Mr. Chairman and members of the subcommittee:
It is a privilege for me to testify this afternoon on the subject of Fannie Mae and Freddie Mac, and I'd like to congratulate and thank you, Mr. Chairman, for taking on an important task that deserves much more attention from Congress than it has received.
It is important to recognize the significance of the accounting problems at Freddie Mac--not because these problems are especially severe, but because they were a surprise and seem to arise from something so routine. From press accounts, it appears that Freddie attempted over many years to manage its earnings by manipulating the valuation of its derivatives. This is known as managing earnings, and its objective is to create a smooth upward curve. Freddie Mac was so good at this that it was nicknamed "Steady Freddie" on the Street. Some attention is now also being paid to Fannie Mae's financial reports, which, despite the vicissitudes of the mortgage market, interest rates and the economy generally, also showed the same smooth upward curve. Managing earnings is very easy to do under Generally Accepted Accounting Principles (GAAP)--so easy that many companies are suspected of doing it.
That is not so much an indictment of these companies as it is an indictment of the excessive reliance that has been placed on GAAP financial disclosure by the SEC, media commentators, and--most recently--Congress in enacting the Gramm-Leach-Bliley Act. The fact is that GAAP financials are highly malleable, and should not be considered an index of the financial condition or prospects of companies. Because the principal constituents of a GAAP earnings statement are predictions about the future--what losses will be suffered on a portfolio of receivables, what reserves should be established for future claims--bottom line financial results reflect simply the judgments of management rather than a true picture of the company's financial condition.
I mention this because many commentators and policymakers seem to believe that requiring Fannie Mae and Freddie Mac to file reports with the SEC will substantially reduce the risks they pose to the taxpayers and to the economy. This idea is as misplaced as requiring a whole new structure to regulate how accountants do audits, as Congress did in the Sarbanes-Oxley Act. No matter how effective an audit, it can never make financial statements prepared under GAAP more "accurate." The key decisions that are made by management in preparing a company's financial statements cannot be made by auditors, who can only determine whether GAAP has been followed. Accordingly, while I believe it would be worthwhile to have Fannie and Freddie register their securities with the SEC, we should not think that doing this--even if it produces improved disclosure--will protect us against the risks they pose.
Because of the uncertainties associated with GAAP, it is not correct to believe that Fannie Mae and Freddie Mac are financially strong companies simply because they are producing earnings or have strong-looking balance sheets. It's likely that they are both profitable and financially strong, but we really can't know for sure. A demonstration of this is the fact that OFHEO--Fannie and Freddie's regulator--was not aware of the true extent of the company's financial problems until advised of them one day before they were announced. If their regulator could not find their financial problems, how is the general public--or Congress--supposed to do it?
This points to another very weak reed in our general defense against the risks created by Fannie and Freddie. We count on regulators to find and correct the most serious problems before they grow out of control. But in relying on regulation we are again deluding ourselves. Occasionally, regulators stumble upon things like bad accounting, but in most cases they are in the dark until someone tells them about the problem. Thus, I don't blame OFHEO, or believe that it is a weak or incompetent regulator because it failed to uncover or understand the gravity of the accounting problems at Freddie. This is what we should expect from any regulator, because it is the most likely outcome. Regulators work in the bowels of the organizations they regulate, but the big decisions--the ones that can really cause the losses at a company--are made at the top level, where regulators generally have no regular access.
Thus, we ought to be clear-eyed about both the effectiveness of regulation and the usefulness of GAAP accounting, and adjust our policies accordingly. In the case of Fannie and Freddie, as I will argue below, we can't afford to make a mistake. If their financial statements do not disclose their real vulnerabilities, and no regulator will find these vulnerabilities, we are courting serious problems if we continue to let them grow with the support of the federal government. In effect, we are creating a new S&L crisis, but one that will be much larger and more consequential for the economy.
We should keep in mind that, together, these companies had--at the end of 2002--approximately $3.3 trillion in liabilities. $1.5 trillion of this was in the form of debt, and $1.8 trillion in the form of guarantees of mortgage backed securities. Against these liabilities they hold only about 3 percent in capital--a percentage far lower than that permitted to regulated commercial banks. As outlined later in this testimony, obligations in this range pose enormous potential problems for the nation's taxpayers and for the economy at large.
