Fannie Mae and Freddie Mac
Fannie Mae (the Federal National Mortgage Association) and Freddie Mac (the Federal Home Loan Mortgage Corporation) are two government-chartered corporations initially created for the purpose of increasing the liquidity of the residential mortgage market. They perform this function by purchasing residential mortgages from banks for their portfolios or by guaranteeing securities based on pools of mortgages assembled by lenders or other mortgage originators. Through different routes, both companies were partially privatized—Fannie in 1968 and Freddie in the mid-1980s—and both are now owned entirely by private investors and listed on the New York Stock Exchange.
Nevertheless, the terms of their partial privatization left both companies with special privileges and links to the government, so that despite their private ownership they are known as Government Sponsored Enterprises, or GSEs. To distinguish them from other corporations with government support, they are sometimes called the Housing GSEs, but in this paper, for ease of reference, I will refer to them simply as Fannie and Freddie or, together, as the GSEs.
The special links to the federal government are numerous and very important. For example:
1. the President appoints five members (a minority) of their boards of directors;
2. they are exempt from state and local taxes;
3. their securities are exempt from registration with the Securities and Exchange Commission;
4. they each have a so-called line of credit at the Treasury under which the Secretary of the Treasury is authorized to invest up to $2.25 billion in their securities; and
5. their securities are eligible for unlimited investment by national banks, and as collateral for Treasury tax accounts deposited with banks.
These ties to the federal government, among several others—and the fact that they are performing a government mission—have apparently convinced the financial markets that the government will not allow the GSEs to fail. This implied government backing, in turn, enables them to borrow money at interest rates that are significantly lower than any private sector AAA credit, and only slightly more than the Treasury itself is required to pay. With this superior financing ability, the GSEs have been able to dominate completely the secondary mortgage market for conventional/conforming residential mortgages (mortgages of less than $275,000), which constitute roughly 70 percent of all residential mortgages in the United States. They now hold in portfolio or have guaranteed securities backed by about 75 percent of all conventional/conforming mortgages, and Bert Ely and I have estimated—in a monograph published last year by AEI—that these two companies will hold or have guaranteed securities representing almost 50 percent all mortgages in the United States by the end of 2003.
The implied government backing for the GSEs has a tangible value, recently estimated by the Congressional Budget Office at $10.7 billion during 2000. According to CBO, about 37 percent of this subsidy is retained by the GSEs, enhancing the value of their shares and increasing management compensation; the balance is passed along to the mortgage market in the form of somewhat lower interest rates. Economic studies have shown that the interest rates on mortgages purchased or guaranteed by the GSEs are about 30 basis points lower than the rates on mortgages above the conventional conforming loan limit of $275,000. Accordingly, in the competitive residential mortgage market, it is essential that a lender be able to resell a mortgage to one of the GSEs, and thus virtually all conventional/conforming mortgages—which means virtually all middle class mortgages with principal amounts less than $275,000—conform to standards established by the GSEs.
Automated Underwriting Systems
Recently, advances in data processing technology have permitted the development of predictive models of creditworthiness. These models use correlations among various data elements to predict the likelihood of default by a borrower. Although the models are relatively new and have not been tested in a serious economic downturn, they have assumed great importance in the credit industry. For one thing, they significantly reduce both the cost and time associated with underwriting a credit, enabling lenders to shorten response times on loan applications and reduce interest rates. When applied to the residential mortgage market, automated underwriting, or AU, has become an essential competitive tool. For obvious reasons, a mortgage lender cannot effectively compete for residential mortgage business unless it can offer the reduced interest rates and rapid response times that have been made possible by AU.
Since the GSEs purchase mortgages from banks and other lenders they have developed their own AU systems. These systems are competitive with and can substitute for similar systems developed by mortgage lenders. The GSEs’ systems are also proprietary; mortgage lenders do not know whether a mortgage will be accepted or rejected by the GSEs’ AU systems until they have been run through.
This is not to say that the GSEs will reject all mortgages that are not approved by their systems; they will accept such mortgages, but with much more stringent representations and warranties that place greater risk on the lender. For example, under certain circumstances, a lender may have to repurchase a mortgage from the GSEs if it was not approved by the GSEs’ AU systems and subsequently defaults. In addition, the GSEs offer their AU systems bundled with other services and software that competitive AU systems cannot match, or they offer their systems at discounts that again undersell independent systems available in the market.
