Money Stuff: The Fannie and Freddie Stakes Are High

The capital structures of Fannie Mae and Freddie Mac are a little complicated, but here’s roughly the situation [1] :

• Fannie and Freddie have total assets of $7.8 trillion ($4.4 trillion for Fannie, $3.4 trillion for Freddie) and liabilities of $7.6 trillion ($4.3 trillion, $3.3 trillion).
• This leaves them with total net worth (shareholders’ equity, assets minus liabilities) of $160.7 billion ($98.3 billion, $62.4 billion).
• The US government has a $348.4 billion senior preferred claim on that net worth ($216.2 billion, $132.2 billion).
• After the US government, there are regular preferred-stock holders with $33.2 billion of preferred claims ($19.1 billion, $14.1 billion).
• After paying out the senior preferred and the regular preferred, whatever’s left over goes to the common shareholders. The biggest common shareholder is also the US government, which gets 79.9% of that; various regular shareholders — of whom Bill Ackman’s Pershing Square Capital Management is perhaps the best-known — get the other 20.1%.

The essential thing to notice there is that the US government’s $348.4 billion senior preferred claim is quite a bit larger than the total shareholders’ equity of $160.7 billion. If Fannie and Freddie liquidated today and returned all their money to shareholders, the US government would get all of it. If Fannie and Freddie’s shareholders’ equity doubled, the US government would still get all of it. The common stockholders, and the holders of regular preferred stock, are underwater by many tens of billions of dollars.

Fannie and Freddie had total net income of $28.8 billion last year ($17 billion and $11.9 billion [2] ); at that rate, it would take about 12 years to earn enough to pay back the government’s $348.4 billion claim and have anything left over for regular preferred shareholders (and another year and change to pay off those preferreds and have anything left over for the common stock). Or it would if Fannie and Freddie worked like normal companies. But in fact, the way they work is that every time their net worth increases, the government’s senior preferred claim — that $348.4 billion — increases by the same amount. So if they earn $30 billion this year, the government’s claim will increase to $378.4 billion, and the shareholders will be no closer to getting paid than they are now. [3]

I do not make the rules! Those are the rules! A lot of people do not like those rules, and that’s reasonable. In the 2008 financial crisis, Fannie and Freddie ran into trouble, and the US government provided hundreds of billions of dollars of cash and credit lines to bail them out. From first principles, there are two plausible ways to do that:

1. Lend them the money, charge them a high interest rate, and take a big equity kicker (say 79.9% of their stock). Then hope they get back on their feet, pay back the money with interest, and create a ton of value for shareholders, including the government.
2. Just nationalize them entirely, zero the shareholders, zero the preferred shareholders, and have the government take all the upside and all the downside.

Either of those approaches could have been fine, and ex ante, in 2008, they didn’t even look that different. But what the government actually did was first No. 1 — bailout with high interest and 79.9% equity stake — and then, later (in 2012), shifting to No. 2, changing the terms of the deal so that the interest on the bailout went from 10% per year to “all of your net income for the foreseeable future.” There is almost no way for the common and preferred shareholders to get anything; whatever money Fannie and Freddie make goes to the government. But the common and preferred still float around, and their holders have ideas. The ideas are mostly “this was unfair and will eventually change.”

And it does seem like something will change. As we have discussed a few times, the Trump administration is talking more and more about ending the government’s control and ownership of Fannie and Freddie. But I started with the capital structure because it seems to me that there is a very wide range of approaches for re-privatizing them.

One extreme of that range would be for the government to stand on its rights: Fannie and Freddie, right now, owe the government more than twice their net worth, and to end government control they would need to pay that back. The arithmetically simple way to do that would be for Fannie and Freddie to go out to the market and raise new capital to (1) pay back the government and (2) be well capitalized as standalone companies. “Well capitalized,” in these discussions, tends to mean shareholders’ equity of at least 2.5% of assets, which works out to about $194 billion of equity. [4] So if Fannie and Freddie went out and did initial public offerings to raise $542.5 billion of new capital — $325 billion for Fannie and $217.5 billion for Freddie — they would have enough to pay back the government and be well capitalized for the future. 

