The Federal Reserve doesn’t create most of the money in the United States — banks do, every time they make a loan and a deposit is born. So when John R. Puri argues in National Review, in his Let Markets Determine the Money Supply, that the Fed’s interest-rate tools should be replaced with a passive “two-way liquidity valve” to cure its “inflationary bias,” he’s picked the wrong villain. His three-step reform: close the discount window, end interest on reserves, and convert open-market operations into a one-price exchange at market prices. Stripped of every active tool, the Fed would still meet the system’s appetite for cash — but the price level, he argues, would no longer be steered by bureaucrats guessing what the economy needs. It is a tidy plan. It would also change almost nothing about the actual cause of inflation.
Here’s the part Puri skips over. When your bank approves your mortgage, nobody goes hunting through somebody else’s savings account to find the dollars it will lend you. The bank types your loan amount into your checking account as a new liability on its own books. A borrower signs a note; the bank credits the borrower’s account with a deposit that did not exist a moment before. The loan creates the money. The deposit you now hold is money — created the moment the loan was approved. No Fed swap counter changes that.
Puri’s reform governs the wrong layer of the money supply. The Fed’s three tools control base money: the reserves banks hold at the central bank to settle transactions among themselves. The money that actually shows up in paychecks, mortgage payments, and grocery bills is broad money — checking accounts, savings deposits, money market funds — and broad money is not handed out by the Fed. It is created when a bank makes a loan. The Fed’s tools, whatever they look like, don’t dictate how much of this money exists. They set the price at which banks are willing to create more. Closing the discount window or ending interest on reserves changes that price. It doesn’t change how many checking accounts get opened.
Watch a real-estate boom if you want to see this in the wild. Banks compete for borrowers, loosen underwriting standards, and the broad money supply swells as new mortgages create new deposits. Prices rise because the new deposits chase the same number of houses. When the boom busts, the deposits vanish as the loans go bad. That cycle — credit up, prices up, credit down, prices down — is the actual inflation mechanism. The Fed can lean against it by raising the cost of reserves, and thus the cost of lending, or it can sit on its hands. Either way, the engine is bank credit, not central-bank discretion.
That matters, because the diagnosis Puri offers is upside down. He says the Fed has “enormous discretion over the money supply’s growth, which it almost always biases toward inflation.” Look at the record. For most of the twelve years before the pandemic, the Fed undershot its 2 percent inflation target. The inflation of 2021 and 2022 didn’t arrive because some Fed official decided to goose the money supply. It arrived because pandemic supply chains snapped and Congress flooded households with cash — a fiscal event the Fed accommodated rather than started. The bias in the data has been toward undershooting, not overshooting. Puri is fighting a phantom. The bias he points at is real enough as an institutional tendency — but it is a tendency toward easier policy when politicians want easier policy, and it lives in the Fed’s posture, not in its plumbing. The bias is real. The cure Puri offers is not.
Then look at the architecture of that cure. Closing the discount window and ending interest on reserves leaves bank credit creation completely untouched. Banks would still underwrite mortgages at whatever spread they could earn above Treasury yields. The broad money supply would still swell and contract with the lending cycle. And the “two-way valve” itself? Puri wants the Fed’s bond operations converted into a pure exchange at “preexisting fair market prices.” But who decides those prices? The Fed does — by virtue of being the only counterparty with the balance sheet to absorb trillions of dollars of Treasuries at a moment’s notice. Under Puri’s plan, banks that need liquidity must sell to the Fed; banks that want to shrink their reserve balances must buy from the Fed. The Fed doesn’t “initiate” transactions anymore, but it is the only door in the building. The price it transacts at is set by the Treasury market it sits inside. That’s not a smaller central bank. That’s the Fed as market-maker of last resort — a more structural, more permanent role than the one Puri claims to be dismantling. He would govern one corner of the system — the corner that, in normal times, does not drive prices at all.
And the corner he wants to dismantle is the one that prevents the worst outcomes. The lender-of-last-resort function — the one Puri relegates to a “severe financial stress” facility — is the worst place to remove a standing tool. The discount-window spikes of 2008, 2020, and 2023, the moments when every solvent bank needed liquidity at once, were not bureaucratic overreach. They were the only job a central bank can do: lend to solvent institutions against good collateral when private counterparties will not. In March 2023, three mid-sized U.S. banks failed within a single week because uninsured depositors could move their money by smartphone. The discount window — the facility Puri wants shut in normal times — handled more than $150 billion in borrowing in that week. There was no time to convene a “stress facility.” There was a window that worked because it was already open. Puri keeps a back-up for “severe financial stress,” but the boundary between normal stress and severe stress is exactly the discretionary judgment he is trying to eliminate. When the next bank run starts at 6 p.m. on a Tuesday, someone has to decide whether the facility is open. That someone is the Fed. Puri does not propose who replaces it. Under his plan, that window would have to be activated after the fact, and the cost of the delay would be measured in failed institutions and broken payrolls.
Puri is right about one thing, and it deserves to be said. Conservative monetary rules — K-percent, Taylor, nominal GDP targeting — really have failed in the field. They look tidy until the velocity of money shifts, or until the output gap turns out to be unmeasurable, or until a pandemic shows up. Anyone selling a clean formula for the money supply is selling something. But the Taylor rule and nominal-GDP targeting are serious proposals with serious academic backing, and a more rules-based Fed would constrain the institutional drift toward whatever politicians want next year. The answer to bad formulas, though, is not no Fed. The answer is a Fed whose job is more honest, not more withdrawn.
What to build instead is harder and less elegant than swapping out the plumbing — which is how you know it’s real. Anchor the Fed to a real mandate: maximum employment at stable prices, judged by outcomes that include wages and who is actually hired, not just the headline CPI. Publish the Fed’s reaction function in real time, so markets and the public can see what the institution is actually doing rather than what its press conferences claim. Separate the Fed’s balance sheet from fiscal policy with a hard rule against monetizing Treasury issuance, even in stress. Recapitalize the lender-of-last-resort function so the Fed never has to choose between solvency and liquidity in the middle of a crisis. Give fiscal policy room to do what it does — invest in housing, child care, clean energy, the productive capacity Puri correctly notes grows over time. And constrain the actual source of money growth — bank lending — with rules that don’t depend on guessing what the Fed will do next: a hard leverage ratio, honest mortgage standards, a credit registry that lets regulators see where new deposits are being born.
Puri’s plan leaves the credit cycle alone and ties a knot in the Fed. That is the wrong way around. The Fed’s discretion isn’t the disease. It’s a thermometer. The fever is in the lending. The prescription is to write rules for the lenders.