Finally, for certain kinds of high loan to value (LTV) mortgages, the GSEs will only accept mortgages that meet their AU standards, so that lenders for these popular mortgages are required to use the GSEs’ systems unless they want to hold the loans in their own portfolios—creating a degree of liquidity risk because the loans could not at a later time be sold to the GSEs or to other financial institutions that are assembling loans for a GSE guarantee.
But the strongest inducement to use the GSEs’ AU systems is the fact that it is the most effective way of assuring that a mortgage loan will be purchased by one of the GSEs—thus reducing the lender’s cost or risk of carrying the loan. And since the GSEs are—because of their government support—the sole economically feasible purchasers of most of the vast majority of all residential mortgage loans made in the United States, lenders that use the GSEs’ AU systems gain considerable cost advantages over lenders that do not. It is of course possible that a lender might develop its own AU system—or purchase an AU system from an independent developer—and use that system to evaluate its loans, but this would represent a higher cost initially, as well as the need to assume costly risks in selling the loan to the GSEs.
As early as February 1999, Morgan Stanley Dean Witter estimated that the GSEs’ combined market share in the use of AU by lenders was 95 percent and likely to grow. As MSDW said in a report on the GSEs’ technology developments: “[W]e believe that automated underwriting systems, of the sort developed by Fannie Mae and Freddie Mac…, constitute a kind of ‘killer app’ for the mortgage sector.”
The Microsoft Decision
In its recent Microsoft decision, the D.C. Circuit confronted Microsoft’s use of another “killer app”—the Windows operating system. The government claimed that Microsoft was monopolizing the market for personal computer operating systems, attempting to monopolize the internet browser market, and illegally tying other products to the windows operating system. The trial court found for the government on all three claims, and Microsoft appealed. (Slip Op., p.5; all subsequent citations will be to the Slip Opinion issued on June 29, 2001)
In its June 29 decision, the Circuit Court affirmed the District Court in part and reversed in part, ultimately holding that Microsoft had violated Section 2 of the Sherman Act by employing anticompetitive means to maintain its monopoly in the operating system market, reversed the District Court’s finding that Microsoft had attempted to monopolize the browser market, and remanded for further proceedings the question of whether Microsoft had violated Section 1 of the Sherman Act by tying its Internet browser to the Windows operating system. The following is a summary of the Circuit Court’s analysis in each major category. (p.6)
The Court noted that monopolization under the Sherman Act has two components: “(i) possession of monopoly power in the relevant market, and (ii) the willful acquisition of maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.” (quoting United States v. Grinnell Corp. 384 U.S. 563, 570-71 (1966)). The District Court had found that that Microsoft possesses monopoly power in the market for operating systems, and maintained this power not through competition on the merits but through unlawful means. (p. 13)
The first question the appellate court addressed is whether Microsoft in fact had monopoly power. It noted that monopoly power exists where a firm has the ability to raise prices above market levels or to exclude competition. “A firm is a monopolist,” said the Court, “if it can profitably raise prices substantially above the competitive level.” (p. 14) Since it is difficult to find direct proof of these circumstances, courts have developed a structural test. “Monopoly power, said the Court, “may be inferred from a firm’s possession of a dominant share of a relevant market that is protected by entry barriers.” ( p.14) The District Court had found that Microsoft has a greater than 95 percent share of the operating system market, and that its market was protected by a substantial entry barrier—the fact that software developers prefer to write for the Windows operating system because it has such market dominance and consumers prefer to use an operating system for which the most software is available. (pp.20-23)
Having concluded that Microsoft in fact had monopoly power, the Circuit Court turned to the second criterion for violation of the Sherman Act—use of anticompetitive or exclusionary conduct to maintain the monopoly position. The Court noted that the development or acquisition of a monopoly as a consequence of a superior product, business acumen or historic accident is not sufficient to constitute monopolization under the Sherman Act; there must be the use of anticompetitive means. (p. 25)
