What's Really Blocking the Fed's Balance Sheet Isn't the Fed

He does not show off his skill, nor claim the credit; he does not do the small labour himself, nor encroach on the work of others.

—— Liu Zongyuan, “The Master Carpenter” (Tang dynasty, c. 9th century; my translation)

Liu Zongyuan was writing about a foreman who never picked up an axe. His whole craft was knowing which jobs belonged to somebody else. That line stayed with me through this episode, because an institution reaching into a drawer that isn’t its own usually didn’t plan to — it reached in once, under pressure, and never pulled back out.

What the episode covers

The 7 September 2026 Macro Musings, with David Beckworth hosting Stephen Miran, who has served both as chair of the Council of Economic Advisers and as a Fed governor. The first half is the balance sheet: his speech “Regulatory Dominance of the Federal Reserve’s Balance Sheet,” and the paper he co-wrote with three Fed staff economists, “A User’s Guide to Reducing the Federal Reserve’s Balance Sheet.” The second half is a new paper with Nouriel Roubini and Peter Ireland titled “A Return to Monetarism?”

He keeps the two documents separate: the paper lays out options without taking a side, and the credit belongs to his co-authors; the speech is his own view, and any blame for it lands on him alone. That split made me more willing to listen, not less.

The main points

The balance sheet is big because of regulation, not because the Fed wants it big. That’s the “regulatory dominance” idea — we talk about fiscal dominance and monetary dominance, this is a third one. Banks hold reserves to satisfy liquidity rules, then hold an extra precautionary buffer because drawing the first pile down draws attention. On top of that, the discount window, intraday credit and standing repo all carry stigma, so banks avoid them. Stack those and reserve demand sits far above what banks need.

Which sets the order of operations: shrink demand first, then supply. Move the demand curve left, and supply can follow without pushing short rates around. Their estimate is one to two trillion dollars of reduction while staying inside an ample-reserves framework. Leave that framework for scarce reserves and you can go considerably further — but not back to the pre-2008 world of $50 billion of reserves against an $800 billion balance sheet. Interest on reserves isn’t going back to zero, and Basel and Dodd-Frank are statutes: the liquidity coverage ratio can be loosened or broadened, not abolished.

His sharpest argument against a large balance sheet is housing. Intervene when there’s a market failure; absent one, let markets allocate. Buying mortgage-backed securities injects credit into one sector by choice, and that choice is political. During the pandemic the Fed kept buying with home prices up 20% year over year. His word for the result was that housing got knocked out of reach for a generation. His analogy: antibiotics are for when you’re sick.

A large balance sheet also bites the Fed itself. Big positions mean loss exposure; post-pandemic the deferred asset ran past a trillion, and remittances to the Treasury swing with it, adding volatility to public borrowing and rates. A central bank can in principle run with negative capital. He doesn’t think it survives contact with practice — what it costs is credibility, and credibility is the thing critics aim at.

The discount window exchange is the part I’d keep. Beckworth gets pushback: you’re a markets person, why would you want banks borrowing from the Fed more often? Miran’s answer is to look at the alternative honestly — skipping the discount window doesn’t buy you a deregulated world, it buys you the same regulatory system implemented less efficiently. Society already decided to regulate banks to avoid deflationary depressions; the job is to do that at the lowest cost. He drops an old economists’ joke along the way: if you never miss a flight, you’re spending too long at the airport. The optimal number of bank failures isn’t zero either.

He inverts the couch-cushion metaphor. Someone asked whether we want banks digging under the cushions for spare change. His reply: piles of cash sitting on top of the couch aren’t health — they mean the money can’t find a use in the real economy, because we’ve regulated lending out of the banks and into private credit. He has nothing against private credit. He objects to borrowing and lending decisions being made to arbitrage the rulebook rather than on fundamentals.