But even these risks might be worth taking if Fannie and Freddie produced substantial benefits for the economy or for homebuyers. However, this is not the case. In fact, Fannie and Freddie deliver relatively little value to the economy or to homebuyers--and certainly not enough value to justify the risks they are creating. Since we are not likely to be protected against those risks either by better financial disclosure or regulation, it seems prudent to consider other ways that the risks might be reduced, preferably without any adverse effects on the mortgage markets or on homebuyers.
In the balance of this testimony, I will try to show that Fannie and Freddie provide relatively little benefit to the economy or to homebuyers generally, and that what little benefit they provide is overwhelmed by the risks they create. Since neither better accounting or auditing, nor better regulation, is likely to save us from the consequences of these risks, we should consider policies that will reduce risks in other ways. I believe the best way to eliminate these risks is to privatize Fannie Mae and Freddie Mac, and break them up into smaller competitive components, but I recognize that this idea is not at the moment politically feasible. So at the end of this testimony I suggest another way to substantially reduce the risks they pose, without either privatizing Fannie and Freddie or adversely affecting the mortgage market.
First, however, I will discuss the balance of benefits and costs that I see in Fannie and Freddie.
Weighing Benefits v. Risks
The case against Fannie Mae and Freddie Mac is very simple: they create enormous risks for the government, for the taxpayers, and for the economy as a whole, yet provide no significant benefit to homebuyers. Accordingly, Congress should take steps to cut their links to the federal government. Like the S&L crisis many years ago, procrastinating will only put off the day of reckoning, and the problem will be worse and more costly when Congress is finally compelled to act. Fannie and Freddie have been doubling in size every five years, and now have combined liabilities of $3.3 trillion. This is not a problem that can be safely or responsibly put off.
Fannie Mae and Freddie Mac were created for a single purpose--to provide liquidity for the housing finance system by creating a market for the mortgages made by banks and other mortgage originators. They did this very well. There is now a vibrant and efficient secondary market for residential mortgages. The technology has been developed, investors have been educated, a distribution system has been established. The structure will now operate without government assistance of any kind. In fact, in the so-called "jumbo" market--mortgages larger than Fannie and Freddie are permitted to buy--it operates entirely without any government backing. So Fannie and Freddie are no longer necessary for their original purpose. They should be thanked and sent home.
In fact, Fannie and Freddie know all this. So they have been diligent in creating a rationale for themselves that does not depend on their providing liquidity to the housing market. They have been advertising instead that they "open the doors to home ownership" by reducing the cost of mortgages, or that they are in "the American dream business" because they enable people to buy homes who might otherwise not be able to do so, or--implicitly--that they help minorities to become homeowners.
However, they do not really do these things. Let me take them one at a time.
Helping people afford homes. The basis for this claim is the correct observation that interest rates on mortgages purchased by Fannie and Freddie are somewhat lower than rates on so-called "jumbo" loans--which are sold in an entirely private secondary market. There have been many studies of the degree to which Fannie and Freddie provide lower interest rates to buyers who can qualify for conventional conforming loans. Table 1, attached, is a compilation of such studies that was presented at an AEI conference in October 2002. It shows that the effect of Fannie and Freddie's activities is to reduce interest rates on home mortgages by a very small amount--somewhere in the range of 25 basis points, or 1/4 of 1 percent. If I can put this in perspective, every time the Fed lowers interest rates by one-quarter point, it has the same effect, and the Fed has done this 12 times in the last two years. Similarly, every time the Fed raises interest rates 1/4 point it has the opposite effect. If that 1/4 point were as important as Fannie and Freddie suggest in their advertising, thousands and thousands of American families would be frozen out of home ownership every time the Fed raises interest rates by 1/4 point.
Moreover, this benefit comes almost entirely from the implicit support Fannie and Freddie receive from the government, not because of anything particularly special that Fannie and Freddie bring to the market. The Congressional Budget Office has estimated that in 2000 Fannie and Freddie received implicit government support with a value of about $10.6 billion, of which about two-thirds was actually made available to the mortgage market through lower rates. The balance, presumably, increased the share values of Fannie and Freddie by increasing their bottom line profitability, and went to management compensation.