In this connection, the District Court found that Microsoft had engaged in a number of exclusionary acts to maintain its monopoly, by preventing the effective distribution and use of products that might threaten its dominant position in operating systems. Here the appellate court noted that it is difficult to distinguish between vigorously pursued competition and illicit exclusion. “The challenge for an antitrust court,” it said, “lies in stating a general rule for distinguishing between exclusionary acts, which reduce social welfare, and competitive acts, which increase it.” (p. 26)
In attempting to discern this line, the Court pointed out that to be illegal monopolization must have an adverse effect on the competitive process, not just on competitors. For this proposition, the Court quoted the Supreme Court in a 1993 case: “The [Sherman Act] directs itself not against conduct which is anticompetitive, even severely so, but against conduct which unfairly tends to destroy competition itself.” (Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 458 (1993)). Nor is the intent behind the act important, said the Court, “except to the extent that it helps us understand the likely effect of the monopolist’s conduct.” (p. 26)
As the Court explained, the courts have over time developed a complicated procedural path in this area. If a monopoly is found, and exclusionary or anticompetitive acts are shown to maintain it, the monopolist may—in justification—show that the alleged anticompetitive or exclusionary acts are procompetitive. And after such a showing, if any, the trial court must weigh one against the other, determining whether the restriction on trade has, overall, a net pro- or anti-competitive effect. (p. 26-27)
Using this analysis, and after considering Microsoft’s arguments that its action were procompetitive, the Circuit Court found that Microsoft had attempted to monopolize the operating system market by (i) its restrictive licenses with original equipment manufacturers (which limited the ability of manufacturers to place icons for competing browsers on their proprietary desktops when they modified the standard windows desktop) (pp. 28-35); (ii) its agreements with almost all Internet access providers, such as AOL (which provided that Microsoft would place their icons on its Windows desktop in exchange for exclusive promotion of Microsoft’s browser) (pp. 40-47); and (iii) a threat against Intel—which was cooperating with Sun Microsystems on a platform for a competitive program (JAVA)—that Microsoft would work with a competing chipmaker unless Intel abandoned the work with Sun. (pp. 52-58)
In each case, the appellate court found that Microsoft had no significant procompetitive justification for the anti-competitive or exclusionary acts alleged. In the absence of a procompetitive rationale for fundamentally anticompetitive or exclusionary acts, the Court held that Microsoft had violated Section 2 of the Sherman Act by attempting to maintain its monopoly of personal computer operating systems.
The government also charged that Microsoft, in violation of Section 2 of the Sherman Act, had attempted to monopolize or gain a monopoly in the browser market by leveraging its monopoly in operating systems. The Circuit Court defined attempted monopolization as follows (quoting Spectrum Sports, supra, 456): “a plaintiff must prove (1) that the defendant has engaged in predatory or anticompetitive conduct with (2) a specific intent to monopolize and (3) a dangerous probability of achieving monopoly power.” (p. 62)
The Court concluded that the government had not shown that there was a dangerous probability of monopolization—one of the three essential elements of the violation—and for that reason reversed the District Court’s finding of attempted monopolization. However, the Court’s view was not based on an analysis of what would be a dangerous probability of monopolization; the Court simply found that the government had failed adequately to identify the relevant market or to demonstrate that substantial barriers to entry protect that market. Without these elements, the Court pointed out, it was not possible to prove a dangerous probability of monopolization. (p. 63)
Thus, the Court’s conclusion was not that Microsoft had not attempted to monopolize the browser market, but only that the government had failed to put into evidence the essential elements of the violation.
Finally, the government argued that Microsoft—by bundling its web browser with its operating system—had illegally “tied” the browser to the operating system in which it had a monopoly. Tying arrangements are generally considered per se unlawful—i.e., wherever they are found no extended factual inquiry into purpose or intent is necessary. (p. 69) This is because the courts had found over time that tying—like price-fixing—could not be justified under any reasonable standard of conduct. The District Court had found that Microsoft had tied the browser to its operating system by integrating the two, and thus found a per se violation of Section 1 of the Sherman Act.
