Two Gates on Fannie and Freddie: One Has a Date, One Has a Price

If you're new here: I own Fannie Mae and Freddie Mac junior preferred stock and have for over a decade. It is most of my net worth. I trimmed two Freddie series in early September for reasons unrelated to anything below, and I still hold seventeen series. Nothing here is investment advice. I write about what I own.

Between September 2 and September 9, three principals of the Trump administration told you, in their own words, the order in which they expect the next six months to unfold. None of them said the word "Fannie." They didn't have to. Every gate I mapped in August between here and the recap sits downstream of the one they were describing.

This post is about why the release has slipped, what is actually blocking it, what the administration is doing in the meantime, and why the delay says more about the calendar than the outcome.

What they said this week

Howard Lutnick, September 2, CNBC. The Commerce Secretary said he expects rates to "stabilize and start to decline over the coming, let's say, six months," driven by growth and deficit reduction. On Iran: "It's really just an economic choke out, which the administration is now executing more broadly." (CNBC excerpts)

Scott Bessent, September 4. The Treasury Secretary said the Iran conflict "will end, interest rates and the spike in headline inflation will come down," and that crude could fall to "$50, $40 maybe, just because there's so much coming online." He added that rates right now have "the highest correlation they've ever had to the oil price." (World Oil) Asked about Norway's sovereign fund rotating out of Treasuries and into agency mortgage bonds, he called himself "the biggest advocate for that." (Fortune)

Donald Trump, September 8, Truth Social. The President reposted two "Great American Mortgage Corporations" promotional videos for Fannie and Freddie, one of them the AI-voiced ad Bill Pulte first shared in January. National Mortgage News noted the branding "has been associated with Trump administration exploration of a new enterprise stock offering" and that the goal "has been less prominent in 2026 than in 2025." (National Mortgage News, HuffPost) Same day, FHFA filed its rehearing petition in the Sweep case. I covered that here.

Donald Trump, September 9, Joint Base Andrews. Asked about oil at war highs, with Brent above $101 and WTI above $96: "Right after the election, oil prices are going to be tumbling downward." And on Iran: "They're desperate to try and affect the election, so that we can get a nice weak group of people in there." (CNBC, AP via US News)

The backdrop, September 10. Freddie Mac's survey rate printed 6.76%. Daily quotes crossed 7%. The 10-year touched 4.95%, a multi-year high. The Fed held at 3.50 to 3.75% in July with three dissents in favor of a hike, and a hike on September 16 is a live possibility. Gasoline is about $4.10 nationally, up roughly 40% since February 28. The Strategic Petroleum Reserve sits at 286.6 million barrels, 40% of capacity. (Freddie Mac PMMS, Yahoo Finance, The Mortgage Reports, SPR data)

Put those together and the sequence is explicit: choke out Iran, oil falls, rates follow, six months. Trump put a date on the oil leg. Lutnick put a duration on the rate leg. Bessent gave you the mechanism. Nobody put a date on Fannie, and that is the point. Fannie is what comes after.

What was on the desk on February 27

To understand the delay you have to remember where things stood the day before the war.