This small one-quarter point benefit, however, is not a very good argument for continuing the implicit government subsidy. First of all, it's a very inefficient way of subsidizing the housing market. About one-third of the benefit the government has conferred on Fannie and Freddie goes to their shareholders and managements, rather than to create lower interest rates. This is surely an extreme form of corporate welfare, in which two managements and their investors are enriched in order to confer limited indirect benefits on homebuyers. If Congress wants to subsidize housing, it should be able to find a more efficient way to do it.
But second, and much more important, it isn't even clear that the subsidy--limited as it is--goes to homebuyers. It's entirely possible that it simply causes home prices to rise. In other words, it is a subsidy to home sellers and developers. I don't know of any studies that show this--nor of any studies that show the opposite--but it is common sense that to the extent that the monthly payments required of homebuyers are reduced, it provides an opportunity for home sellers to raise their prices.
Putting people in homes. Fannie and Freddie argue that the small reduction in interest rates that they pass along to the mortgage markets out of their implicit government subsidy contributes to the growth of home ownership in the United States by helping people buy homes. However, a study by the Census Bureau, also presented at an AEI conference in October, showed that the monthly cost of owning a home is not the obstacle that prevents renters from buying homes. The obstacle is the down payment. Renters do not generally have the financial resources necessary to buy their first home. Accordingly, the claim that Fannie and Freddie put people in homes by reducing interest rates is not true. No amount of interest rate reduction will make it possible for renters to become homeowners, because the problem for them is not the carrying cost of owning a home--it is the fact that they cannot accumulate the necessary down payment.
This reality led my colleague at AEI, Professor Charles Calomiris, to propose that Fannie and Freddie be completely privatized and the implicit subsidy they now receive used to provide down payment assistance to families who would otherwise be unable to purchase a home. Professor Calomiris estimated that this use of the Fannie and Freddie subsidy would permit more than 600,000 families, now renting, to buy homes.
Helping minority families. Through their advertising, which prominently displays photos of minority families in or in front of what are presumably their homes, Fannie and Freddie suggest that they provide special assistance to minority families hoping to become homeowners. And if they did this disproportionately--that is, helped minorities or low income borrowers more than they helped middle class borrowers--that would be a powerful argument for preserving their current status.
But they do not do this. Instead, according to a study by Jonathan Brown of Essential Information, a Nader-related group, Fannie and Freddie buy proportionately fewer conventional conforming loans that banks make in minority areas than they buy in middle class white areas. Other studies have shown that the automated underwriting systems that Fannie and Freddie use to select the mortgages they will buy approve fewer minority homebuyers than similar automated underwriting systems used by mortgage insurers. There is at least one lawsuit against Freddie Mac by a minority homebuyer, arguing that he was unable to get a conventional conforming mortgage because of the exclusionary nature of Freddie's automated underwriting system.
The sad fact is that Fannie and Freddie--two government sponsored enterprises that have a government housing-related mission--do less for minority housing than ordinary commercial banks. Studies have repeatedly shown that banks and other loan originators make more loans to minority borrowers than Fannie and Freddie will buy. That in itself should be a scandal, together with the fact that both companies seek through their soft-focus advertising to create the impression that they are actually using their government benefits for the disadvantaged in our society.
So the U.S. housing finance system gets very little benefit from the continued existence of Fannie Mae and Freddie Mac as government sponsored enterprises. The reduction in interest rates that they can point to as a result of their activities is really the result of their implicit government support, which is small in any case, and is swamped by macro changes in interest rates as a result of economic conditions. In any event, it isn't even clear that the lower rates operate as a benefit to homebuyers rather than home sellers. This small reduction in interest rates does not put people in homes or improve homeownership rates in the United States because most renters lack the down payment necessary to buy a home, not because they could not afford the monthly carrying cost of homeownership. And finally, despite the implications of their advertising, Fannie and Freddie seem to discriminate against minority homebuyers rather than assist them.
The Costs of Fannie and Freddie
So the benefits of continuing Fannie and Freddie as GSEs are meager to non-existent. What then are the costs?