However, the Circuit Court held that the question of tying in this case should be reviewed under the so-called “rule of reason.” The Sherman Act prohibits any contract “in restraint of trade,” but as courts noted early on in Sherman Act litigation all contracts restrain trade to some extent, and it could not have been the intention of Congress to forbid contracts in general. Accordingly, the courts developed a “rule of reason” in analyzing Sherman Act cases, under which all the facts of the alleged violation were evaluated to determine whether the defendant had violated the standards of conduct that Congress likely had in mind. In taking this position, the Court believed that the special circumstances in the computer market and the unfamiliarity of the courts with the way the technology works could produce harm if the traditional tying analysis was inflexibly applied. Accordingly, the Court remanded the case for additional trial activity and an assessment of the tying charge under the so-called rule of reason. (p. 69; and pp. 77-86))
Nevertheless, the Court’s analysis of what constitutes tying is instructive. The Court found four elements to the tying violation: (i) the tying and tied goods must be two separate products; (ii) the defendant has market power in the tying product; (iii) the defendant affords consumers no choice but to purchase the tied product from the defendant; and (iv) the tying arrangement forecloses a substantial volume of commerce. (p. 70)
On the question of whether two goods are in fact separate products, the Court examined existing precedent, which indicated that the tying good and the tied good were separate products if there was a separate demand for each in competitive markets. (p. 72) If competitive firms—that is, firms without market power—offered the goods separately they should be considered separate products. However, the Court was reluctant to conclude that the browser was a separate product on the basis of this analysis alone. For one thing, Microsoft had argued that integrating the browser into the operating system improved the functioning of both, and the court was concerned that applying a per se rule might stunt valuable innovation. For these reasons, the Court drew no conclusion on the separate products question, but sent it back to the District Court for resolution—insofar as possible—through additional proceedings at the trial level. (p. 77)
The Court then provided guidance for considering a tying claim under a rule of reason standard. First, the Court said, the government must show that the integration of the browser and the operating system “unreasonably restrained competition”—an inquiry, said the Court, into its “actual effect.” (p. 86). In addition, the government must show that Microsoft’s conduct was, on balance, anticompetitive. As in the monopolization analysis, a defendant may demonstrate that the alleged anticompetitive conduct has a procompetitive effect, and the government has the burden of showing that the anticompetitive effects outweigh the procompetitive. (p. 87) In particular, the trial court should attempt to determine whether the operating systems are in general sold with bundled browsers by other marketers of operating systems , and—even if so—whether these companies will sell their browser separately or offer a discount if the browser is not included in the operating system. (pp. 88-90)
The GSEs Use of their Automated Underwriting Systems
Although the Circuit Court was only able to conclude that Microsoft had violated the Sherman Act in one respect—its effort to maintain its monopoly in the Windows operating system—the Court’s analysis shows that Microsoft escaped judgment on the attempted monopolization claim and the tying claim only because of special circumstances. In the case of the attempted monopolization claim, the Circuit Court concluded that the government had failed to establish two factual predicates that were essential to its argument. In the case of the tying claim, the trial had not produced sufficient evidence on certain technical matters concerning the computer software business to give the appellate court confidence that the standards for illegal tying had been met under a rule of reason analysis.
These problems are not likely to be present if the Court’s analysis, as outlined above, were applied to the conduct of Fannie Mae and Freddie Mac in their use of their respective automated underwriting systems. Despite the fact that the AU systems are the products of technological developments, and are accessible over the Internet, they do not depend for their competitive effect on their technological nature. They would have the same effect—and would be subject to the same analysis—if they were simply a set of rules applied on a case by case basis by employees in the GSEs’ offices. In other words, their effect is likely to be judged under the conventional tests applied in antitrust law—all of which were carefully outlined by the Circuit Court.
Nevertheless, as the Microsoft decision shows, antitrust law is highly fact-specific. It is difficult, without a detailed inquiry at a trial, to differentiate between aggressively competitive conduct and exclusionary or anticompetitive action. There is a great deal that is still unclear about how the GSEs use their AU systems, and for that reason it is not possible to draw firm conclusions about how a court would analyze the GSEs’ conduct in an antitrust context. However, there is enough information to make start on such an analysis from the standpoint of a court using the Circuit Court’s Microsoft decision as its analytical framework.