This is arithmetically simple but practically impossible, because you can’t raise $542.5 billion in an IPO (or two). (Particularly not just to repay the government and leave the companies with less book equity than the amount they raised.) That’s more than the total amount of money raised in all US initial public offerings over the last 10 years. If somehow you did do it, existing shareholders would be massively diluted. Pro forma for half a trillion dollars of new equity, the capital structure would be something like 99.99% new investors, 0.008% the US government (which owns 79.9% of the equity now) and 0.002% Bill Ackman and friends. Fannie’s common stock closed at $10.78 yesterday, Freddie’s at $7.975. Those shares would be — not literally zeroed, but pretty much zeroed, if Fannie and Freddie had to go out and raise new capital to pay back the full amount they owe the government.

The other end of the range would be for the government to say “you know what, never mind, this $348.4 billion number is kind of fake, that claim is unfair, and we are going to write it down to zero.” If you do that, then Fannie and Freddie can just keep their existing shareholders’ equity ($160.7 billion). That is almost enough for them to be well capitalized, though they’d still have to do very large IPOs — perhaps $30 billion total — to get all the way to 2.5%. [5]

If you do that, then the existing shareholders will be in good shape. Ignoring a potential new equity raise, the existing shareholders would own 20.1% of the companies (the US government would own the other 79.9%), with a book value per share of around $14 or $15. [6] The existing common shareholders would get something like $25 billion of book value in the companies, the preferred shareholders would get about $33 billion (the face amount of their preferred), and the US government would have a 79.9% equity stake with something like $100 billion of book value.

Notice that the $100 billion for the government here is less than the $348 billion in the other scenario. The other scenario is not realistic, though, because it relies on (1) making all of the existing shareholders really mad and then (2) doing a half-trillion-dollar IPO. The softer scenario — the one where the government gives up its $348 billion claim — seems more realistic, and would actually get the government $100 billion (or more) worth of stock.

Still the point here is that there is a range, and the government can kind of do whatever it wants within that range. It can stand on its rights! It can demand the $348.4 billion, which would probably prevent the re-privatization of Fannie and Freddie, or it can give up the whole $348.4 billion, or it can do something in between. “Do a $60 billion IPO and give us half the money to cancel that senior preferred claim” seems fine, for instance. “Cancel the senior preferred claim, but give us 95% of the common equity instead of 79.9%,” why not.

One possibility here is for the government to cancel its entire $348.4 billion senior preferred claim for the benefit of existing common and preferred shareholders. That is obviously what those shareholders want! But the government doesn’t have to do that, and it can drive a harder bargain.

Which it might do if … it … wants … money? For instance? Bloomberg’s Scott Carpenter reports:

Signs are emerging that the Trump administration may be less willing to give up control of mortgage giants Fannie Mae and Freddie Mac than investors have bargained for, as policymakers scrounge for ways to close US budget gaps.

In recent social media posts, Donald Trump said he’s exploring the sale of new shares in the two companies, which play a key role in determining how much Americans pay for home loans — but he also made clear the government will keep a strong oversight role.

And in recent interviews, Federal Housing Finance Agency director William Pulte said the administration is considering a public offering without actually exiting conservatorship, the quasi-government ownership imposed on the two companies since a 2008 bailout.

“Maybe there’s a way to take these companies public and use these companies for what they are, which are assets for the American people,” Pulte said Monday in an appearance on Fox Business. …

This might deeply disappoint Wall Street investors such as Pershing Square Capital Management’s Bill Ackman who’ve been counting on a windfall if Fannie and Freddie are set free. Since Trump’s election win, shares of Fannie Mae have been up some 700%, hovering near their highest levels in two decades. The bet is that removing the so-called government-sponsored enterprises or GSEs from conservatorship will unleash a torrent of pent-up earnings power.

I have no idea what this means in practical terms. But as a very general matter, Fannie and Freddie are giant companies that generate a lot of money. Right now, the US government is entitled to essentially 100% of that money. To re-privatize them, it would have to give up some of that money. The expectation, which is not unreasonable, is that the government will give up some of that money to Bill Ackman. But that is not an absolute requirement of re-privatizing them, and if the government is looking to keep as much money as possible, that might not be great for the shareholders.

Everyone who has thought about finance for a while has had at least these two thoughts:

1. “Wouldn’t it be cool if people could sell stock in themselves?” [7]
2. “Wouldn’t it be cool if homeowners could sell part of the equity of their homes?”

“The great late-night dorm-room question of financial capitalism,” I have called the first thought, in part because selling equity in yourself is mostly a way to replace student loans; the second thought comes later (when you buy a house). What is great about these thoughts is that everyone has had them, but only at most, what, 500 people have started companies to address them? We have talked about the first question any number of times, because people keep starting businesses of the form “you can sell stock in yourself.” We have talked about the second question fewer times, [8] but not none. There have been some businesses of the form “you can sell stock in your house”; we talked in 2016 about one called Point, which as far as I can tell still exists.