The monetarism half turns on measurement, not doctrine. Simple-sum M2 adds the money you buy a soda with to the money you park for a year, at equal weight. Money market funds arrive, the savings half explodes, and the aggregate stops telling you anything about transactions. Divisia weights the components by how money-like they are — the way a price index weights housing above video games. Weight it properly and velocity stabilises, and the history lines up: after the financial crisis, simple-sum M2 exploded and people predicted hyperinflation, while the weighted series showed deflationary pressure from shadow banking and said policy wasn’t easy enough. In 2021 it flashed clearly that policy was behind the curve. He repeats twice that this is one more tool and a cross-check, not a replacement — and that the data right now are not saying the Fed is behind. The gap worth watching is broad M4 running above M2: if it widens, it may be describing froth in financial markets rather than the real economy.

Going further

”The policy moved. Why didn’t my holdings?”

I’ve mistaken one thing for another here more than once, and this topic is a clean place to practise.

Walk the chain: regulation loosens, banks hold fewer reserves, the Fed shrinks the balance sheet. Did the price of money change? No. What changed is whose ledger the money sits on — from a reserve account at the Fed to a lending relationship in the private sector. But the headline reads “Fed shrinks balance sheet,” which sounds like draining.

The question I ask myself now: does this change the price of money, or its location? Price changes — the policy rate, term premia — press directly on the discount line under every valuation. Location changes are plumbing; they alter friction in some markets and leave future cash flows alone. Keeping the two apart at least stops me explaining my own P&L with a neutral piece of news.

“Everyone says this indicator stopped working. Should I still look at it?”

M2 is the textbook case. From the 1980s money demand grew unstable, forecasting power faded, and eventually three Fed chairs said it carried no useful information. Then the Divisia literature went back and found the problem wasn’t that money stopped mattering — it was adding two different functions together at equal weight. The tool was fine. The ruler was broken.

I use that on my own instruments. Price-to-earnings, for a company shifting from hardware to subscriptions, isn’t measuring the same object before and after the revenue recognition changes. Dividend yield misses half of shareholder return once buybacks take over.

So when I hear that some indicator has stopped working, I start with a dumber question: have its components or its definition changed lately? If they have, try reweighting before discarding. Throwing a ruler away takes a second; building one takes years.

”The experts say it can’t be done. Can it?”

Miran jokes in the speech that he has a personality problem: tell him something can’t be done and he has to find out whether that’s true. He describes how the conversation used to go — “you can’t do that.” “Why?” “You just can’t, short-term markets would go haywire.” And it stopped there.

So he sorted the constraints into two piles. Genuinely binding: the liquidity coverage ratio is in statute and can’t be abolished. Habit: the stigma on the discount window, and conventions about what collateral counts toward which requirement. Once sorted, the room to move was wider than anyone assumed.

Applied to my own investing rules, most of my “can’t” turns out to be habit. “I can’t sell this one” usually means I don’t want to book the loss. “I can’t buy now” usually means I’m waiting for a price that doesn’t exist. The lines actually binding me are two or three: available cash, tolerable drawdown, when I need the money. The rest is negotiable.

Worth a look

• The episode itself (7 September 2026, guest Stephen Miran), split evenly between the balance sheet and monetarism
• “A User’s Guide to Reducing the Federal Reserve’s Balance Sheet” — the summary table alone works as a catalogue of options, each with a rough estimate of the reserve demand it frees up
• “Regulatory Dominance of the Federal Reserve’s Balance Sheet” (speech, November 2025), where the phrase comes from
• “A Return to Monetarism?” with Nouriel Roubini and Peter Ireland
• Divisia aggregates are published monthly by the Center for Financial Stability in New York, from M1 through M4

The one thing to take away

The idea I’m keeping: before you accept a line that blocks you, work out whether it’s law or habit. They look identical from the front — both stop you, both come with someone insisting it can’t be done. The difference is that law you have to route around, while habit only has to be touched once before it moves. What Miran spent a paper doing was picking up the lines in front of the Fed one at a time, and finding most of them were the second kind.

Here’s what I’ve tried, if you want it: take something you’ve said “I can’t” about this month — it doesn’t have to involve money. “I can’t raise this with my manager.” “My mother would never accept it.” “I couldn’t sign up for that course.” Write the sentence down word for word, then add one line underneath: who made that rule? If you can name a person, a clause, a policy in writing, keep the line and find a route around it. If all you can produce is “that’s how it’s done” or “I tried once,” then touch it this week — send the message, make the call, ask whether it’s possible.

Raw Markdown

This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.