. In every program there is at least a minimum of uncertainty about its efficacy it is an unavoidable result of any federal interventionbut such small levels may simply be the cost of doing business. However, uncertainty that pervades a program should bother the public and policymakers; after all, federal intervention almost always involves the use of taxpayers' resources. The need for review is not always apparent, though, because a program's design may highlight its virtues while downplaying its true performance.
Fannie and Freddie provide a good case study to explore uncertainty in federal intervention. The agencies' circuitous subsidy transmission is outside the control of the federal government; as a result, policymakers have a hard time discerning if Fannie and Freddie pass almost all of the subsidies to borrowers or if the agencies retain a significant portion. Also, the indirect delivery of the federal subsidies hinders analysts' ability to determine if Fannie and Freddie add value by offering new products and services.
This uncertainty would not raise red flags if the enterprises were sure to pass on their federal benefits to borrowers. In reality, Fannie and Freddie's orientation toward profit-making creates a strong incentive to retain some of the subsidies, although the ability to act on this incentive is tempered by Fannie and Freddie's housing responsibilities and limitations on their activities.
For the public and legislators, then, the questions of uncertainty surrounding Fannie and Freddie regard the efficiency of using the agencies to transfer a subsidy that is intended to benefit home buyers: Can Congress or the public really feel comfortable with Fannie and Freddie's use of their resources? Does the apparent success of Fannie and Freddie mask deeper uncertainties about the agencies? Is there a better delivery system for the subsidies?
Fannie and Freddie's claims of costless benefits are suspect. They make this claim because the benefits do not involve explicit transfers of cash. However, taxpayers also incur costs whenever valuable public goods and services are provided by the government to beneficiaries for free or at a discounted price. The following hypothetical example illustrates this point: Mr. and Ms. Olson go to a local bank to apply for a loan to renovate their home. After reviewing their application, the loan officer informs the Olsons that she will only make the loan if they offer their car as collateral and pay a fairly high interest rate. The bank is concerned, she explains, that the Olsons will not have the resources or the desire to repay the loan.
As they trudge away from the loan officer's desk, the Olsons pass a brochure touting the federal government's brand new "Cosigning for America" program. Under this program, the couple reads, the federal government will cosign any home renovation loan without any charge to the borrower. That is, the federal government will repay the loan to the bank if the borrower(s) cannot. With completed brochure in hand, the Olsons return to the loan officer who now agrees to approve the loan at a very low interest rate and without requiring collateral. The bank knows that the ability of the federal government to repay the loan is unquestioned.
Has the federal government given the Olsons anything of value? Yes. Without the federal government's cosign, the Olsons would have had to pay a higher interest rate and would have lost their car if they defaulted on the loan. In fact, the Olsons would have paid the government to cosign the loan.
Is "Cosigning for America" costly for taxpayers? Yes. As noted, the federal government could have sold its cosigning authority to the Olsons. The money raised by the sale of the cosigning could have been used to reduce taxes. The giveaway of the cosigning has a real cost to taxpayers even if the federal government never writes a check under the program. This last point is worth repeating as it often causes confusion. A program of free government cosigning is costly to taxpayers even if all the borrowers with whom the government cosigns repay their loans without government help. The cost arises because the government is giving away a valuable service for which borrowers would pay. The borrower is willing to pay for the cosigning even before they start repaying the loan, as the benefits they receive from the cosigningparticularly the more favorable interest ratesare built into the terms of the loan.
A variety of firms sell the equivalent of "cosigning" services. For example, many home buyers purchase mortgage guarantee insurance. For a fee, these insurers will make up some of the shortfall on the mortgage that the lender suffers if the home buyer defaults and the sale of the property does not cover the loan.
Like the Olsons, Fannie and Freddie receive, and still benefit from, valuable services and rights from the federal government. The federal government exempts Fannie and Freddie from both state and local income taxes on their earnings, and the costly registration process of the Securities and Exchange Commission. But the biggest subsidy Fannie and Freddie receive involves an implicit federal guarantee. Investors believe that the special status and powers granted to Fannie and Freddie by the federal government create an implicit, but not legal, guarantee on Fannie and Freddie's financial obligations. Investors have concluded that the federal government would almost certainly pay them the interest and principal on Fannie and Freddie's obligations if Fannie and Freddie could not. This perception is not unintended. The obligations of Fannie and Freddie share many characteristics with U.S. government securities. Moreover, the federal government did not repudiate a similar implied guarantee of the Farm Credit System when it had the chance in the late 1980s.