The first question a court would face is whether Fannie and Freddie are competitors. Fannie and Freddie, together, account for almost 100 percent of the secondary market for conventional/conforming loans, and about 70 percent of the secondary market for all residential mortgages. Of these totals, Fannie has about 60 percent and Freddie 40 percent in each market. If the two companies were truly competing, it would be difficult to charge them with monopolization, but all indications are that they are not competing.
In an analysis prepared for the Department of Housing and Urban Development in 1995 and published in 1996, Hermalin and Jaffee concluded—based on a painstaking review of a number of factors, including the GSEs’ extraordinary returns on equity—that Fannie and Freddie were tacitly colluding. That is, they were not competing with one another, and the market they dominate—the secondary mortgage market—was not a competitive market. This does not mean, of course, that Fannie and Freddie are acting unlawfully. Tacit collusion, unlike actual collusion, is not a violation of law. However, if Fannie and Freddie are tacitly colluding with respect to the use of their AU systems—and that tacit collusion allows each of them to act like a monopolist with respect to those who use their respective AU systems—then they could be treated in antitrust analysis as though they were one company holding a monopoly in a single AU system. Of course, in a trial, it might be determined that Fannie and Freddie are actively colluding; a conclusion of this kind, however, was beyond the scope of the Hermalin-Jaffee study.
The next question for a court would be whether the GSEs do in fact have a shared monopoly in the secondary mortgage market and in AU systems. Based on known facts, this is not a difficult question. As noted above, the GSEs are virtually the only purchasers of conventional/conforming mortgages and thus, together, have a duopsony in that market. Needless to say, for purposes of this analysis, there is no significant antitrust difference between or among a duopsony (two buyers), a duopoly (two sellers) and a shared monopoly. Moreover, a February 1999 Morgan Stanley report on the GSEs cites a survey that shows Fannie and Freddie together had 95 percent of the AU market in 1998—that is, the use of AU systems for the assessment of mortgage credit. It is doubtful that this number has declined. A 95 percent penetration of a market—especially under the circumstances present in the secondary mortgage market—would on its face meet anyone’s definition of a monopoly.
However, as discussed by the Circuit Court in Microsoft, mere domination of a market is not sufficient to find illegal monopolization. The monopolist must also have monopoly power—i.e., the power to raise prices or exclude new entry by competitors. On the question of excluding new entry, that condition certainly exists in the secondary mortgage market, where the GSEs hold exclusive federal charters, together with government-granted advantages that provide them with lower interest rates and other advantages that potential competitors cannot match. Accordingly, it will not be difficult for a court to find that the GSEs have a monopoly in the secondary mortgage market, and that monopoly in turn confers a monopoly in the AU market.
In addition, the GSEs appear to have the ability to raise prices for their AU systems without regard to actual costs or the prices that would be charged by their competition. Estimates of the cost of underwriting a mortgage loan on competitive AU systems are in the range of $15. According to sources in the mortgage lending industry, the GSEs’ regular charges for the services of their AU systems are significantly higher than $15—indicating again that they have monopoly power in the AU market.
The Circuit Court also pointed out that holding a monopoly, and having monopoly power to maintain it, is not sufficient to find a violation of the Sherman Act. The monopolist must take exclusionary or anticompetitive acts to maintain the monopoly. Here, there seems to be a great deal of published evidence that the GSEs have been taking such acts to maintain their monopoly. It should be noted that these acts, which might be simply aggressive competition in a competitive market, assume a different character when taken by a monopolist in a non-competitive market.
The GSEs have offered advantages to companies that agree to use their AU systems in preference to competitive systems. These advantages include (i) relief from representations and warranties that can result in a lender having to repurchase a loan from the GSEs—thus increasing the lender’s risk; (ii) discounted AU fees; (iii) providing free or discounted additional software for users of their AU systems; and (iv) bundling their AU systems with software that enables lenders to reduce the time and cost of obtaining credit reports, verification of assets, appraisals, and verification of income. Finally, the GSEs refuse to accept any loans made under their popular low-downpayment loan programs that have not been run through their AU systems.
Under the Circuit Court’s analysis in Microsoft, all of these actions—and especially the last—would be regarded as exclusionary and anticompetitive, taken in an effort to maintain a monopoly. They are akin to Microsoft’s actions in support of its Windows operating system monopoly that were found to have violated Section 2 of the Sherman Act.