• September 11, 2025. Lutnick on CNBC: an offering "could well be a this year thing." The aim was "not to sell a lot" but to show taxpayers a "mark-to-market" on what they own. Five to fifteen percent, roughly $30 billion, on a combined valuation around $500 billion. (HousingWire)
• November 2025. Pulte: the companies would "remain in conservatorship" through a sale of up to 5%, and "the president will make a decision either this quarter or early next year." (HousingWire)
• January 8, 2026. Trump orders Fannie and Freddie to buy $200 billion of mortgage-backed securities to push rates down. (CNBC) I wrote about how that fits under the PSPA portfolio caps here.
• February 3. NPR runs "A Fannie/Freddie stock offer could be a boon for rich Trump donors." (NPR) File that headline. It comes back.
• February 6. Pulte on Fox Business: the companies are "definitely ready" for a 2.5 to 5% sale. Conservatorship exit is "an eventual goal." Decision rests with the President. Sullivan & Cromwell is engaged. No banks have been mandated. (National Mortgage News) The same month, House Financial Services Chair French Hill says they are not ready. (Politico Pro)
• February 28. The United States and Israel strike Iran. WTI goes from $67 to $98.71 by March 13. Iran closes the Strait of Hormuz in late March. The administration releases 172 million barrels from the SPR. (Wikipedia timeline, SPR deep dive)
• Early March. Senator Tim Scott, asked about the GSEs: "I assume that he meant after the midterms. Because frankly, it's nearly impossible for us to get there." (Eleanor Mueller) Note the "us." He was talking about Congress, not about what Treasury and FHFA can do by letter agreement.
• March 30. Bill Ackman: "Fannie and Freddie are stupidly cheap. Asymmetry at its best. They could be a 10X." FNMA rises 41% in a day. (Fortune) The market never priced the war as a thesis kill. It priced it as a timing hit, and then it kept bleeding.
• April through July. A ceasefire on April 8, failed talks in Islamabad, a June memorandum, a June 28 agreement to stop attacks, and a collapse on July 8.
• June 2 to August 3. Pulte serves as acting Director of National Intelligence while keeping FHFA and both board chairmanships. (CNBC, CNN) TD Cowen tells clients the spin-off has stalled. (CNN Business) Trump, on Air Force One, says the IPO is still under consideration and there is "no rush." (HousingWire)
• July 20. Jonathan McKernan, Treasury's Under Secretary for Domestic Finance, the desk that would draft any PSPA amendment, leaves after nine months in the seat. His own post says he stepped down. More on him below. (McKernan on X)
• Mid-August. Treasury doubles its long-end buyback operation to at least $4 billion and calls it a "Treasury Twist." (CNBC) Bessent is fighting the long end of the curve with the balance sheet.
• September 1. Hostilities resume. The 10-year hits its highest level in nearly 20 months while Bessent, at the G20 in Asheville, touts the bond market's performance. (CNBC)

So: on February 27 there was a decision on the President's desk. Whether the deal on that desk was a sale inside conservatorship or a sale that required exit is something the public record cannot settle, and I'll come back to that. What the record does settle is that the war arrived at the exact moment the deal needed a decision, and it took away the bandwidth, the market window, and the politics all at once.

Gate one is political, and it has a date

Here is the problem with doing anything visible on Fannie and Freddie before November 3.

Voters cannot separate oil-driven rate moves from GSE-driven rate moves. Nobody can, in real time. So any action the administration takes on the companies before the election gets blamed for whatever mortgage rates do next, and mortgage rates are going up for reasons that have nothing to do with the companies. The attack ad writes itself: he did the deal, rates went to seven and a half, and his hedge fund friends cashed out. NPR wrote the first draft in February.

I think ending the conservatorships makes mortgages cheaper, not dearer, once the capital structure is fixed and the guarantee is explicit. But I should be honest that this is contested. The Furman Center and a good chunk of the Street think exit without an explicit backstop adds somewhere between 20 and 90 basis points. In a rising-rate backdrop, that claim is unfalsifiable. A mainstream economist goes on television and says the IPO raised your rate, and there is no counterfactual to show. That is not just optics. That is real ammunition, and the administration knows it.

The thing about this gate is that it closes on election night regardless of where oil is. After November 3 there is no election for two years. If the House flips, the hearings come anyway, so the marginal political cost of acting drops further, not less. The political gate has a date, and the date is not about Fannie.

Gate two is mechanical, and it has a price

The second gate is about the sale itself, and it does not care about the calendar.

You cannot run a common-stock offering whose public story is "lower mortgage rates" into a 10-year at 4.95%, a survey rate at 6.76%, daily quotes above 7%, and a Fed that might hike next week. In August I wrote that the administration's own stated precondition for an offering was tight mortgage spreads. That precondition is violated right now. Underwriters know it. A seller who says "no rush" and has said the companies are worth many times more does not sell FNMA at $5.56, down roughly 64% from the September 2025 high.