Fannie and Freddie benefit from the implied guarantee when they borrow money. In some cases, they issue debt on the national capital markets, use the proceeds to purchase mortgages from lenders and hold the mortgages in their respective asset portfolios. This activity is called portfolio lending. Both firms also fund mortgage lending by securitizing mortgages (for example, acquiring mortgages, grouping the loans and issuing a specialized type of debt backed by this loan pool). The payments that homeowners make on the underlying mortgages are used to pay the interest and principal owed to the investors who bought the so-called mortgage-backed securities. Fannie and Freddie guarantee the payment of principal and interest on the mortgages underlying the mortgage-backed securities.
The implied guarantee is valuable because it substitutes the federal government's ability to repay debt in place of Fannie and Freddie's. This substitution, as was the case for the Olsons, reduces Fannie and Freddie's cost of raising funds because lenders will accept a lower interest rate when they are virtually certain to get their money repaid. Many private firms make money by selling their guarantee of repayment to other firms that are not as creditworthy. In contrast, the federal government gives away its implicit credit guarantee to Fannie and Freddie. The amount of this implicit taxpayer coverage, as of March 1996, was about $360 billion of one- to four-family mortgages in Fannie and Freddie's portfolio and about $1 trillion of outstanding guarantees on mortgage pools. By way of comparison, these financial obligations were about equal to all the debt outstanding of state and local governments and were equal to about 35 percent of all outstanding Treasury securities.
Confusion about the cost of a subsidy is not limited to the case of Fannie and Freddie. For years, policymakers claimed that the provision of deposit insurance and other financial guarantees by the federal government had no cost to taxpayers. These programs lend themselves to uncertainty because nothing tangible is exchanged between the government and the recipient and the ultimate cost of the program may not become clear until far into the future, long past the initial provision of the guarantee. However, even on a superficial level, the subsidy of Fannie and Freddie is harder to grasp than that of many guarantee programs because of its implicit nature and the public ownership of Fannie and Freddie.
The initial murkiness engulfing the cost of Fannie and Freddie's subsidy is exacerbated because Fannie and Freddie currently enjoy more information on, and day-to-day power over, the federal subsidy than Congress or the president. As the budget office of Congress recently reported, policymakers cannot look to the federal budget for answers on how much of a subsidy will be paid to Fannie and Freddie; neither does any appropriations bill nor any other accounting record measure the size of the federal subsidy.
In fact, the federal government does not set the size of the subsidy it provides Fannie and Freddie each year; the magnitude of the subsidy from the implied guarantee, for example, depends on the difference between how much Fannie and Freddie would pay to raise money if they did not have special status with investors (that is, their true borrowing costs) and how much they actually pay with the perceived credit support of the federal government. The size of the subsidy also depends on Fannie and Freddie's outstanding obligations. The weaker the financial condition of Fannie and Freddie, the higher their true borrowing costs and the greater the subsidy provided by the implicit guarantee. The more obligations the federal government guarantees, the greater the subsidy.
The ability of Fannie and Freddie, rather than Congress, to manage the size of the subsidy can manifest itself in subtle ways. Fannie and Freddie can take on new risks, for example, by choosing to fund mortgages through portfolio lending instead of mortgage-backed securities. By issuing debt and holding mortgages, Fannie and Freddie may bear the risk that interest rates will fall, thereby encouraging home buyers to prepay their mortgages, possibly leaving Fannie and Freddie with less income to repay the debt they have issued. The implied guarantee practically, but not legally, shifts this risk to taxpayers. In contrast, these risks are borne by investors when Fannie and Freddie issue mortgage-backed securities. The greater the risks taxpayers bear, the greater the subsidy they provide to Fannie and Freddie. Thus, Fannie and Freddie can increase their subsidy through the portfolio lending option.
It is much more difficult to quantify the subsidy than to describe the activities that alter it. The quantification could require, for example, an estimate of how much it would cost Fannie and Freddie to raise money if they did not have the implied guarantee. One way to try to get at the cost advantage of federal credit support is to compare mortgage-backed securities or debt issued by Fannie and Freddie with similar obligations issued by firms in the same general lines of business as Fannie and Freddie but without their favored status. But finding comparable companies and obligations and trying to control statistically for the differences between Fannie and Freddie and their comparable entities is, by definition, very formidable, because Congress designed Fannie and Freddie and their obligations to be unique. In addition, all estimates are highly dependent on the time period under examination. The same method that found a large subsidy in one period will find close to no subsidy in another.