I suppose that in 2025 everyone who has thought about finance has had a third thought, which is “if I just buy a pot of Bitcoins will that make the otherwise normal thing that I am doing more exciting?” Like, if you run an enterprise software company, that’s fine, whatever, but if you run an enterprise software company that also owns a bunch of Bitcoins, that is a capital markets revolution and will reliably get you a valuation that is many times the value of your pot of Bitcoins. So, whatever you are doing, you might as well buy a pot of Bitcoins too. This is not investment advice, but we have talked a lot about the vogue for Bitcoin treasury companies. 

So the obvious move right now is something like “what if people could sell stock in themselves for Bitcoin,” or “what if people could sell some of their home equity for Bitcoin.” Two birds one stone etc. Here’s a tweet from Joe Consorti, who runs a thing called Horizon:

Strategic leverage for bitcoin accumulation is the play.

The cost of capital to do this at an individual level via a HELOC or reverse mortgage, in addition to your mortgage payment, is way too much for most homeowners.

We've removed monthly payments, removed interest charges, and we're not even long the Bitcoin.

You get the Bitcoin right away, and by the way, there are no term limits.

Essentially, Horizon is a perpetual house-backed Bitcoin call option.

If you're betting that BTC (min. 24% 4yr CAGR) will outrun your home (avg. 4% national CAGR) over the next 5, 10, 20, 50 years, Horizon lets you arb that bet, without making any lifestyle changes.

It's essentially the Bitcoin Treasury Strategy at an individual level.

Near-term, we get every homeowner in America to stack bitcoin using their trapped home equity.

Long-term, we extract the monetary premium on real estate and siphon it into BTC.

That is: You have a house, you have $100,000 of home equity, you sell 20% of the equity to Horizon, you use the $20,000 to buy Bitcoins, and now you have diversified your “trapped home equity” into “stacking Bitcoin.” Here is Horizon’s website (“Transform Dormant Home Equity into Bitcoin.”) As far as I can tell it’s the standard way to structure this idea that you would come up with in your dorm room: They give you money for a fixed percentage of your home value, you make no payments, and you pay them back the same percentage when you sell or refinance. Amusingly Bitcoin has nothing to do with it; deep in the FAQ there’s this:

Do I have to use the funds to buy Bitcoin?

No. While Horizon is designed to make it easy to invest a portion of your home equity in Bitcoin, how you ultimately use the funds is entirely up to you. There are no restrictions — once the funds are disbursed, they're yours to use as you choose.

Sure. “You can sell stock in your house” and “you can buy Bitcoin” are completely separate things, but there are only so many ideas in the world, and combining them in new ways can be good marketing.

I should say! We talked last month about the idea that people should mortgage their retirement portfolios, in part because a lot of people really do have a lot of their net worth concentrated in undiversified home equity. “Sell some of your home equity to diversify into financial assets” is not a crazy pitch, though having a lot of your net worth concentrated in undiversified Bitcoin has its own issues. 

Anyway, if you have started a company that lets people sell stock in themselves to buy Bitcoin, I guess let me know. Or don’t honestly, it’s fine.

Naively, the way a private equity fund works is that a sponsor raises money from limited partners, and then it uses their money to buy some companies, and then it spruces up those companies’ operations and capital structures, and then when they are all spruced up it sells them for cash (to other buyers or to the public), and then it takes the cash and distributes it to the limited partners (minus fees). This whole process (buy, spruce, sell) takes a finite and pretty predictable amount of time: It would be weird for a private equity fund to hold a company for six months, or 15 years. So when the sponsor raises a fund, it can say to investors “we plan to give your money back within 10 years” or whatever, and the investors can plan on that.

And then the big story recently is that private equity funds have had a hard time selling their companies at prices they like, and the limited partners are getting antsy: They were expecting to get their money back already, and they haven’t. The private equity industry has dealt with this problem in various ways, including borrowing money to pay out the limited partners, but particularly including “ continuation funds,” where the sponsor raises a new fund to buy companies from its old fund. This is controversial and sort of miscellaneously embarrassing, but it makes sense. The private equity holding period is longer than expected, so you need new fund structures to deal with it.

