September 8 Was the Tell, and It Came In the Other Way: FHFA, Fannie, and Freddie Ask the Full D.C. Circuit to Rehear the $812 Million Verdict

If you're new here: I'm Glen Bradford. I'm long Fannie Mae and Freddie Mac junior preferred shares and have written the full Fanniegate thesis for years. On July 24 the D.C. Circuit unanimously affirmed the $812 million judgment against FHFA, Fannie, and Freddie for breaching the implied covenant of good faith when they adopted the Net Worth Sweep">Net Worth Sweep in 2012. Three weeks ago I wrote that September 8 was the tell: a quiet day meant the cheap route to further review was closed, and a rehearing petition would "reset the cert clock entirely" and slide the whole calendar into 2027. On the evening of September 8, the last day of the window, FHFA, Fannie, and Freddie filed a petition for panel rehearing or rehearing en banc. This post is what's in it, what it does to the schedule, and what I think it means. I got the base case wrong. Here's the honest update.

What the petition actually says

I read the whole thing. The PDF is here; the first eighteen pages are the petition; the addendum is the July 24 opinion plus the boilerplate certificates. The structure is tight because the rules force it to be: Rule 40 caps a petition at 3,900 words and requires an en banc petition to open with a statement showing a conflict with Supreme Court precedent, a conflict with another circuit, or a question of exceptional importance. The government claims all three.

The framing sentence. The petition opens by describing the panel's rule as one "with extraordinary consequences for the exercise of federal conservatorship authority: FHFA may act within the authority Congress conferred, choose a course the Supreme Court held it reasonably could take, and still have that choice result in hundreds of millions in state-law contract damages because a jury concludes that shareholders did not anticipate that particular exercise of conservatorship authority." That's the whole petition in one sentence. Everything after it is support.

Argument I.A: Collins leaves no gap. The implied covenant of good faith is a gap-filler. Under Delaware and Virginia law it can't override conduct the contract expressly authorizes. The shareholder contracts incorporate HERA. HERA's "best interests" provision lets the conservator act in the interest of the Agency or the public, even at the shareholders' expense, and the Supreme Court in Collins said HERA "authorized the Agency to choose" the Sweep. So, the argument goes, there was never a gap for the covenant to fill: the contract already told shareholders whose interests FHFA could serve, and Collins already said the Sweep was inside that grant. The panel got around this by saying HERA didn't "specifically authorize" the Sweep itself. The petition calls that a "transaction-specific requirement" that "manufactures a contractual 'gap' whenever a party exercises broadly conferred authority in a previously unspecified way."

Argument I.B: Collins foreclosed "arbitrary or unreasonable." To win an implied-covenant claim, plaintiffs have to show the other side acted arbitrarily or unreasonably. Collins held FHFA "could have reasonably concluded" the Sweep served the public interest and "reasonably viewed" it as better than the alternatives shareholders proposed. The panel called that a "different type of reasonableness," statutory rather than contractual. The petition says the distinction doesn't survive: once the Supreme Court has said the choice was reasonable, "a jury could not deem that decision arbitrary or unreasonable merely because the shareholders had not anticipated it." It also points out that plaintiffs tried the case on the same theory Collins rejected: sustained profitability was coming, write-downs would become write-ups, and paying the dividend in kind was a "foolproof solution."

Argument II: damages "affect" the conservator. HERA says "no court may take any action to restrain or affect the exercise of powers or functions of the Agency as a conservator." The panel, following the D.C. Circuit's 2017 Perry decision, held that a damages award doesn't restrain or affect anything because it neither undoes the Sweep nor prevents another one. The petition says that reads "or affect" out of the statute and conflicts with the Third Circuit's Jacobs v. FHFA (2018), which held the inquiry "turns on the practical effect of the relief, not 'damages versus injunctions,'" and which, as the petition notes, expressly acknowledged its conflict with Perry at the time. Imposing "massive liability because FHFA selected one authorized course over another necessarily 'affects' that exercise."

The procedural hook. The cleverest part of the petition is a case called N.S. v. Dixon from 2025, which the government cites for the proposition that "an intervening Supreme Court decision adopting a 'broader' statutory interpretation than prior circuit precedent effects a 'change in circuit law.'" This is the answer to the panel's framing. The panel asked whether Collins specifically displaced Perry's holding on the implied covenant, and said it didn't because Collins was an APA case that never addressed contract claims. The petition says that's the wrong question. The right question is whether the claim Perry allowed remains viable after the Supreme Court held the underlying conduct was authorized and reasonable. That is a question only the full court can answer, because only the full court can overrule Perry. So the request for en banc review is, in that narrow sense, the correct vehicle.

What the petition concedes by silence

This is the part that matters more than the arguments.

The July 24 opinion had four sections. The petition attacks one of them. It does not challenge the panel's holding that plaintiffs proved harm, or that the $1.6 billion single-day drop in share value on the day the Sweep was announced was a proper measure of it. It does not challenge the holding that shareholders who bought after August 2012 have standing because the claim travels with the shares. It does not challenge the class definition or the plan of allocation. It does not revive the "anticipatory breach" argument. The "unprecedented" reasoning it attacks is attacked only for what it implies about reasonableness, not for what the jury found about expectations. And the opinion the government attached to its own petition records, on page 12, that "the FHFA does not challenge the sufficiency of the evidence with respect to the parties' reasonable expectations." Nothing in the petition walks that back.

So the record as it stands is this. A jury found the government frustrated the reasonable expectations of the shareholders. The panel held that finding was supported. The government's position on rehearing is not that the finding was wrong. It is that the finding is legally irrelevant, because HERA lets the conservator disappoint shareholders' reasonable expectations whenever it decides the public interest requires it, and no court may make it pay for that.

