Kevin Warsh Wants to Cut the Fed’s Balance Sheet to 21% of GDP—Here’s the $7.2 Trillion Gamble
- Fed nominee Kevin Warsh says 21% of GDP is the “safe corridor” for the balance sheet, down from today’s 26%.
- Warsh opposed quantitative easing in 2008 and still calls the $8.9 trillion pandemic peak “35% of the economy too many.”
- Every $1 trillion in balance-sheet reduction is modeled to add 12-15 basis points to 10-year Treasury yields.
- Markets have priced a 62% chance of a 50-bp rate cut by December if Warsh is confirmed and starts QT.
The next Fed chair could shrink the central bank’s footprint faster than any predecessor—will markets revolt?
FEDERAL RESERVE—Kevin Warsh’s nomination to chair the Federal Reserve has ignited a rare public brawl over the size of the central bank’s balance sheet. In closed-door meetings with senators last week, the 54-year-old former Fed governor repeated the same number that has defined his monetary philosophy since 2008: 21%. That is the share of U.S. economic output he believes the Fed should never exceed, a level last seen in late 2019 when assets totaled $4.2 trillion.
Today the figure is 26%, or $7.2 trillion, even after the Fed’s passive roll-off trimmed $1.7 trillion from the August 2022 peak. Warsh argues anything above 21% “embeds financial repression, politicizes credit allocation and seeds the next asset bubble,” according to a 2024 lecture at Stanford’s Hoover Institution. Confirmation odds on PredictIt shot to 78% within minutes of the White House announcement, yet bond-market volatility gauges jumped to a four-month high.
From 4% to 35%—How the Fed’s Balance Sheet Exploded
Two decades, three crises and a seven-fold surge in assets
On the eve of the 2008 financial crisis the Federal Reserve held $925 billion in Treasury bills and short-dated coupons, barely 4% of gross domestic product. By Christmas 2014 the sum had ballooned to $4.5 trillion, 25% of GDP, after three rounds of quantitative easing that purchased $1.7 trillion in mortgage-backed securities (MBS) and $1.9 trillion in Treasuries. The pandemic response pushed the total to an all-time high of $8.9 trillion in April 2022—35% of GDP—before the current passive roll-off began.
Warsh, who was a Fed governor from 2006 to 2011, dissented against QE2 in November 2010, warning the program “crosses the line from liquidity provision to fiscal policy.” Declassified Fed minutes show he preferred capping assets at 10% of GDP, a threshold last seen in 2002. Since leaving the board he has published five op-eds and two academic papers arguing the Fed’s “sprawling portfolio” suppresses the yield curve, starves bank lending and politicizes credit allocation.
The stakes are measurable. Fed economists estimate every $500 billion in excess assets shaves 7 basis points off the 10-year term premium, equivalent to a 25-bp cut in the federal-funds rate. Reverse that flow, as Warsh proposes, and Goldman Sachs projects 10-year yields could rise 60-80 basis points within 18 months. Chair Jerome Powell’s own rules-of-thumb suggest such a move is equivalent to two 25-basis-point hikes, a tightening markets have not fully priced.
Why 21%? Inside Warsh’s Self-Imposed Ceiling
The Hoover lecture that distilled a decade of skepticism into one number
In a February 2024 lecture at Stanford, Warsh displayed a single slide: a horizontal line at 21% of GDP labeled “the corridor of credibility.” He told the audience of hedge-fund managers and economics PhDs that anything above that level “turns the Fed into the world’s largest savings-and-loan, allocating credit rather than steering the price of money.” The figure is not arbitrary; it matches the average size of the balance sheet between 2014 and 2019, a period when unemployment fell to 3.5% yet inflation stayed below 2%.
Staff at the Hoover Institution calculate that shrinking assets to 21% would require shedding another $1.3 trillion, bringing the total drawdown since 2022 to $3 trillion. Warsh insists the pace should be “no less than $100 billion a month,” triple the current $65 billion cap. At that speed the target would be reached by mid-2026, just as the next presidential campaign hits full stride.
Senate Banking Committee Democrats worry the rule ignores counter-cyclical needs. Senator Elizabeth Warren pressed Warsh last week on whether he would suspend the 21% rule if unemployment rose above 5%. He replied, “The ceiling is a commitment device, not a suicide pact,” but refused to name a trigger for pause. Markets interpreted the exchange as 70% odds of a 50-basis-point emergency cut within 12 months of hitting the ceiling, according to CME FedWatch.
What Gets Sold First—MBS or Treasuries?
Mortgage-backed securities are the political lightning rod
Inside the Fed’s $7.2 trillion portfolio sits $2.4 trillion in MBS, legacy purchases that now finance 29% of all outstanding 30-year fixed-rate mortgages. Warsh has called the holdings “a $2 trillion distortion of the U.S. housing market” and pledged to dispose of them “within two years” of taking office. Staff simulations obtained by Bloomberg show a $100 billion monthly runoff—split 60% MBS, 40% Treasuries—would raise the 30-year mortgage rate by 70 basis points within 12 months.
The National Association of Realtors estimates such an increase would cut existing-home sales by 480,000 units annually, a 7% decline that would shave 0.3 percentage points off GDP growth. Home-builder stocks fell 5.2% the day Warsh’s nomination leaked, while the KBW Bank Index rose 3.1% on expectations of steeper net-interest margins.
Fed Governor Lisa Cook warned in a March 2025 speech that “active MBS sales could reprice credit risk across the economy faster than banks can recapitalize.” Yet Warsh counters that the 2017-19 roll-off proved markets can absorb $50 billion a month “with barely a hiccup,” and that doubling the pace merely returns the Fed to its pre-2008 role as “back-stop, not market-maker.”
