For this second post since “Howard on Mortgage Finance” went live, I thought I would start by briefly summarizing the goal of the site and addressing two possible concerns people might have about my motivations for creating it.
The site’s goal is simple: to improve the quality of information available about the relative merits of various approaches to reforming the U.S. mortgage finance system. There is a large amount of misinformation currently being circulated on this subject, which in many cases my background and experience put me in a position to correct. Where I can I will try to do so.
Few disagree on the basic objectives of mortgage reform: to develop and implement a capital-markets based secondary market system capable of financing very large volumes of 30-year fixed-rate mortgages, throughout the business cycle, for the largest number and the widest range of qualified homebuyers, at the lowest cost consistent with a defined standard of taxpayer protection. Where the disagreements arise is over whether other proposed features of the system—be they ideological imperatives or restrictions, or a prominent role for certain institutions or business processes—unnecessarily compromise the attainment of those objectives, at homebuyers’ expense.
If only because of the lawsuits against the Net Worth Sweep, the current “indefinite conservatorship” of Fannie Mae and Freddie Mac is not sustainable over the long term. For both political and practical reasons, few support a formally nationalized secondary mortgage market. That leaves only two realistic alternatives: a system based on a reformed version of the entities we already have—Fannie and Freddie—or a de novo system, such as the one proposed in the Johnson-Crapo legislation.
Between these, my knowledge and experience lead me to favor the former. I think it’s for valid reasons—many of which I will discuss in future posts—but I recognize that some may suspect that my views reflect a bias towards my former employer, or, as a holder of Fannie Mae shares, a financial motivation for advocating a future that would make those shares more valuable. I’d like to quickly address both concerns.
The first is easy: the Fannie Mae that employed me for 23 years no longer exists. The aftermath of the 2004 “accounting scandal,” the subsequent dismantling of the credit risk management systems I helped put in place, the mortgage and housing bubbles, the 2008 conservatorship, and the mandatory shrinkage of the company’s mortgage portfolio all changed it irrevocably. There is no going back to the version of Fannie Mae I worked for; what is possible is to take the best features of that company—which I know well—and incorporate them into the system we design for the future.
I do own shares of Fannie Mae (but not Freddie Mac). When I left the company I held common shares and vested stock options, all obtained as compensation there. On advice of counsel I sold no shares, and took shares (not cash) upon exercising stock options, until Fannie Mae completed its earnings restatement in December 2006. From January through June 2007 I sold approximately half of my and my family’s holdings of Fannie Mae common stock. I held the rest when the company went into conservatorship, and in August 2009 sold half those shares and used the proceeds to buy Fannie Mae series N preferred stock (picked because it was the last preferred stock issuance the company did when I was CFO). I have done no transactions in either the common or preferred shares of Fannie Mae since then. These shares now have little value. Having this value increase would be a pleasant surprise for me, but it is not why I have involved myself in the mortgage reform debate. Much more is at stake than that.
THE COMING BAILOUT CHARADE
If the past two years are any guide, Fannie and Freddie could release their fourth quarter and full year 2015 earnings as early as the end of this week. Now, therefore, is a good time to address the coming bailout charade that’s been set up by Treasury.
With Fannie and Freddie currently permitted to hold only minimal amounts of capital (which will disappear entirely by the end of next year), a quarterly loss by either company could trigger a capital draw from Treasury. Already, we’ve heard critics of the companies claim that this further “bailout” will be a wake-up call about how truly risky Fannie and Freddie are, and why they must be replaced with a more stable mechanism in order to protect taxpayers.
I doubt that either company will require a draw this quarter, but as more time passes and their capital gets closer to zero, the likelihood increases that there will be a loss at some point. If and when there is, the way commentators react to it will be a litmus test of whether they’re being honest with their audiences.
Fannie and Freddie have simple business models. They have just three main sources of revenue—net interest income from their portfolios, fees on the mortgage-backed securities they guarantee, and relatively small amounts of miscellaneous income—and two main sources of expense, credit losses and administrative costs. Because both companies prudently underwrite the mortgages they purchase or guarantee, and closely hedge the interest rate and option risks on their portfolios (each of their “duration gaps” have been at zero for all but two months in the past two years), they typically have stable and predictable earnings on an economic basis.
But the companies don’t report their earnings on an economic basis; they report them under Generally Accepted Accounting Principles (or GAAP). And since 2001, when a new standard for accounting for derivatives, FAS 133, was introduced, GAAP has required Fannie and Freddie to use historical cost accounting for some parts of their business and fair value (or “mark to market”) accounting for others. Post-FAS 133, the companies’ biggest financial risk—at least in normal times—hasn’t been interest rate risk, option risk or credit risk; it’s been accounting risk: the risk that the different accounting conventions the companies are required to use for assets and liabilities that are well matched economically will produce large and unpredictable amounts of GAAP earnings volatility on a quarterly basis (which, over longer periods, net to zero, or close to zero).
It generally is not possible to hedge against accounting risk (although I do believe both Fannie and Freddie could, and should, look harder at the permissible accounting choices that exist for reducing the large swings in their quarterly GAAP net income). The only sure way to protect against it is to hold more capital. Prior to 2001, when Fannie and Freddie each were permitted to account for all of their business at historical cost, Fannie typically held about $300 million in excess capital, relative to its required minimum. I was CFO at that time, and just three years after FAS 133 took effect we felt it was necessary to boost our excess capital to nearly $3 billion at the end of 2003—predominately as a bulwark for accounting risk. Both companies’ accounting volatility has increased significantly since that time.
On a normalized basis, Fannie should be able to produce annual pre-tax economic earnings of about $12.0 billion per year, and Freddie about $9.0 billion. That’s $3.0 billion and $2.25 billion per quarter, respectively. Soon those quarterly economic earnings will be the companies’ only defense against accounting volatility, and they likely won’t always be enough. If the Net Worth Sweep were annual, only the first quarter of each year would pose significant risk of an accounting-driven loss. But Treasury made the sweep quarterly (probably just for this reason), and that means the companies will have to be lucky every quarter. They very probably will not be.
If and when the companies do post a loss large enough to require a further draw from Treasury, there will be only one correct response: “Because of the Net Worth Sweep, this was inevitable and is of no consequence. Yes, it’s a current cost to taxpayers, but only because taxpayers already have benefited from tens of billions of dollars of the companies’ previous earnings. Moreover, this quarter’s cost will be recouped (along the rest of the companies’ earnings) in a future quarter, when the accounting volatility runs the other way.”
Commentators who call the next Treasury draw a bailout—or claim it reflects some fundamental flaw in the companies’ business model—either will be grossly uninformed or deliberately saying something they know not to be true. And those who say things like, “I understand what’s happening, but it’s still a problem because…” will be dissembling. They surely will know better, because they’ll know about the accounting volatility, and they’ll know about the Net Worth Sweep.