These difficulties do not mean that estimates are a waste of time. The Congressional Budget Office (CBO) provided useful information when estimating that the value of the subsidy for both enterprises was about $6.5 billion in 1995. But these are far from exact calculations. They are better understood as educated estimates on the orders of magnitude of the subsidy at a given time, that are subject to challenge and reflect inherent measurement obstacles. Thus, while it may be tempting from the onset to focus on estimates, policymakers may be better served by considering the costs, and possible cures, of their disadvantages in gathering information on and exerting control over Fannie and Freddie's subsidy.
Given that they do not set the size of the subsidy, it is not entirely surprising that policymakers only vaguely know who benefits from the subsidy. Certainly, Fannie and Freddie are the initial beneficiaries of federal subsidies, and they claim to effectively pass the subsidy on to home buyers. If Fannie and Freddie faced vigorous competition, they would have no choice but to pass the subsidy to the firms from whom they purchase loans. Because a very high level of competition exists among mortgage originators, these firms would then pass the subsidy to mortgage borrowers. If one originator tried to hold on to the subsidy, another firm could gain an advantage by lowering its prices by the full amount of the subsidy. Other originators would follow suit to remain competitive.
However, there are some indications that Fannie and Freddie are not subject to the type of competition that would force them to pass the full subsidy on to originators. In their securitization market, for example, Fannie and Freddie, in effect, vie only with each other for business. Fannie and Freddie also report financial performance, such as very high returns to shareholders, and market valuations not entirely consistent with full competition. While this evidence appears persuasive, the CBO reported in 1993 that Fannie and Freddie's markets, even with only two competitors, could theoretically generate competitive pricing and production.
It is difficult to determine how much of the subsidy goes to aid the homeowner, even if one knew that Fannie and Freddie were not subject to full competition. Some analysts have calculated the difference in mortgage rates between those mortgages that fall below the conforming limit and those that rise above in order to determine how much Fannie and Freddie lower interest rates. Fannie and Freddie cannot finance mortgages whose original principal amount exceeds the conforming limit, $207,000 in 1996. These calculations find that Fannie and Freddie lower interest rates on conforming loans by about 0.3 percentage points (from 6.3 percent to 6 percent, for example). But, the complexity in trying to separate out how the conforming status affects the cost of a mortgage, all other variables being equal, makes the estimate of reductions in mortgage costs uncertain.
Nonetheless, analysts try to figure out how much of the subsidy actually goes to borrowers by combining their estimate of the size of the subsidy with estimates of how Fannie and Freddie's activities affect interest rates on home loans. Through such a calculation the U.S. Treasury found that about two-thirds of the total subsidy went to borrowers, while Fannie and Freddie retained about $1.9 billion in 1995. The Treasury notes, however, that "Although estimates ... give a general sense of the magnitude of the subsidies involved, no single point estimate should be viewed as a firm indicator of the benefits [Fannie Mae and Freddie Mac] receive or pass through. The calculations ... omit important elements of both benefits received and benefits passed through."
One should not equate the U.S. Treasury's imprecision with the notion that it overestimated the retained subsidy; rather, the imprecision itself is the point. The uncertainty of the calculations means that Congress is always going to face challenges in distinguishing between years when Fannie and Freddie pass on almost all of the subsidy from the years when they retain much of the subsidy. As the CBO acknowledged even while presenting its own estimates, "... the complexity of operations conducted by [Fannie and Freddie] and the proprietary nature of information about their business combine to make it almost impossible for the Congress to be sufficiently well informed about the conduct of Fannie Mae and Freddie Mac."
Again, Fannie and Freddie are not entirely unique. There are a number of cases in which it is difficult to determine who ultimately benefits from a federal subsidy. Firms and households that receive subsidies, and the firms and households they interact with, will change their behavior in ways that make it difficult to determine who exactly receives the economic benefits of the subsidy. The difficulty in analyzing the distribution of benefits, however, is not equal among all programs. The more hands through which the subsidy has to pass to get to the intended borrower, for example, the more difficult it will likely be to determine the distribution of benefits. A subsidy that travels from the federal government to the staff and owners of Fannie and Freddie to mortgage bankers to home buyers via the capital markets is passed quite often.