But notice what this gate binds and what it does not. It binds the sale. It does not bind the paper. A PSPA amendment, a resolution of the senior preferred, a tailored capital rule, the end of the consent orders: those are letter agreements and rulemakings. They price at zero. They can be signed with the 10-year anywhere. And historically they get signed at holidays and transitions: December 24, 2009; December 21, 2017; January 14, 2021; January 2, 2025. I laid that pattern out in the gate map.

So the refined read is this. The calendar gates the restructuring. The price gates the sale. Those are different gates, and they open at different times.

What they are doing while both gates are shut

If you cannot touch release before the election, you pull every affordability lever that does not look like release. That is what the last five weeks have been.

1. The $200 billion MBS buy (January 8). Retained-portfolio purchases under the PSPA caps. Roughly 25 basis points of spread compression, 15 to 20 of pass-through to borrowers, per the Street estimates I used in August. It is the only rate lever the administration owns outright.
2. Build-to-order pressure (August 7). Pulte threatens homebuilders who retreat from first-time buyers with GSE pricing curbs. (National Mortgage News)
3. LLPA review (August 21). Pulte: "We are nearing the end of our review of certain LLPAs. Decisions coming soon." Fannie board member Barry Habib has been running the review since last fall. Analysts expect targeted tweaks, not a broad cut, because a broad cut hits earnings and earnings are the capital rebuild. (National Mortgage News) As of the September 9 LLPA matrix, the decision has not landed. (Fannie Mae LLPA Matrix)
4. VantageScore for every lender (September 3). Pulte orders Fannie and Freddie to approve all lenders for VantageScore 4.0 immediately. FICO falls hard. This costs the Enterprises nothing. (Seeking Alpha, Fannie Lender Letter LL-2026-06)

Add Treasury's Twist on the long end and you have five moves in five weeks, all aimed at the same number, none of them a letter agreement.

Here is what matters about all five: they are spread tools. Oil sets the level. The $200 billion buy did exactly what it was supposed to do to the mortgage basis, and mortgages still crossed 7%, because a 20 basis point spread win gets swamped by a 50 basis point move in the 10-year. That is the cleanest evidence in this whole story that the oil gate is real and that the administration knows it. It is why the principals spent this week talking about crude instead of Fannie. They have run out of things they can do to the spread. The level is on the other side of Iran.

There is one lever they have not pulled, and it is the interesting one. The capital rule. ERCF is a 2020 regulation, and the housing executive order already tasks FHFA with tailoring it. A lower requirement makes LLPA cuts affordable, because the same pricing earns a higher return on less required capital. It also shrinks the shortfall to exit. It is the only lever that serves both gates at once. It is also the one that reads as "lowered the safety standard to help the hedge funds," which is why I think it lands with the paper after the election rather than with the spread tools before it. The executive order's 120-day report on this was due around July 11. I have not found a public trace of it.

Why the original design failed

I want to be careful here, because this is the part I am least able to source and most convinced of.

The original pitch, in Lutnick's own words, was a mark-to-market: sell a small slice inside conservatorship to show the taxpayer what the asset is worth. I do not think anyone was actually asking institutions to buy common stock still sitting behind roughly $190 billion of senior preferred face and $373 billion of liquidation preference. Any real offering would have restructured the balance sheet first, with the senior preferred resolved one way or another, so that the thing being sold had a bottom. That part is paper, and paper can be signed.