Read that argument again with the recap in mind. The petition, on page 15, says the panel's rule "materially curtails FHFA's ability to respond to future financial crises. The next unforeseen exercise of FHFA's conservatorship authority may not be another Sweep. Based on the panel's rationale, another authorized and reasonable response to new circumstances may again expose the Enterprises to substantial damages simply because shareholders had not foreseen it." That is FHFA telling the court, in September 2026, that it wants unreviewable room to do something to the capital structure that shareholders haven't foreseen. I am not going to pretend that sentence is about 2008. The conservator is defending its toolbox, and the $812 million is the price of admission for the argument.

What it does to the calendar

Here is the mechanical part, with the rules linked so you can check them.

Two things a pending petition does not do. It does not stay the accrual of interest; the 5.01% post-judgment rate runs daily on the full balance whatever the government files. And it does not require a bond, because HERA bars execution against the conservator, which is also why nobody has ever been able to force the issue. The government is not paying to keep the money. It is paying to keep the argument.

How likely is it to work

I'll give you my honest read and the evidence for it, and you can weight it.

Against the petition. The D.C. Circuit's handbook says en banc petitions are "frequently filed but rarely granted," and the rule itself says en banc rehearing "is not favored." The panel was unanimous and ideologically mixed: Ginsburg is a Reagan appointee, Walker a Trump appointee, Childs a Biden appointee. There is no dissent for the full court to rally around. The petition's core argument is the same one the panel heard, considered at length, and rejected; the only new move is the N.S. v. Dixon framing. And the "circuit split" with Jacobs has existed since 2018. The Third Circuit flagged it then, and the Supreme Court hasn't taken it up since.

For the petition. The question is genuinely legal, not factual, and the panel's own opinion contains the seam the petition exploits: it acknowledged FHFA "acted within its authority" and still sustained the award. A judge who thinks Collins should mean more than the panel let it mean has a clean hook. And an $812 million judgment against a federal conservator for an action the Supreme Court unanimously called authorized is the kind of thing a court can call "exceptional importance" with a straight face.

My number. I'd put a grant of en banc review well under one in five, and panel rehearing lower than that. But the odds that matter for the timeline are not the odds the court grants. They are the odds the government keeps going after it loses. On August 18 I wrote that "if they will not make the cheap filing, the expensive one is unlikely." They made the cheap filing. The conditional flips. I now think a cert petition, or at least a Rule 13.5 extension application, is more likely than not after a denial, because a party that spends the last day of the window on 3,899 words about a circuit split is building a record for the Solicitor General, not for Judge Ginsburg. The SG still has to sign for FHFA. The companies can petition through their own counsel without him. Both signed this one.

What it means, and what I got wrong

I said a quiet September 8 was the base case. I gave it the base case because private counsel for a Republican administration's FHFA, run by a Director who has spent the year promoting these companies for a public offering, would have to decide that fighting their own shareholders for another year was worth $812 million plus interest. I underweighted how little that decision costs the people making it. The conservator does not pay the judgment; the companies do. The companies do not pay it either, in any way their controlling shareholder feels, because every dollar of net worth they don't pay out sits in Treasury's liquidation preference anyway. The people who decided to file are defending a legal principle with someone else's money, and the principle, unreviewable conservator discretion to subordinate shareholders, is one they may want intact for reasons that have nothing to do with 2012. I should have seen that. I wrote the sentence about the toolbox and then didn't apply it to the filing decision.

What hasn't changed is the thing the thesis actually rests on. Nothing in this petition disputes that the government frustrated shareholders' reasonable expectations. The jury found it, the panel sustained it, and the government's own filing declines to contest it. The petition's whole ask is that the finding be made to not matter. If the full court denies, the affirmed finding stands one more time, now with the government having tried and failed to make it go away twice. If the court grants, the finding still stands; only the legal consequence is in play. Either way, every brief and every Treasury talking point about how the juniors should be treated in a restructuring still has to carry the sentence "a jury found, and a unanimous appellate panel sustained, that the last time the government reached into this capital structure it violated its duty of good faith." That sentence was worth more to me than the cash on August 18 and it still is.

The price implication is the one I laid out in the reaction-function table in the gate map: legal-timing news bleeds, it doesn't gap. A rehearing petition is the definition of legal-timing news. I'd expect a soft tape and a few "government fights on" headlines, and then the same slow re-rate whenever the denial comes. The catalysts that gap, the letter agreement on EDGAR and the record-date order, are untouched by this filing. One of them just moved a few months to the right.

What to watch, in order

• The D.C. Circuit docket, 25-5113. An order directing the class to respond would mean at least one judge is interested; that is the only early signal. Silence followed by a one-line denial is the modal outcome.
• The denial date. Add 7 days for the mandate, 90 for the cert deadline. That is the new finality date if the government stops.
• Late in the cert window. A Rule 13.5 extension application, or a petition. I now think one of them is more likely than not.
• The Q3 10-Qs, late October. The "further action" language will update, and the judgment accrual will tick up.
• Lamberth's docket. He has the April 2025 notice motion fully briefed. He can act on it before the mandate issues, since his judgment was never stayed, but I expect him to wait.
• EDGAR. A letter agreement amending the PSPA remains the event. This filing does not touch it.

Disclosure

Long the junior preferred, same as I've been for years. I trimmed two Freddie series in early September for reasons unrelated to the case. This filing doesn't change what I own or why.

I hold long positions in Fannie Mae and Freddie Mac junior preferred shares. This post is my personal opinion and is not financial advice. Quotations are from the petition and the July 24 opinion; the procedural rules are linked where cited; the odds and the reading of what the filing implies are mine. Do your own research. The full thesis is at glenbradford.com/fanniegate.