When evaluating the performance of Fannie and Freddie in meeting their policy objectives, policymakers will need to consider Fannie and Freddie's argument that they do not just deliver a subsidy. Instead, they claim to create value for society through activities such as developing innovative mortgages and investing in new technologies that will reduce the cost of home purchases. Normally, the success or failure of a product or service in the market signals if it is creating value. The long-term willingness of consumers to buy, and producers to provide, a product means it passes the market hurdle and adds value.
The introduction of the federal subsidy fundamentally skews this market test. Once any product is subsidized, it becomes difficult to determine if its success reflects the inherent benefits consumers derive from it instead of its below-market prices. Likewise, the success of Fannie and Freddie's products and services may simply reflect their subsidization by federal taxpayers rather than any unique value the products or services provide.
There is one area in which Fannie and Freddie have added value: Fannie and Freddie integrated national capital markets with local mortgage markets, and thereby helped eliminate funds shortages at a time when local and regional lenders could not raise enough money to make all the home loans demanded by borrowers. But, even these benefits created by Fannie and Freddie may not justify past and current subsidies. Congress may have been able to reduce regional funds shortages, in part, simply by revoking policies that limited how much interest banks could pay to depositors and where banks could operate. In addition, unsubsidized private firms could now probably integrate local mortgage markets with national capital markets.
Policymakers may view the recent debate on estimates of how much subsidy Fannie and Freddie retain as just one of many disagreements among policy wonks. Reaching that conclusion, however, would relegate legislators to a self-imposed and perpetual uncertainty about the performance of Fannie and Freddie in meeting policy goals.
Policymakers may find the charters of Fannie and Freddie reassuring in the face of this ongoing uncertainty. They may assume that, by detailing the responsibilities of Fannie and Freddie to assist borrowers, the charters will act as a beacon and guide Fannie and Freddie toward effective subsidy provision. However, the financial incentives facing Fannie and Freddienot just a vague charteralso guide their behavior. A legislator may be willing to accept the uncertainty concerning one form of federal subsidy transmission if the program's incentives ensure that the intended party receives the subsidy. However, in this case, Fannie and Freddie have a financial incentive to enlarge the amount of the subsidy they receive and retain.
Like other for-profit firms, Fannie and Freddie have one overriding incentive: to maximize profits. These incentives result directly from the organizational form and legal status of Fannie and Freddie.
According to Fannie Mae's 1995 annual report, its "... mission has been to provide financial products and services that increase the availability and affordability of housing for low-, moderate-, and middle-income Americans." Yet, management of Fannie Mae or Freddie Mac cannot, over the long-term, simply allocate resources to the worthwhile cause of its choice. After all, these firms are owned by shareholders. What drives these owners of Fannie and Freddie? While Fannie and Freddie's shareholders do not have an official mission statement, their unofficial motto might read something like, "We demand the highest return available on our investment for our level of risk tolerance." The owners' creed does not make shareholders an avaricious lot, unsympathetic to the cause of housing availability and affordability. Rather, enterprise owners, such as several Fidelity mutual funds and Warren Buffet's Berkshire Hathaway, have a duty to their own shareholders to maximize value.
When the activities of Fannie and Freddie conflict with shareholders' goals, the owners would try to influence managementeven install new managementto bring corporate behavior in line with shareholder objectives. For example, shareholders of the Student Loan Marketing Association (Sallie Mae)a firm organized like Fannie and Freddie to increase funds for student loanssuccessfully installed eight dissident members to its board of directors when management sought to cut ties with the federal government. This step would have reduced the ability of shareholders to benefit from the federal subsidies Sallie Mae receives.
If Fannie and Freddie continually engaged in activities that did not maximize profits, their shareholders could sell their stock and drive down the value of Fannie and Freddie's stock. A reduction in the value of enterprise stock would directly affect senior management, who receive some of their compensation in the form of stock. The potential for unpleasant shareholder actions gives Fannie and Freddie a long-run incentive that should differ from their mission statement in one crucial aspect: Fannie and Freddie will increase the availability and affordability of housing but only in a manner that maximizes shareholder returns. Alternative scenarios in which these investors willingly give up returns on their investments, or are systematically duped by Fannie and Freddie for long periods of time, seem less plausible.