I have already cited the CBO estimate that Fannie and Freddie receive an implicit subsidy from the U.S. government--in effect an extension of U.S. government credit--with an annual value of at least $10.6 billion. That, however, is not the extent of their cost to the taxpayers. Because their securities directly compete with Treasury securities--in fact they have begun to issue securities on a regular schedule, just like Treasury, in order to be a more effective substitute--they cause Treasury interest rates to rise slightly, probably by a few basis points. On a total Treasury debt of several trillion dollars, those few basis points amount to hundreds of millions of dollars annually.
But these two costs do not begin to describe the potential costs to the government, the taxpayers and the economy of allowing Fannie Mae and Freddie Mac to continue to grow. Because Fannie and Freddie are implicitly backed by the U.S. government, financial problems at either of them could require a government bailout. This is what Congress has had to do with other GSEs--most recently the farm credit system in the mid-1980s--and there is no reason to suppose that Congress would not step in if Fannie or Freddie, or both, were in financial trouble.
Until June of this year, when Freddie Mac dismissed its top three officers and announced that it would have to do a considerably bigger financial cleanup than we initially thought necessary, it was possible to say that both Fannie and Freddie were in strong financial condition and that there was no prospect of a bailout. Since then, however, there has been much more scrutiny of the financial statements of both companies, and at least some observers have pointed out that while Freddie might have been more profitable than it reported during the three years ending in 2002, Fannie Mae might actually have lost money, or made no profits, last year. That is not what Fannie reported, which was of course another huge annual increase in profitability. The problem is, because of the malleable nature of Generally Accepted Accounting Principles (GAAP), we don't really know how these complicated companies are doing. We would get a better picture of Fannie and Freddie's actual condition with better cash flow reporting, but that is not currently required by GAAP or the SEC.
In any event, however they are doing today, changes in interest rates and the economy generally could have a significant adverse effect on their financial health in the future, and the taxpayers are ultimately responsible for assuring that they meet their obligations. It is important to remember in this connection that, at the end of 2002, Fannie and Freddie had an aggregate of $3.3 trillion in liabilities. Even a small part of this obligation--if it has to be made up by the taxpayers--will make the S&L bailout look like a dimestore operation.
But even that does not end the risks we all face with these two companies. Because they are integral to the health of the housing market, the failure of either of them could have a systemic effect--meaning an adverse effect on the economy as a whole. It's relatively easy to see how this might happen. Fannie and Freddie, together, purchase almost all the conventional conforming mortgages that come on the market each year. They currently hold or guarantee 75 or 80 percent of all conventional conforming mortgages and almost half of all residential mortgages in the United States. If either Fannie or Freddie were to lose the confidence of the capital markets, and were unable to purchase their share of new mortgages as these came on line, the entire residential finance system would be seriously disrupted--at least temporarily. Interest rates would rise and residential mortgages would be harder to get. This would rapidly affect the rest of the economy. Home sales would decline, construction would fall, sales of home furnishings and appliances would suffer.
This effect would be bad enough as it ripples through the economy. Much worse would be the effect on the financial system as a whole. Large numbers of banks and other financial institutions are major investors in the securities of Fannie and Freddie. They are encouraged to buy and hold Fannie and Freddie securities by a statutory exemption for these securities from regular restrictions on loans to one borrower. Declines in the value of Fannie and Freddie securities will reduce, and in some cases impair, the capital of all these financial institutions. Reduced or impaired capital will reduce the amount of credit they can provide, even outside the mortgage markets.
Altogether, then, the effects of a failure or severe financial crisis at either Fannie or Freddie could be systemic in character, not limited to the home mortgage markets. And since there are only two of these companies, it is accurate to say that the continued health of our economy depends on decisions by only two corporate managements. If one of them makes a grave mistake, the entire economy could suffer. And the recent events at Freddie Mac show that management judgments are far from infallible. We don't know the extent of the problems at Freddie, but we do know that the top management made serious errors of judgment. These, fortunately, do not appear to threaten systemic effects, but errors of judgment come in many shapes and sizes, and one day the error may be of a kind that cannot be repaired by accountants working around the clock.