Naively, the way a private credit fund works is that a manager raises money from limited partners, and then it uses their money to make some loans (mostly to private equity buyouts), and those loans have fixed contractual terms, and as those loans get paid back, the money goes to the limited partners (minus fees). The whole process takes a finite and very predictable amount of time: If you do a five-year term loan, you pretty much get paid back in five years. So when the manager raises a fund, it can say to investors “we plan to give your money back within 10 years” or whatever, and the investors can really plan on that. The loans have fixed terms.

Or not! The Wall Street Journal reports:

Antares Capital has brought in Ares Management to help cash out investors in two older funds, forming one of several large private-credit continuation funds to hit the market in recent months.

Chicago-based Antares, which mainly extends loans to midsize businesses under private-equity ownership, raised over $1.2 billion for a fund to acquire the holdings from investors that chose to sell their share of more than 100 outstanding loans, according to a joint statement. 

Ares’s secondary funds put up a majority of the capital with Antares also making a commitment to the continuation fund, according to the statement. Investment bank Evercore advised Antares on the process. …

The use of continuation funds in the private-credit market has spiked in the past 18 months as managers look for ways to help fund backers, known as limited partners, monetize their illiquid investments.

The typical life of a private-credit fund has lengthened as private-equity managers extend the maturities of loans their portfolio companies have taken out rather than sell the companies. That has resulted in backers of those credit funds to remain invested for longer than anticipated. 

I have to say, I didn’t realize we were so far into the private-credit cycle that anyone would even need continuation funds. My impression was that the dollar-weighted average age of private credit funds was like six weeks.

Anyway I suppose the point here is that there is a category that is like “companies that are acquired by alternative asset managers’ funds,” whose capital structures consist of (1) equity owned by private equity funds and (2) debt owned by private credit funds. [9] Technically the debt has a maturity and the equity doesn’t, but practically they all have similar, finite, and somewhat flexible maturities: Everything kind of continues until the company gets sold, and if the company doesn’t get sold everyone has to wait longer. 

Roughly once a week I get an email to the effect of “I asked ChatGPT to read your columns and then write about _____ in the style of Money Stuff,” which is obviously a waste of time. What you should do is ask ChatGPT to read 200,000 lines of Cobol code written by Dave at Chemical Bank in 1973, which now underpin the entire US payments system, and which no one at any bank has understood since Dave passed away in 2005, and then have it rewrite the code in the style of Dave, but in Python, and with comments, so everyone else can read it. Morgan Stanley knows:

Morgan Stanley is now aiming artificial intelligence at one of enterprise software’s biggest pain points, and one it said Big Tech hasn’t quite nailed yet: helping rewrite old, outdated code into modern coding languages.

In January, the company rolled out a tool known as DevGen.AI, built in-house on OpenAI’s GPT models. It can translate legacy code from languages like Cobol into plain English specs that developers can then use to rewrite it. 

So far this year it’s reviewed nine million lines of code, saving developers 280,000 hours, said Mike Pizzi, Morgan Stanley’s global head of technology and operations. …

But when it comes to full translation, the technology still has some room to mature, he said. It can technically rewrite code from an old language like Perl in a new one like Python, but it wouldn’t necessarily know how to write it as efficient code that takes advantage of all Python’s capabilities, he said. And that’s one big reason humans are staying in the loop, he said.

I assume that approximately five people graduate from college each year with a degree in computer science (Cobol specialty), and then they immediately get hired at big banks at $1 million a year to sit in Dave’s seat and make sure nothing breaks. But now AI is coming for that very niche job.

JPMorgan Hands Lake, Seen as a Contender to Succeed Dimon, More Responsibility. Trump’s ‘Revenge’ Tax Could See Dollar Dive 5%, Allianz CIO Says. Thoma Bravo’s $34 Billion Fundraising Haul Bucks Private-Equity Slowdown. Private credit could ‘amplify’ next financial crisis, study finds. At 10 AM, Stock Options Soar as Retail Traders Unleash New Bots. Meta to Buy Nuclear Power From Constellation as AI Demand Soars. Apple Challenges EU Order to Increase Compatibility with Rivals’ Products. U.S. Scientists Warn That Trump’s Cuts Will Set Off a Brain Drain. New spying claims emerge in Silicon Valley corporate espionage scandal. Doritos, M&Ms Could Be Forced to Include Warning Labels in Texas. Mongolian prime minister ousted after public anger over son’s luxury lifestyle.

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks!