Skip Navigation

June 21, 2001
 

Mr. Franklin D. Raines
Chairman and Chief Executive Officer
Fannie Mae
3900 Wisconsin Avenue, NW
Washington, DC 20016

Dear Mr. Raines:

I am writing to respond to your written comments, dated May 9, 2001, on the Congressional Budget Office's draft study on the housing GSEs and to reply more generally to some related arguments, including ones in a recent speech by Tim Howard, "Fannie Mae's Benefits to Home Buyers: The Business Perspective."

The first set of comments concerns the calculation of tax and regulatory benefits that the GSEs receive. You argued that the subsidy is exaggerated when figured on a pretax basis. CBO routinely does all such calculations before taking federal taxes into account. Moreover, it is entirely possible that if calculated on an after-tax basis, the ratio of benefits to Fannie Mae and benefits to homeowners would increase.

The second set of issues you raise involves the calculation of the debt benefit. You asserted that the spread between portfolio returns and debt should be 78 basis points rather than 90 basis points. Ninety basis points, however, is the spread reported in Fannie Mae's annual report, which is based on mortgages rather than the GSE's overall portfolio. You also made the point that the total debt in Table 2 of CBO's report includes debt that is not used exclusively for conforming mortgages but also for multifamily mortgages and nonmortgage investments, whereas the subsidy to multifamily borrowers is not included in the pass-through estimates. In fact, we based the subsidy estimates on the full increase in the amount of debt outstanding and so included a subsidy to multifamily borrowers. The total subsidy calculation implicitly includes all subsidies that are passed through.

As to the issue of the appropriate comparison group, I refer you to the study, which explains in greater detail why CBO used that particular set of firms. We understand that this comparison could overstate the subsidy calculation but not by much. Your own consultants produced estimates that are consistent with ours. Further, many other assumptions that we made are favorable to Fannie Mae and Freddie Mac.

You also objected to our citing the 1996 Treasury Department study that relied on a comparison group of firms with single-A credit ratings, given that Fannie Mae and Freddie Mac have since been rated AA-. Treasury's approach is still valid. The department noted that a single-A credit rating is typical of large high-quality fully private financial firms holding portfolios of residential mortgages.

You raised several objections to the division between short- and long-term debt. In the main, we assume that Fannie Mae would not synthetically create long-term debt if it was cheaper to simply retire its short-term debt and issue long-term securities. Therefore, the same subsidy must apply to "effective" long-term debt--unless there is some reason that Fannie Mae or the markets are not efficient in this regard.

On a related point, we did not claim that Fannie Mae acts to maximize the subsidy. We believe, however, that your institution operates as efficiently as you assert it does and that it would not unnecessarily pass up opportunities to reduce costs and pass savings on to home buyers.

Regarding the calculation of the subsidy to mortgage-backed securities (MBSs), there are fundamental reasons, discussed below, to believe that the GSEs can charge more than a private firm can for an otherwise identical guarantee. The fact that the structure of private guarantees is often different is noted in the study, but that point is not relevant to the issue of whether the GSEs have an advantage, arising from their government backing, in the MBS market.

Perhaps most fundamental is the question, also raised by Mr. Howard, of whether Fannie Mae and Freddie Mac receive a federal subsidy. It is true that federal subsidies to GSEs do not appear in the federal budget and are largely determined by the GSEs' own actions (since they grow in proportion to the GSEs' business activities). Nevertheless, subsidies exist in the form of tax and regulatory advantages as well as lower borrowing rates and higher guarantee fees resulting from the perceived government guarantee and other legal preferences. Those subsidies represent costs to the economy and the public, and they are highly valued by their recipients. As Mr. Howard noted, "Do some segments of the credit markets receive less financing, or pay a higher rate for that financing, because of Fannie Mae's and Freddie Mac's effectiveness in channeling capital to homeowners? Almost certainly yes, as most economists would attest." But with regard to the retained subsidy, the question is not, as Mr. Howard suggests, what would happen to the subsidy if the GSEs' charters were taken away completely but, rather, how much Fannie Mae or a competing financial firm would be willing to pay for a similar charter.