The GSEs would, under the Court’s analysis, have an opportunity to show that their actions were in fact procompetitive, and a trial court would then weigh whether the procompetitive actions outweighed the anticompetitive. It is of course difficult to predict at this stage how such a balancing would turn out, but it is also difficult to discern what the GSEs might say in defense of their actions. Clearly, there is nothing explicitly procompetitive in what they have done, but the court might allow them to argue that their use of their proprietary AU systems was necessary in order to protect themselves against the possibility that they might be offered low quality mortgages. Although this is not strictly procompetitive, it is a reasonable strategy.
However, the GSEs would not be able to stop there. They would have to show that their proprietary systems were superior to others, and that they could not reasonably be expected to accept mortgages that were underwritten by other AU systems. This, however, is likely to be a difficult standard to meet. It is doubtful that the GSEs have ever done a comprehensive analysis of the quality of competing AU systems. Indeed, at least one lender that had developed its own system asked Fannie Mae to run a test that would compare both systems. According to that lender, Fannie Mae refused. If—as this incident indicates—the GSEs have not bothered to test potentially competitive AU systems before rejecting them, it is doubtful that a court would accept their argument that they were required by sound business judgment to encourage or compel lenders to use the GSEs’ proprietary AU systems.
Finally, the GSEs’ AU systems are proprietary “black boxes.” It is not possible for a lender to know—before a loan has been run through these systems—whether it will be purchased by the GSEs. Under these circumstances, lenders—to be certain that they are committing to a loan they can subsequently sell to Fannie or Freddie—must pay the cost of using the GSEs’ AU systems. In this way, the GSEs’ monopoly in the secondary market is being used to maintain and extend their monopoly in AU systems.
It seems, therefore, that a strong argument could be made, following the Circuit Court’s reasoning in the Microsoft case, that the GSEs have acted unlawfully to monopolize and to maintain and extend their monopoly of AU systems.
Attempt to Monopolize
On technical grounds, the Circuit Court in Microsoft rejected the District Court’s finding that the company had attempted to monopolize the browser market. The government, in the Court’s view, had not defined the browser market or shown that there were barriers to the entry of competitors, and thus could not show that there was a danger of monopolization. As a result, the Court never discussed in detail the other elements of an attempt to monopolize—specific intent to monopolize and predatory or anticompetitive acts.
In the case of the GSEs, there might be two markets in which a court might find an attempt to monopolize—the AU systems themselves and software related to or used in the mortgage lending process. Attempt to monopolize is a separate violation of the Sherman Act from monopolization itself. If for any reason the GSEs are not found to have monopolized the AU market in violation of Section 2 of the Sherman Act, or to have acted to maintain or extend their monopoly, it would still be possible to conclude they had attempted to monopolize the AU market.
Each of the actions cited above as anticompetitive or exclusionary could also be cited as evidence of an attempt to monopolize the AU market. They are each anticompetitive in nature and evince an intent to monopolize, meeting two of the three tests set out by the Circuit Court in Microsoft. Moreover, unlike the Microsoft case, there is no question concerning whether there is a danger of monopolization; the GSEs already hold an impregnable monopoly in the secondary mortgage market; any service they require a lender to take from them could be an attempt to monopolize if the service is otherwise available from others.
Clearly, the most difficult element of proof in the context of showing an attempt to monopolize would be the showing of a specific intent. All the various communications among Microsoft officials that were used to demonstrate Microsoft’s own intent are of course not available today in the case of the GSEs. They would have to be discovered in the files of the GSEs in the course of a litigation.
However, there are known acts of Fannie Mae’s Chairman and Chief Executive Officer that could be used as evidence of a specific intent to monopolize the AU market. In early March 2001, the Wall Street Journal reported that the Franklin Raines, the Chairman and Chief Executive Officer of Fannie Mae had threatened three executives of large financial institutions—Wells Fargo Bank, American International Group and GE Capital Services—that Fannie would withdraw business from their companies if they remained active in FM Watch, a lobbying group that was attempting to convince Congress to limit the growth or range of activities of the GSEs. One of the activities about which FM Watch had complained was the GSEs’ attempt to monopolize AU systems.