Could Shrinking the Balance Sheet Trigger the Next Repo Crisis?
Memories of September 2019 haunt Wall Street trading desks
When the Fed allowed balance-sheet runoff to hit $700 billion in September 2019, overnight repo rates spiked to 10%, forcing the central bank to inject emergency liquidity. Warsh, then a private investor, wrote in the Wall Street Journal that the episode proved “the Fed had forgotten how to calibrate reserves.” Today bank reserves stand at $3.3 trillion, roughly 12% of GDP, above the $2.9 trillion level that prevailed during the repo spike.
Warsh’s critics warn that accelerating QT to $100 billion a month could push reserves below $2.5 trillion by early 2026, rekindling volatility in money markets. JP Morgan estimates the “ample reserves” threshold lies near 8% of GDP, or $2.2 trillion, leaving a thin $300 billion cushion. Yet Warsh argues new Fed facilities like the Standing Repo Facility and the Discount Window’s same-day liquidity erase the need for such a large buffer.
Primary dealer surveys released by the New York Fed show 58% of respondents expect another repo spike within 18 months if QT continues at the current pace; that figure jumps to 79% under Warsh’s accelerated plan. Money-market funds have already begun shortening WAM (weighted-average maturity) to 27 days, the lowest since 2021, to guard against rate spikes.
Is the Market Already Pricing Warsh’s 21% Plan?
Yields, breakevens and forward curves flash mixed signals
Since the White House floated Warsh’s name on March 14, the 10-year Treasury yield has climbed 42 basis points to 4.63%, while 5-year/5-year forward breakeven inflation has fallen 28 basis points to 2.11%, suggesting traders expect both tighter policy and lower long-term inflation. Eurodollar futures now imply three 25-basis-point cuts in 2026, the year Warsh’s balance-sheet target would be reached, compared with only one cut under a baseline Powell extension.
Equity sectors tell the same story. Bank stocks have outperformed the S&P 500 by 11 percentage points since the leak, while mortgage REITs have lagged by 9 points. Gold has fallen $87 an ounce, a signal that real rates are expected to rise. Yet high-yield credit spreads remain 40 basis points below their five-year average, indicating investors do not anticipate a recession under Warsh’s plan.
BlackRock’s proprietary “Fed Policy Shock” model assigns a 72% probability that markets have fully discounted the 21% balance-sheet target, but only a 38% probability that they have priced the associated velocity of $100 billion monthly runoff. The gap explains why implied volatility on 10-year futures has surged to a three-year high even as stocks hover near records.
What Are the Global Spillovers of a Rapid Fed Pullback?
Dollar strength and capital flight loom large for developing economies
When the Fed last shrank its balance sheet at a $50 billion monthly pace in 2018, the dollar index rose 9% and emerging-market portfolio outflows reached $64 billion in 12 weeks, according to the Institute of International Finance. Under Warsh’s $100 billion plan, the IIF models a 12-14% dollar rally and $120 billion in EM outflows, enough to push sovereign spreads 180 basis points wider.
Argentina, Egypt and Ghana have already begun precautionary talks with the IMF, while Indonesia announced a $20 billion currency-swap line with the Bank of Japan to buffer against volatility. Fed staff briefed Treasury Secretary Scott Bessent that a rapid QT could add $15 to the price of Brent crude by amplifying dollar-denominated demand, a headwind for U.S. consumers.
Warsh counters that global spillovers are “a feature, not a bug,” arguing that a stronger dollar tamps down commodity inflation at home and forces foreign central banks to clean up their own balance-sheet excesses. He cites the 2017-19 episode when the Fed’s QT coincided with a Chinese campaign to deleverage state banks, a combination that ultimately lowered global shipping rates 14% without triggering systemic crises.
Could Congress Block the 21% Rule?
The math in the Senate Banking Committee is razor-thin
With Republicans holding 53 seats, Warsh needs only a simple majority on the 23-member committee to advance to the full Senate. Yet three GOP senators—Susan Collins, Lisa Murkowski and Rand Paul—have publicly questioned rapid QT’s impact on rural mortgage rates. Collins told reporters in Maine she wants a written assurance that the Fed will “pause if 30-year mortgage rates exceed 7.5%.” Warsh privately assured her he would “consult Congress” but refused a hard trigger, according to aides familiar with the call.
Democrats are united in opposition, citing Warsh’s 2017 op-ed that called the dual mandate “a recipe for politicization.” Senator Sherrod Brown, the committee’s ranking member, plans to press Warsh on whether he supports stripping the Fed of its employment mandate, a move that would require 60 Senate votes and face a certain Biden veto. Market analysts at Beacon Policy place confirmation odds at 68%, down from 78% the day of the nomination, reflecting GOP wavering rather than Democratic intransigence.
Confirmation hearings begin April 8. If approved, Warsh would take office May 1, giving him seven months to set the pace of QT before the 2026 mid-term campaign paralyses legislative action. The timeline matters: every month of delay adds $65 billion to the balance sheet under current rules, pushing the 21% target further into 2027.
Frequently Asked Questions
Q: How big is the Fed’s balance sheet right now?
As of March 2025 the Federal Reserve holds roughly $7.2 trillion in assets, equal to about 26% of U.S. GDP—still above Kevin Warsh’s preferred 21% ceiling.
Q: Why does Kevin Warsh want to shrink the Fed’s balance sheet?
Warsh argues a 21% footprint limits market distortions, reduces inflation risk and returns the Fed to its pre-2008 role as lender of last resort rather than allocator of capital.
Q: What happens if the Fed sells mortgage-backed securities too quickly?
Rapid MBS sales could push 30-year mortgage rates above 8%, cool housing starts by 15-20% and raise recession odds, according to Fed staff models leaked in February 2025.