The harder problem is the word "inside." Even a cleaned-up balance sheet would have been sold by a conservator, and for seventeen years the conservatorship has been the vehicle through which shareholder rights under HERA were set aside: the Sweep, the suspended dividends, the delisting, the exit that never came. An institution buying into that structure is not buying a company. It is buying a promise from the same agency that wrote the Sweep. And that agency's own rehearing petition, filed Monday, tells the D.C. Circuit it must preserve conservatorship authority because the "next unforeseen exercise" of that authority "may not be another Sweep." That is the regulator telling a court, in writing, that it wants to keep the power to do it again. You can fix the capital table with a signature. You cannot fix that sentence while the conservatorship is still in place. Mark-to-market is not a pitch you can make to someone who cannot see what the mark is against, and inside conservatorship the mark is against FHFA's discretion.

The record is consistent with the market saying no. No bank was ever mandated. French Hill said not ready. Congress pushed back through the spring. (National Mortgage News) Michael Burry called the IPO "a 2027 proposition at best" after "the lukewarm reception on Wall Street." (Seeking Alpha) And Ackman, the largest outside holder, spent all of 2025 publicly warning against a rushed offering and proposing the opposite sequence. (Scotsman Guide, December 31, 2025, Ackman's post)

Ackman's sequence, from the Pershing Square deck of January 16, 2025, is worth stating precisely because his timing predates the war, predates Tim Scott, and gets to the same window from the balance sheet instead of the calendar. Treasury acknowledges the senior preferred as repaid. Treasury exercises its warrants. A 2.5% capital standard replaces the current framework. Fannie re-IPOs by year-end 2026 raising about $5 billion; Freddie follows by year-end 2027 raising about $15 billion. Treasury sells down over five years after each. The whole thing depends on retained earnings closing most of the gap first, which is why he put it in late 2026 and not 2025.

I do not share his view on the senior preferred. I do not expect Treasury to write it down or forgive it. My base case, laid out in It's All One Move, is a conversion at face value into common, which is what CBO has scored. That is worse for the common than Ackman's plan and irrelevant to the junior preferred, which sit above the common either way. But notice what conversion does to the capital math: it reclassifies roughly $190 billion from senior preferred into common equity, and combined CET1 goes from negative $46 billion at Q2 to comfortably positive without a dollar of new money. That is why I think the offering is the last step and not the first. You do not need a mega-IPO to fix the capital. You need a signature.

Both Ackman's version and mine agree on the thing that matters for this post: the senior preferred gets resolved before any equity gets sold. Sell-then-exit became fix-then-exit-then-sell. That is a much heavier piece of paperwork, and it is Treasury Domestic Finance paperwork, and the war took the bandwidth for it at exactly the moment it needed to be done.

The Treasury desk

The friction between Treasury and FHFA over how to do this is not gossip. It has been public record for a year.

On September 3, 2025, at the Executive Branch club in Washington, Bessent confronted Pulte over what Pulte had said about him to the President and threatened to punch him. Politico broke it and every outlet picked it up. (KATV, citing Politico) Barron's framed the underlying dispute the following week: Pulte wanted a near-term sale of about 5%, while Treasury was running a longer, comprehensive reform track through a series of stakeholder meetings on what to change about the conservatorships themselves. (Barron's) Bessent's on-air response, when asked, was that Treasury secretaries "dating back to Alexander Hamilton have a history of dueling."

So the two designs I described above were not abstractions. One lived at FHFA and one lived at Treasury, and the men running them were not getting along.

The Treasury desk that owns the comprehensive track is Domestic Finance. Jonathan McKernan took it in October 2025. His resume matters here. He was financial policy adviser to Senator Bob Corker from 2017, when Corker's housing finance bills would have wound the companies down and replaced them. He was senior policy counsel at FHFA from 2019. He sat on the FDIC board from 2023 to 2025. (Wikipedia) He left Treasury after nine months. His own words were that he stepped down and was grateful for the opportunity. I have heard other accounts of why he left. I cannot verify them, so I will not repeat them.

What I can say is this. The Treasury official most associated with the wind-down school is no longer on the desk that would write the paper. I cannot find a confirmed successor. And Bessent's public posture since July is the one you read at the top of this post: rates follow oil, the conflict ends, and if the world's largest sovereign fund wants agency mortgage paper, he is "the biggest advocate for that."