You contend that competition forces the entire subsidy (or "benefits") to be passed through to mortgage borrowers. We maintain that shareholders and other stakeholders of the GSEs are able to retain a portion of the subsidy because of the competitive advantage conferred by the GSEs' special status. Contrary to what you assert, we explicitly took the structure of the industry into account, as did our many outside reviewers.

Several times in your commentary on our study, you mentioned our alleged failure to correctly characterize the "markets." For example, you stated that the housing GSEs have only a 27 percent share of the mortgage market. It is clear, however, that they control close to three-quarters of the relevant market (30-year conforming fixed-rate mortgages) and are continuing to expand that share.

Similarly, according to Mr. Howard, "There is no shareholder 'subsidy' in this [MBS] fee. The fee is determined by competitive negotiations among Fannie Mae, Freddie Mac, and hundreds of individual lenders, including many large lenders with significant market leverage." Of course, the number of buyers is largely irrelevant to whether sellers have a competitive advantage. In fact, sellers have more power in a market with many competing buyers than in a market with a small number of buyers who can effectively collude.

The real question is whether there are close substitutes for a guarantee sold by Fannie Mae or Freddie Mac, and the answer is, there are not. A guarantee offered by a GSE is clearly more valuable than an otherwise similar guarantee offered by a private firm, owing to the perception that the GSE guarantee has the added protection of government backing. Put differently, a GSE can offer a guarantee with less capital backing than a private firm requires, but the GSE can demand a similar price because the government's perceived backing is a substitute for the comfort of additional capital. The GSEs have a lower cost of capital; the difference between the associated capital costs of a private firm and those of a GSE constitutes a subsidy.

The argument that many other participants in financial markets also receive some form of government assistance and are therefore effective competitors is not persuasive with regard to credit guarantees. The fact that the GSEs are virtually the only guarantors of conforming mortgages is strong evidence that they have a competitive advantage over other financial institutions. The GSEs' issuance of more debt and MBSs than the amount of new 30-year fixed-rate mortgages further confirms their dominance of this market.

Another point of disagreement between us is whether the GSEs can retain a portion of the subsidy associated with their portfolio investments--the arbitrage that drives their earnings. Let us begin with those facts that we agree on. We agree that the interest rates on GSE debt issues are determined in competitive markets with many investors and that investors consider the credit quality of GSE debt issues and of MBSs with a GSE guarantee to be the same. We also agree that GSE debt and MBSs differ in their overall risk characteristics and therefore are not perfect substitutes. (For instance, MBSs have prepayment risk but noncallable GSE debt does not.)

As Mr. Howard states, "Many investors do not want to bear the prepayment or cash flow uncertainty of MBSs. Instead, they prefer the cash flow certainty of bullet debt, or the relative cash flow predictability of callable debt." You have argued that the potential exists for arbitrage between the two markets, causing the GSEs to relinquish the subsidy. But that loss is impossible precisely because of those risk differences.

In fact, it strains credulity to suggest that someone, through arbitrage, can strip the GSEs of their subsidy. Of course, anyone can mimic Fannie Mae's portfolio, but unless a private firm receives similar advantages, its risk profile will be much higher (or the MBS markets are badly mispricing Fannie Mae's guarantees).

To elaborate on that issue, an investor taking the financial position suggested by Mr. Howard (selling Fannie Mae's debt short and buying MBSs) is exposed to risk because the payment characteristics of the two kinds of securities differ considerably, as described earlier. That is not arbitrage but, rather, a risky investment. Taking a so-called short position involves significant costs, which the GSEs do not face as the original issuers of debt; thus, the positions of the GSEs and unsubsidized investors are not equivalent. If they were, every firm would finance all of its investments by shorting subsidized risk-free debt rather than issuing its own securities. While the risk and return characteristics of such a portfolio structure might be attractive to Fannie Mae, particularly given its ability to hedge elements of its risk in the derivatives market relatively cheaply, there is no reason to suppose that an unsubsidized investor would find it profitable to hold a similar portfolio.