Shareholders' objectives seem to give Fannie and Freddie a significant business incentive to treat their federal subsidy as another tool to increase profits. We should thus expect Fannie and Freddie to retain portions of the subsidy. We should also expect Fannie and Freddie to try to increase profits by expanding the activities in which they deploy the subsidies and gain market dominance. Analysts have reported behavior consistent with the expanded deployment of the subsidy.
Freddie Mac's use of debt to fund mortgages offers an interesting example of how shareholder ownership could influence enterprise behavior. As previously noted, Fannie and Freddie can increase their subsidization, and thus increase returns to shareholders, through portfolio lending. During the 19 years prior to its change in ownership status, Freddie Mac did not engage in much portfolio lending. Its mortgages in portfolio increased from zero to about $21 billion in that period. In the six complete years from 1990 to 1995, following the conversion to public shareholder ownership, mortgages in portfolio increased from about $21 billion to about $108 billion.
This incentive does not mean that Fannie and Freddie's charters are blissfully ignored, though; shareholders do not expect Fannie and Freddie to exploit their subsidy in an unconstrained fashion because they know that Fannie and Freddie must meet very specific targets to assist a variety of home buyers, including those with lower incomes. Fannie and Freddie face legal restrictions on the types of activities in which they can engage. However, these restrictions are not so severe relative to the benefits of their favored status, so as to compel Fannie and Freddie to give up their federal charters.
Owners also realize that Congress may decide to eliminate the subsidies to Fannie and Freddie if policymakers perceive that shareholders capture too much of the subsidy. Shareholders want to avoid this threat because a reduction of the subsidy means a reduction in shareholder wealth. For example, Fannie and Freddie's stock price fell on Friday, Oct. 20, 1995, after a senior congressman proposed levying a user fee on Fannie and Freddie that would effectively reduce their subsidy. However, by the following Monday the stock price was back up. According to Bloomberg Business News, "... analysts all over Wall Street said that the user fee idea would blow over and that investors should take the opportunity to buy ... " Shareholders should be willing to give up some subsidy today to ensure that they reap the benefits of subsidies into the future.
Policymakers have many methods for providing housing subsidies. A direct subsidy transmission from the government to targeted home buyers would not suffer the same level of uncertainty or the same incentives. The direct subsidy could take any number of forms: One method would be to directly provide checksthat could be restricted to housingto all qualified home buyers (those in a certain income class, for example). Congress would annually appropriate the funds for this program and enact rules on eligibility and the percent of program costs that could go to administrative expenses. If it so desired, Congress could require that private firmswhose contracts reward them for cost-effectivenessdesign and operate the program. Whatever its form, the cash flows of the direct subsidy would be recorded in the federal budget.
Some analysts oppose the replacement of Fannie and Freddie's subsidies with an appropriated subsidy, believing the switch will lead to a reduction in housing subsidies. A restrictive budgetary climate, they argue, would prevent Congress from approving the same level of subsidies for a direct subsidy program that Fannie and Freddie are believed to receive. But a housing subsidy whose survival depends on being hidden from view, sheltered from budgetary constraints, and shielded from comparisons with other uses of society's resources should be re-examined.
A direct subsidy program would not be immune from uncertainty. It may be difficult, for example, to determine how much of the subsidy benefits landlords who raise prices in response to the new subsidy. However, the direct approach provides a clearer and much more public framework with which to analyze the variety of issues that all subsidy programs inevitably raise.
It is not always obvious if a federally subsidized intervention achieves its goals. Of course, a little doubt about the effectiveness of a government program is practically unavoidable. However, the uncertainty concerning the effectiveness of a federal program can be pervasive and inherent in the subsidy transmission vehicle policymakers have chosen. Those cases deserve policymakers' and the public's special attention, and Fannie Mae and Freddie Mac are examples of such vehicles.
Upon closer inspection, there are at least two major implications of using organizations like Fannie and Freddie to provide subsidies.
The combination of an environment of uncertainty and this financial incentive is a virtual formula for an inefficient use of taxpayer resources. If so inclined, Congress could provide a direct subsidy to homeowners and reduce the uncertainty associated with Fannie and Freddie. At a minimum, these implications suggest that Congress consider their ability to evaluate the effectiveness of a subsidized program, and the incentives provided by the program, when determining if it should be implemented.
Editor's note: One of Fannie and Freddie's most significant new initiatives involves automating mortgage underwriting. A follow-up article in the next issue of The Region will discuss the uncertainty surrounding the potential benefits and costs of these automated systems.