Whatever the reason for the departure, the direction of travel at Treasury is toward the paper, not away from it. That is a direction signal, not a timing one. Losing the desk's occupant in July, in the middle of a war, is one more reason the paperwork did not move this summer.

The honest paragraph

Was there a deal ready to go on February 27 that required taking the companies out of conservatorship? I do not know, and I do not think anyone outside the building does. Pulte's public framing on February 6 was still "eventual goal." Nobody in the administration said "exit first" on the record before the war. If the design changed in January because the market pushed back, that change never surfaced publicly. If the design never changed and the war simply froze an in-conservatorship sale, then the market's refusal is still waiting for them when they come back.

Here is why I do not need to resolve it. Both readings land in the same place. If exit was already the plan, the post-election paper is the deal resuming. If exit was not the plan, the market forces the same paper as the precondition to any sale. Either way the senior preferred is resolved before equity is sold. Either way the political gate has a date and the sale gate has a price. The conclusion is robust to the thing I am unsure about.

Where the clocks converge

Now stack the dates.

• Trump: oil falls "right after the election."
• Lutnick: rates decline over "let's say, six months" from September 2, which is roughly March 2027.
• Ackman, from the balance sheet in January 2025: Fannie by year-end 2026.
• The letter-agreement pattern: December 24, December 21, January 14, January 2.
• The Sweep judgment: FHFA's rehearing petition killed the October 22 finality date; denial in October or November restarts a 90-day cert clock and puts finality somewhere between January and March 2027. (my rehearing post, payout timeline)

Four independent clocks, none of them set by the same person, all landing in the window from Thanksgiving to early spring. The paper in the holiday window, the sale when the oil forecast has either come true or not.

The bear case, stated fairly

This administration's forecasting record on this war is poor. Ackman called it "one of the most one-sided wars in history" on March 30. Two ceasefires have collapsed since. Hostilities resumed nine days ago. "Right after the election" is what you say when you cannot fix it before the election. If oil stays above $90 into 2027, Lutnick's six months becomes twelve, the sale gate stays shut, and the paper gets harder to sign because the political case for it weakens without a rate story to attach it to.

The rehearing petition is not the language of an agency packing up. TD Cowen's "stalled" call from June has aged well so far. Burry's "2027 at best" was right on the offering. The goal is, in National Mortgage News's phrase, "less prominent in 2026 than in 2025," and reposting two ads does not change that. And if the Fed hikes on September 16, the front end moves against every spread tool the administration has.

All of that is true, and all of that is about timing. None of it changes who owns the companies, what the senior preferred is worth at face, or whether a President who reposts Fannie Mae ads on the same day his own agency files a rehearing petition intends to leave this alone. He has said he will do it. He has said there is no rush. Both of those are seller's sentences.

What I am watching

1. September 16, FOMC. A hike moves the front end against the spread tools.
2. Brent. The oil gate is the only gate with a price you can watch every day.
3. The LLPA decision. "Coming soon" since August 21. Targeted or broad tells you how much of the capital rebuild they are willing to spend before the election.
4. The rehearing order. A one-line denial versus an order directing a response. Then the new cert deadline.
5. Q3 10-Qs, late October. The "further action" language on conservatorship. Any change in wording from "remain in conservatorship" toward "exit" is the tell that the design changed.
6. The week after November 3. If the paper is going to move, the pattern says look at the holidays.

Prices as I write, September 11: FNMAS $10.13, about 40% of par. FMCKJ $9.14. FNMFN $15.54. FNMA $5.56. FMCC $5.00.

---

Disclosure: I am long Fannie Mae and Freddie Mac junior preferred stock across seventeen series. I trimmed two Freddie series in early September for reasons unrelated to the analysis above. I am not long the common. Nothing here is investment advice. I hold what I write about, and I have been wrong about timing before, including Monday's base case. More at /fanniegate.