To repeat, we agree that both MBSs and Fannie Mae's debt are priced in two very competitive markets and that the market prices of the securities offer investors a fair return, given the risk of each type--again, a statement that is true but irrelevant because it does not mean that the GSEs compete with other potential purchasers of mortgages on a level playing field. The markets are not interchangeable, and operating in both, as the GSEs do, represents an opportunity for profit that is not available to other investors. The problem with Mr. Howard's argument is that it neglects the difference in risk tolerance between different sectors of the market, the costs of short selling, and the differences in counterparty risk.

Another way to understand the advantage is to consider a non-GSE whose portfolio has the same structure as Fannie Mae's. Such a firm would receive an equal return on mortgages but would pay more for the debt used to finance those mortgages. Furthermore, if the private competitor increased the size of its portfolio, the rates that its debt holders would demand would climb as well. The interest rates on the GSEs' debt are less sensitive to portfolio risk because of the perceived government guarantee, allowing the GSEs to take on risk without facing immediate discipline by the market.

It is worth repeating the fact that in every year since 1997, the debt issued and MBSs guaranteed by Fannie Mae and Freddie Mac have exceeded the amount of new fixed-rate conforming conventional single-family mortgages. The housing GSEs are borrowing and guaranteeing more than the annual supply of new mortgages by using their funding advantage to bid more for mortgages than their unsubsidized competitors can pay and to make a profit on mortgages financed with debt issues. This means that the GSEs' relevant market share, now close to three-quarters, will continue to grow. It also means that there are very few new conforming mortgages that are not touched by the GSEs--hence, there is little benefit to the "rest of the market," as you claim.

As your final point, you asserted that because other financial intermediaries also receive various types of government backing, ". . . developing a 'subsidy' measure that applies only to Fannie Mae and Freddie Mac is not a meaningful or relevant exercise." There is, of course, no logical inconsistency in asking how much of a subsidy the GSEs receive, whether or not other institutions are also subsidized. The real issue is whether those subsidies to other institutions imply that the GSEs have no competitive advantage.

Once again, the dominance of the GSEs in the conforming mortgage market suggests that other financial institutions cannot effectively compete. That inability can be explained by differences in the nature of the federal subsidies that are conferred. The subsidy to Fannie Mae and Freddie Mac is more tightly targeted than are subsidies to other financial intermediaries. Fannie Mae and Freddie Mac can take advantage of the favorable rates on their long-term debt only by borrowing and investing, and their investments are limited to conforming mortgages.

That limitation is not surprising because the government's assistance to the GSEs is clearly targeted toward the mortgage market, whereas the advantages conferred on other financial institutions are tied to other markets and are more diffuse. The banks with which the GSEs compete have no particular advantage on long-term debt, although they may receive a subsidy on short-term deposits as a result of deposit insurance. The fact that banks have captured a large part of the market for adjustable-rate mortgages (ARMs) is consistent with the idea that they can effectively compete for short-term investments but not for fixed-rate mortgages.

I hope and expect that the fact that the housing GSEs have a funding advantage is not at issue--if it were, the GSEs would have no need to resist so-called "privatization." Rather, you take issue with some but not all of the assumptions we made (and made explicit) in our analysis--but only those assumptions that you feel hurt your case and not those that work the other way. You also selectively characterize markets in a way that seems to bolster your case--characterizations that are often irrelevant and occasionally misleading. Fannie Mae and Freddie Mac intermediate a large and growing share of the 30-year conforming fixed-rate market. With only two competitors, that result is not surprising.

Sincerely,

Dan L. Crippen
Director