Yep it's Fannie Mae, says it right on the sign. Photographer: Andrew Harrer/Bloomberg
Yep it's Fannie Mae, says it right on the sign. Photographer: Andrew Harrer/Bloomberg

Fairholme Capital Management's theory of Fannie Mae and Freddie Mac appears to be that if you pretend hard enough that they're regular companies, it might come true. This is not the worst theory? As theories go. For instance, Fannie and Freddie have boards of directors. Since those directors agreed to pay the Treasury all of their companies' profits, their job is basically to punch shareholders in the face every quarter on behalf of the government. But in a normal company, the board would be tasked with running the company on behalf of shareholders. You could see why shareholders like Fairholme would prefer that.

So the latest is that Fairholme sent some letters to Fannie's and Freddie's boards, asking them to act more like normal boards of directors and less like wards of the state. Here are the letters, and they're pretty punchy. Most immediately, Fairholme wants the enterprises to stop sending all of their profits to Treasury every quarter; more broadly, it wants the directors to set themselves up as an independent force, taking the side of the shareholders against the government. On the first point, here is, generally, how you make corporate directors pay attention:

It is common sense that no Board should approve cash distributions without independent financial advice as to the effect of such payments on the Company’s safety, soundness, and liquidity. Moreover, corporate laws generally prohibit the payment of dividends in many circumstances, imposing personal liability on Directors for illegal dividends – a liability that, pursuant to the Housing and Economic Recovery Act of 2008, is not assumed by the Conservator.

Fannie is scheduled to pay a $7.2 billion dividend to the Treasury this month, and I am imagining a court finding that dividend illegal and Fannie's 12 directors having to come up with half a billion dollars or so each to pay back to the company.

Homeownership Reconsidered

I am not imagining it too hard, though.1 And neither, it seems, is the board: Fannie's chairman, Philip Laskawy, had a curt response to Fairholme's letter, saying that he is "confident that the Board is doing the job it has been given," which is not the job that Fairholme wants it to do: "FHFA has retained certain authorities for its exclusive determination and control, as provided by federal statute, including all decisions relating to the declaration and payment of dividends to the United States Treasury." So Fairholme's immediate goal of getting the board to stop paying Treasury seems a bit unlikely.

Still you can't fault them for trying. The larger goal of Fairholme's campaign is, first of all, to create a good record for its ongoing litigation over who gets Fannie's and Freddie's profits. But another goal seems to be to shift perceptions within Fannie and Freddie. The boards, made up mostly of people with private-sector experience who are used to the idea of public companies operating as fiduciaries for their shareholders (and the related idea of directors being sued personally if they don't), might have some sympathy for the shareholders' predicament, and so Fairholme might as well try to use that sympathy to divide the government regulators of Fannie and Freddie from the directors they've installed.2 Getting the enterprises' directors to think like directors of normal companies is a plausible first step to turning them back into normal companies, even if the directors can't currently do much.

Of course, the main problem is that Fannie and Freddie are not really normal companies. They're weird quasi-agencies of the U.S. government, guaranteeing mortgage debt with the not-quite-full-faith-and-credit of the government behind them.3 And nobody quite knows what normal-company versions of Fannie and Freddie would look like.4

John Carney has a good column today doing some math on how much Fannie and Freddie depend on the government. Part of this dependency is that Fannie and Freddie have massive securities portfolios funded by massive below-market borrowings, but the core business -- guaranteeing mortgages against default -- is even more weirdly government-dependent:

Fannie's other big source of income is its mortgage-guarantee business. This generated about $12.3 billion of income last year, excluding loan-loss releases, on about $3 trillion of single-family and multifamily mortgage guarantees. This is the business that Fairholme Capital and other fund managers see as most attractive. Again, though, this business is so profitable only because of government support, something the firm doesn't pay for.

Yet the bailout agreement envisioned the government levying a fee for what is effectively reinsurance. While its size was never determined, the fee was waived in the 2012 change to the bailout terms.

Remember that part of the reason Fannie and Freddie don't pay for a government guarantee is that, officially, they don't have a government guarantee, but leave that aside! The point is that normalcy in this business is hard to imagine, because there's not much precedent for it.

Just for fun, I've taken a swing at imagining it; that swing is in the footnotes for those who share my view of fun.5 The numbers are fake, but the important point is that your conclusions will be entirely a product of your assumptions: A highly capitalized Fannie/Freddie paying a big reinsurance fee to the government will not work as a private entity; a thinly capitalized Fannie/Freddie relying solely on its own credit (and no too-big-to-fail guarantee hahahaha) to insure mortgages could easily earn double-digit returns. If markets cooperate, that is; I'm not sure that a non-government-backed guarantee from a private, thinly capitalized Fannie and Freddie would be worth quite as much as the current quasi-governmental version but, you know, let's find out.