Fannie’s action in this respect—an effort to cause the demise of a competitive group through threats from a monopolist—would be analogous to Microsoft’s threat against Intel, which the Circuit Court found to have showed an anticompetive or predatory intent. Accordingly, it could be considered in the case of the GSEs as evidence of a specific intent to monopolize.
In addition, other publicly known acts of the GSEs could be interpreted by a court as evidence of a specific intent to monopolize the AU market. In particular, the GSEs have offered to evaluate through their AU systems loans that by law they are not permitted to buy or guarantee. A likely reason for this is to prevent the development of competing AU systems that might be established to evaluate loans GSEs cannot buy.
Accordingly, there seems to be ample reason to believe that—using the framework articulated by the Circuit Court in Microsoft—a court could find the GSEs are attempting to monopolize the market for AU systems.
A further question is whether the GSEs’ activities amount to an attempt to monopolize related areas of the mortgage finance process. This is a far more speculative area, but facts as they develop over time could demonstrate that the GSEs are using their monopoly in AU systems or in the secondary market generally to drive out of the mortgage finance process unrelated or uncooperative companies that are engaged in offering title insurance, credit scoring, appraisal, the development of mortgage-related software, credit reporting and mortgage insurance. In some cases, Fannie and Freddie have invested in companies engaged in these activities; in others, they appear only to have established contractual relationships that might involve cross-referral or other arrangements. The exact nature of these relationships are obscure, but in an antitrust case discovery would permit their nature to be exposed and analyzed.
Discussions with lenders indicate that the GSEs may be well along in attempting to monopolize the appraisal market—especially the business of electronic appraisal. The traditional appraisal is done by a specialist in residential values, who visits the property and estimates its value for the lender based on comparable homes and locations. However, for many loans there is now sufficient data on particular properties and locations that it is possible for the appraisal to be done electronically, through data processing, without a visit to the site, or with only a viewing of the exterior of the property. In these cases, which are a substantial portion of all mortgage loans, the GSEs insist on the use of their own electronic appraisals—and charge for this service—even though the electronic appraisals of lenders are as likely to be accurate and are less expensive for the lender and the homebuyer.
Indeed, circumstances in the electronic appraisal market bear a strong resemblance to what has already happened in the AU market. The GSEs’ monopoly in the secondary mortgage market is being leveraged to give them control over other areas of the mortgage lending process.
Efforts on the part of the GSEs to integrate their AU systems with other services would provide evidence that Fannie and Freddie are attempting to monopolize products and services other than AU. In this connection, it is significant that Fannie Mae has recently announced that all access to its AU system, after September 30, 2001, will have to be channeled through MornetPlus 2000, a proprietary Fannie Mae online system that links lenders, realtors and others with providers of other services in the mortgage finance process such as appraisal, credit reports, and title insurance. This will provide Fannie with the opportunity to create and profit from significant competitive advantages granted to selected suppliers of these services. Since Fannie has a monopoly of the AU system that lenders need in order to sell loans to Fannie, its selection of particular service providers—and not others—may amount to a direct suppression of competition.
It is important to recall in this connection, as the Circuit Court emphasized, that the Sherman Act is directed at the protection of competition, not competitors. The ability and willingness of the GSEs to favor certain companies over others—thereby excluding some companies in related fields from the competitive playing field—may amount to an attempt to monopolize under Section 2 of the Sherman Act or contracts in restraint of trade under Section 1.
Illegal tying occurs when a monopolized product or service is used as a lever to require a customer to purchase a product or service for which there would otherwise be a competitive market. The indications, outlined above, that the GSEs are requiring lenders to use the GSEs’ AU systems and electronic appraisal systems could be strong evidence of illegal tying. The tying product or service in this case would be the GSEs’ monopoly in the secondary mortgage market, and the tied product or service would be the GSEs’ AU and electronic appraisal systems. Alternatively, the tying product or service could be the AU system and the tied product or service the electronic appraisal system.
As explained by the Circuit Court in Microsoft, tying is normally a per se violation of the Sherman Act. However, in the Microsoft case, because of the technical nature of the tying and tied products, the Court refused to affirm the District Court’s finding of a per se violation through tying. Instead, the Court held that the issues associated with a tying violation should be reviewed under the “rule of reason” and remanded the case for further proceedings that would resolve some of the issues that seemed unclear.