What to Do
So what is to be done? We have a situation in which two companies create enormous risks for the taxpayers and the economy, but offer little in the way of benefits to anyone. Congress has it within its power to change this calculus in a number of ways. My preferred answer would be to privatize Fannie and Freddie and at the same time break them up into five or six smaller entities. In nature, diversity protects a species; in finance, diversity can protect an economy.
However, I am aware that this solution is not for the moment on anyone's radar screen. So I have a more modest proposal: Congress should prohibit Fannie and Freddie from buying back or accumulating any substantial portfolio of mortgages or mortgage backed securities (MBS).
Today, these companies do business in two very different ways: (i) they create pools of mortgages which are used to collateralize MBS that they guarantee and sell to investors, and (ii) they buy whole mortgages and repurchase the MBS they have already sold to investors.
These are two very different ways of performing their functions, and have very different consequences. When Fannie and Freddie create pools of mortgages and sell MBS backed by these pools, they are guaranteeing that investors will receive a stream of revenue derived from the interest and principal paid into the pools by homeowners paying off their mortgages. In this case, Fannie and Freddie are taking only credit risk--the risk that homeowners will not meet their mortgage obligations. This is not a very significant risk, especially today, when losses on mortgage pools have been running at 1 or 2 basis points.
However, buying and holding mortgages or MBS is an entirely different story. In that case, Fannie and Freddie must take interest rate risk in addition to credit risk. Interest rate risk--that rates will rise or fall--is a far greater risk than credit risk, and requires Fannie and Freddie to buy derivatives of various kinds to protect themselves against the vicissitudes of the credit markets. To put this in perspective, it was interest rate risk that caused the failure of the S&Ls. They were holding mortgages that were paying, say, 5 percent, but in order to finance these loans they had to pay 10 or 12 percent for their funds when interest rates rose. With a negative spread like that, they weren't solvent for very long. Fannie and Freddie are in the same position, but their risks run two ways. The same thing happens to them if interest rates suddenly go up--they are holding mortgages that may yield less than the new rate they have to pay for their funds. But they also run risks if interest rates go down, since a portion of their portfolio is funded with longer term debt. If interest rates decline, homeowners refinance, and Fannie and Freddie end up holding, say, 4 percent mortgages, that they've funded with 5 percent liabilities. Another losing proposition.
Why, you might ask would Fannie and Freddie take such risks? Why would they buy back from investors the MBS on which the investors are already taking the interest rate risk? The answer is that, even after buying all that hedging protection through derivatives, it is still profitable for them to buy and hold their own MBS. In fact, it has been estimated that Fannie and Freddie own, in the aggregate, 34 percent of all MBS currently issued. With their government backing, they can borrow money at rates low enough so that they can do a rather simple arbitrage, profiting from the spread between their cost of funds and what the MBS are yielding.
Now one might think that somehow buying back their MBS will have the effect of lowering interest rates for mortgages, but this is not the case. Economists point out that borrowing funds to buy back other credit instruments is simply a wash. It doesn't have any effect on mortgage rates, which are a product of all the funds available in the capital markets.
Accordingly, what we have here is the classic case of privatizing the profits and socializing the risk. Fannie and Freddie profit from arbitraging their government backing, but the people really taking the risk are the taxpayers.
Accordingly, if Congress does not currently have the stomach for privatizing Fannie and Freddie, it can at least reduce the risk they pose to taxpayers and to the economy generally by prohibiting them from buying back the MBS they issue and from holding a large portfolio of mortgages. Instead, their activities should be limited to forming pools of mortgages and selling MBS that they guarantee. The risks on this--which is simply credit risk--are far less than the interest rate risk they have been taking, and it would have no effect on mortgage interest rates.
This is at least a temporary solution to the problems posed by Fannie Mae and Freddie Mac. Because we cannot rely on either accounting or regulation to protect the taxpayers and the economy against a serious mistake by the managements of Fannie or Freddie, or both, we should be thinking of ways to reduce that risk. By prohibiting the purchase of their own MBS or accumulating a large portfolio of mortgages, we can significantly reduce the risk that these two enterprises currently create for taxpayers and the economy, without any effect on interest rates on conventional conforming mortgages.
That concludes my testimony, Mr. Chairman.
Peter J. Wallison is a resident fellow at AEI.