Obviously, some people want a Fannie and Freddie that can't work as private entities and that remain highly profitable wards of the Treasury forever. And obviously others, such as Fairholme, want a Fannie and Freddie that do work as private entities and are highly profitable for their existing shareholders rather than for Treasury.6 In a normal world, you'd expect the board of directors to be on the side of the shareholders, and of profitability. Clearly, Fannie and Freddie don't live in that normal world, but it's not impossible that frequent reminders of normalcy will push the directors in that direction.

1 Fairholme doesn't make a particularly strong case for the illegality of the dividends: While they are apparently being paid out of borrowed money, and leave Fannie/Freddie, I think it's safe to say, under-capitalized on a sort of comparison-to-other-regulatory-capital-regimes basis, Fairholme doesn't argue that the dividends leave the enterprises insolvent, and after all the regulator approved/demanded the dividends.

2 An obvious comparison is to AIG, another ward of the state that had a shareholder who tried to get the board to join his lawsuit against the government over the terms of the company's bailout. That didn't work. So far it hasn't worked with Fannie and Freddie either. Here's Freddie's 10-K:

We received a letter dated October 16, 2013 addressed to the Chief Executive Officer, the Board of Directors and the then Acting Director of FHFA, purportedly on behalf of holders of common stock and junior preferred stock of Freddie Mac. We received a similar letter dated January 6, 2014, and two more dated January 7, 2014, each on behalf of a plaintiff in the consolidated lawsuits. The letters demand that Freddie Mac commence legal action against the U.S. government to recover all losses sustained by Freddie Mac as a result of the August 2012 amendment to the Purchase Agreement. The letters also demand that Freddie Mac take action to terminate the August 2012 amendment to the Purchase Agreement. On January 15, 2014, FHFA (as Conservator) informed the purported shareholders named in the October 16, 2013 letter that the Conservator does not intend to authorize Freddie Mac or its directors or officers on behalf of Freddie Mac to take the actions that such shareholders demand.

3 It bears repeating that the prospectuses for Fannie's and Freddie's mortgage bonds say, on their cover pages, in bold, "We alone are responsible for making payments under our guaranty. The certificates and payments of principal and interest on the certificates are not guaranteed by the United States and do not constitute a debt or obligation of the United States or any of its agencies or instrumentalities other than Fannie Mae." It's a strictly implicit, and explicitly disclaimed, government guarantee.

4 Don't say "they'd look like Fannie and Freddie in 2007." Come on.

5 Again, this is strictly for fun, and not with any representation that it maps to any actual or possible state of the world, I did some arithmetic on what a Fannie Mae guarantee business might look like (stripped, as Fairholme sort of seems to want, of its legacy of investing in lots of bonds). Here is a very simple model of a business that (1) insures around $3 trillion of mortgage debt, (2) does it with an (entirely equity-funded) reserve of 2 percent to 10 percent of the notional amount insured, (3) invests that reserve in Treasuries and (4) either does or does not buy reinsurance from Treasury (Fairholme seems to think that this reinsurance would not be necessary and that a purely private-capital Fannie/Freddie could work). Here are some illustrative numbers:

For comparison purposes, new Fannie guarantee fees run 57.4 basis points (see page 5 of the 10-K), and credit losses in 2013 were 14.7 basis points, though they were 48.2 basis points in 2012 and got up to 77.4 basis points in 2010 (page 63). I have no particular idea how much capital the guarantee business should have (though I guess that 77.4 basis point credit loss ratio in 2010 provides some guidance? or Fairholme's proposal to bring "approximately $52 billion of private capital to support credit risk on $1+ trillion of new mortgages" maybe implies they want 5-ish percent capital?), or how much the government reinsurance should cost (FDIC assessments run from 2.5 to 45 basis points, though that's not a real comp), so that's most of what I'm varying here.

6 Though, also, Treasury is effectively a 79.9 percent shareholder, so I guess that's some argument to keep operating on its behalf. In normal companies, the board is a fiduciary for all shareholders even if there's a single controlling shareholder, or such is the theory.

To contact the writer of this article: Matt Levine at mlevine51@bloomberg.net.

To contact the editor responsible for this article: Tobin Harshaw at tharshaw@bloomberg.net.