There is no reason to treat the GSEs’ use of their secondary market or AU monopoly as anything other than a standard case of possible tying. For reasons outlined above, there is no sense in which the GSEs’ secondary market activities or AU systems present the difficult technical issues that compelled the Circuit Court to require a “rule of reason” test. Therefore, if tying can be found, under prevailing antitrust precedents it would be a per se violation of Section 1 the Sherman Act.
Accordingly, it will be necessary only to show that the tying products or services—the GSEs’ secondary mortgage market monopoly and/or their AU systems—and any tied products or services, are in fact distinct products; that the GSEs have market power in the tying product or service; that consumers have no choice but to purchase the tied product or service; and that the tying arrangement forecloses a substantial amount of commerce.
There appears to be more than enough available data to make out a strong case of tying against the GSEs with respect to their AU systems. There, it seems clear that their secondary market monopoly has been used to require customers to use their AU systems, and that this has foreclosed a substantial amount of commerce. This may also be true of electronic appraisals.
On the other hand, there is not sufficient data yet available publicly to make out a case of tying against the GSEs for all the other services—credit reports, mortgage insurance, title insurance, closing services and the like—that they seem to be integrating with their AU systems. However, as the Circuit Court’s Microsoft analysis becomes more widely known, it may be that victims of tying—who don’t yet realize that the antitrust laws are applicable—will come forward to provide evidence of tying arrangements.
The Microsoft case provides a road map for analyzing the potential antitrust liabilities of the GSEs in three areas: efforts to maintain or extend their AU monopoly; attempts to monopolize the AU market; and tying of other services or products to their monopolized AU systems. On the basis of the Circuit Court’s analysis in Microsoft, a strong case can be made, using information already in the public domain, that the GSEs have violated and are violating the Section 2 of the Sherman Act by monopolizing and acting to maintain their monopoly of the AU market. A similarly strong case can be made that the GSEs are attempting to monopolize the AU market and the market for electronic appraisals. There also appears to be a strong case that the GSEs are illegally tying their AU and appraisal systems to their secondary market monopolies This would be a per se violation of Section 1 of the Sherman Act. However, there is as yet not enough information currently available to permit a conclusion that—other than electronic appraisals—the GSEs are tying other goods or services to their AU systems. If such a case could be made it would be a per se violation of the Sherman Act.
Section 2. Monopolizing trade a felony; penalty Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons, to monopolize any part of the trade or commerce among the several States, or with foreign nations, shall be deemed guilty of a felony, and, on conviction thereof, shall be punished by fine not exceeding $10,000,000 if a corporation, or, if any other person, $350,000, or by imprisonment not exceeding three years, or by both said punishments, in the discretion of the court.
 Section 1 of the Sherman Act (15 USC Section 1) states:
Section 1. Trusts, etc., in restraint of trade illegal; penalty Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal. Every person who shall make any contract or engage in any combination or conspiracy hereby declared to be illegal shall be deemed guilty of a felony, and, on conviction thereof, shall be punished by fine not exceeding $10,000,000 if a corporation, or, if any other person, $350,000, or by imprisonment not exceeding three years, or by both said punishments, in the discretion of the court.
 Wallison, Peter J. and Ely, Bert, Nationalizing Mortgage Risk: The Growth of Fannie Mae and Freddie Mac, American Enterprise Institute, 2000.
 Hermalin, Benjamin E. and Jaffee, Dwight M., The Privatization of Fannie Mae and Freddie Mac: Implications for Mortgage Industry Structure, in Studies on Privatizing Fannie Mae and Freddie Mac, Department of Housing and Urban Affairs, 1994, p. 225 et seq.
 Statement of Donald Lange in Bank Technology News, November 1999; statement of Becky Poisson in Mortgage Technology, July 1998.
 Report on decision of St. Francis Capital to use Fannie Mae AU system, Milwaukee Business Journal, July 6, 1998.
 The Detroit News, October 13, 1997; Newsday, December 12, 1997.
 Wall Street Journal, March 8, 2001.
 See, e.g., FM Watch, GSE Mission Creep: The Threat to American Consumers, March 2001, p.3