Fannie and Freddie Rally on Ackman's Exit Call — but the Rulebook Doesn't Exist and the Capital Gap Is $205 Billion

Ackman's intervention was a political argument, not a financial plan. He argued that delay lets the GSEs' cash-generating capacity become a political asset for the other party if it regains power in 2028 — his words were \"massive war chest\" — and that swift action avoids the \"waste, fraud, and abuse\" he says the administration claims to be targeting. The same day, the White House published the fact sheet for the executive order \"Decreasing Housing Costs for the American People,\" which directs FHFA to instruct the GSEs to take a set of actions. Two documents moved the stocks and neither one is an exit rule.

The round trip is the point. Pershing Square told shareholders in its second-quarter letter that Fannie Mae and Freddie Mac shares fell 39% and 41% in the first half of 2026 after both more than tripled during 2025 — the run that produced the $12.39 and $11.80 highs. Ackman is now pressing for the policy step that his own portfolio says is missing. The man urging the exit is the man telling the market the exit has no date.

The Q3 2026 filings are the cleanest evidence available, because they quantify the exit in the regulator's own arithmetic. Fannie Mae reported net income of $4.18 billion on net revenues of $7.65 billion and GAAP net worth of $119.6 billion — and still reports a $205 billion shortfall against its adjusted total capital requirement including buffers, and a $173 billion shortfall against its common equity tier 1 requirement. Profitability is not the constraint; qualifying capital is.

The mechanism is more important than the headline gap. Under the Senior Preferred Stock Purchase Agreements, retained earnings do not accrue to the common; they compound into the preferred's liquidation preference. Fannie Mae's claim grew $3.8 billion in the third quarter alone and is contractually scheduled to reach $241.1 billion by Dec 31, 2026. Freddie Mac has already paid Treasury $119.7 billion in cash dividends against $71.6 billion of capital it received — a net cash outflow of roughly $48 billion. Every quarter of waiting transfers another slice of enterprise value from the common to the preferred.

The executive order and the exit call arrived on the same page of the same day, and they point in opposite directions. The order directs FHFA to build a pilot that proactively cancels borrowers' mortgage insurance premiums once they satisfy equity requirements measured against current market values, and to turn the title-waiver pilot into a permanent program for refinance closings. It also notes that in January 2026 the GSEs were directed to purchase $200 billion of mortgage-backed securities. Both instructions move credit onto the GSE balance sheet, and credit on the GSE balance sheet is exactly what the capital framework charges for.

On Sept. 30, 2026, FHFA cut its inspector general's budget to $20 million — 64% below the amount the White House submitted to Congress. The inspector general, James Hodges, said the office would have to cut 70% to 80% of staff and discontinue essentially all criminal investigations into mortgage, bank and other fraud schemes. FHFA's defence was that the office is 18% of the agency's workforce and had requested 16% of the budget, against roughly 2% at other regulators. The agency writing the exit rule stopped funding the office that audits the books the rule will be written about.

That matters for the exit specifically. A conservatorship release has to rest on a capital rule, and a capital rule has to rest on books somebody has verified. With the fraud investigator gutted at the same moment the GSEs are being asked to take on more mortgage risk, the probability that the rulebook arrives late, loosely drafted and legally contested rises. The litigation record already shows how contested this is: on July 25, 2026 the D.C. Circuit affirmed an $812 million judgment that FHFA breached the implied covenant of good faith and fair dealing by adopting the net worth sweep — the very mechanism that drains enterprise earnings. The instrument that collects the common's earnings has already been ruled unlawful, and the replacement has not been written.

Contrast the scorecard with the credit-scoring reform, where the rules were written and published before the market repriced. FHFA said every GSE-approved lender could use either Classic FICO or VantageScore 4.0 for loans sold to the enterprises, and applied a single loan-level price-adjustment grid to both models. FICO fell 26% in one session on Sept. 29 — its worst day ever — and TransUnion extended VantageScore 4.0 at $0.99 per standalone mortgage origination score through 2028. Where the rulebook exists, the market repriced in a day; where it is missing, the market has been repricing on headlines for three weeks.

The tell is management behaviour. FICO chose price over volume — holding its mortgage score price while conceding that the score can be bought elsewhere for less than a dollar — and paid for it with 15% of its workforce and roughly $27 million of severance charges in the September quarter. By contrast, the GSEs' capital requirement has no draft, no comment period and no agency docket. A rulebook you can read is a risk; a rulebook you cannot read is an option, and options are not priced by earnings models.

The premise that this trade has no obvious owner is half right, and the half that is wrong is the actionable half. The GSE common is a call option on a document that does not exist, with a payoff that depends on four things nobody has published: the post-exit capital framework, the disposition of the $238.0 billion and $150.4 billion preferred claims, the treatment of warrants that expire Sept 7, 2028, and the form of the exit itself. That is a position-sizing problem, not a valuation problem. The owners of this policy are the names where published rules are already moving the numbers — the score vendors, the insurers being displaced, and the originators buying the cheaper score.

The one channel with genuinely no owner is the mortgage-backed securities basis and the private-label spread. The agency bid is set administratively — the GSEs were directed in January 2026 to buy $200 billion of mortgage-backed securities — so the marginal buyer is a policy instrument rather than a price. And the boundary at which a re-capitalised GSE starts competing with private-label credit for riskier segments is undefined until the exit framework exists. That is judgment, not measurement: no public series isolates the effect, which is precisely why it has no natural owner. Unowned, unobservable and un-sizeable is a description of an absence, not an investment.

Over the next few quarters, the drivers are dates and dockets, not fundamentals. The preferred claim steps up again on Dec 31, 2026 to $241.1 billion, FHFA still has to publish the terms of the mortgage-insurance cancellation pilot and the permanent title-waiver program, and the midterms decide whether a fast exit is worth the political cost — exactly the argument Ackman is making. Expect the tape to keep moving on headlines and expect the fundamental numbers to keep improving while the gap stays put.

There is a governance overlay that most price action ignores. The FHFA director is also chairman of both enterprise boards, so the same official who would write the exit rule controls the boards that would implement it, and the Administrative Procedure Act question of when a conservatorship directive becomes regulation has not been settled. A rule written by the conservator, adopted by the boards he chairs, and audited by an office running on 30% of its staff is a rule with a long litigation tail — and the D.C. Circuit has already demonstrated willingness to enforce the implied covenant against the agency. That is why the sensible assumption is a late rulebook rather than a clean one, and why the sequencing of headlines matters more than the headlines themselves.

What is established: the exit is gated by a capital rule that has not been written, and the two Q3 2026 filings put the size of that gate at a $205 billion adjusted total capital shortfall for Fannie Mae and $158 billion for Freddie Mac, against $119.6 billion and $81.0 billion of GAAP net worth, with the common payout ratio fixed at 0% and the senior preferred claims compounding to $238.0 billion and $150.4 billion. What is inferred: because the Oct. 22 order pushes credit risk onto the GSEs while the agency writes their exit terms, the rulebook arrives later and more contested than the rally assumes. What is not knowable: the price of the option, because no draft exists to model. A trade with three states — written rule, late rule, no rule — is a position, not a valuation.