Latest / Monetary Policy Explained with Fexingo: Central Banks, Money Supply, and Interest Rates / How Central Banks Use Quantitative Tightening
Transcript
- Lucas: So the Federal Reserve’s balance sheet peaked at just under nine trillion dollars in mid-2022. By June of 2026 — today — it’s down to about six and a half trillion. That’s quantitative tightening, or QT, in action. Luna: And QT is basically QE in reverse, right? Instead of buying bonds, they just let them roll off. Lucas: Exactly. The Fed stops reinvesting the proceeds from maturing Treasury securities and mortgage-backed securities. The bonds expire, the cash goes back to the Treasury, and the central bank's liabilities — mainly bank reserves — shrink. Luna: But why did they start? Inflation was the obvious trigger — but there's more to it than that. Lucas: Right. The stated goal was to tighten financial conditions without relying solely on rate hikes. By reducing the size of the balance sheet, the Fed pulls liquidity out of the system. But there's a second, less discussed reason: they wanted to rebuild 'headroom' for the next crisis. If you never shrink the balance sheet, you lose the ability to expand it again when you need to. Luna: So it's like reloading the cannon. Lucas: Exactly. The Fed's own research suggests that balance sheet reductions act as a supplementary tightening tool — equivalent to maybe one or two quarter-point rate hikes over the entire runoff period. But the real risk is operational: how low can reserves go before the plumbing breaks? Luna: And we have a pretty vivid example of that breaking. September 2019. Lucas: Right. Overnight repo rates spiked to nearly ten percent because reserve balances had fallen too far — down to about one point five trillion dollars. The Fed had to step in with temporary repo operations to calm things down. That episode is basically the cautionary tale for every subsequent QT cycle. Luna: So what's the magic number this time? How low can reserves go before we get another 2019? Lucas: The Fed doesn't have a precise number, but they've been watching a few indicators. The overnight reverse repo facility — or ON RRP — is one. That facility drained from over two trillion dollars in mid-2023 to essentially zero by early 2025. Once ON RRP hits zero, further QT starts eating directly into reserves. That's when things get sensitive. Luna: And the Fed did slow the pace of QT in the spring of 2025, didn't they? Lucas: Yes. In May 2025, they reduced the cap on Treasury runoff from sixty billion a month to thirty billion. That was a clear signal that they were feeling around for the 'ample reserves' floor. The MBS runoff cap stayed at thirty-five billion, but they also said they'd reinvest any excess MBS prepayments into Treasuries to avoid distorting that market. Luna: Which is basically the taper tantrum lesson — don't let the market think you're on autopilot. Lucas: Right. Communication is half the battle. The Fed has been very explicit: we want to stop when reserves are 'somewhat above' the level we think is ample. Not at it, not below it. That's a softer landing than 2019. Luna: But there's also a political dimension. Some members of Congress have criticized QT for adding to the government's borrowing costs, because the Treasury has to issue more debt to the public rather than having the Fed buy it. Lucas: It's a subtle point. When the Fed reinvests, the Treasury effectively gets cheap financing. When the Fed lets bonds roll off, the Treasury has to find buyers in the open market — at higher yields. That puts upward pressure on long-term rates, which the Treasury pays. So QT is effectively a fiscal tightening as well. Luna: So who's buying all those Treasuries that the Fed isn't buying anymore? Lucas: A mix. Domestic banks, pension funds, money market funds, foreign central banks — including Japan and China, though their buying has been more selective. And the Treasury itself has shifted toward shorter-dated issuance to keep demand strong. But the key point is that QT requires willing buyers. If the market can't absorb the supply, yields spike and QT becomes disruptive. Luna: Which brings us back to the balance sheet size question. Where do they land? Lucas: The consensus among Fed watchers is that the balance sheet will settle somewhere between five and five point five trillion dollars. That's roughly where it was in mid-2020, after the initial COVID expansion but before the full QE program. The precise number depends on currency demand and reserve needs, which grow with the economy. Luna: And if you're a listener who wants to follow the Fed's balance sheet data, what do you look at? Lucas: The weekly H.4.1 release, published every Thursday at 4:30 PM Eastern. It shows total assets, securities holdings, and reserve balances. That's the scorecard. The New York Fed also publishes a quarterly survey of primary dealers that asks about their expectations for the balance sheet path. Luna: So QT is still ongoing, just slower. Do you think they'll finish this year? Lucas: If the current pace continues — thirty billion a month in Treasuries, thirty-five in MBS — the balance sheet could reach around five point eight trillion by year-end. But the Fed has said they want to stop when reserves are ample, not when they hit a specific number. So we may see a gradual taper to zero over the next few quarters, with a final pause sometime in 2027. Luna: And just to be clear — QT isn't the same as selling bonds outright, right? It's passive. Lucas: Right. The Fed is not actively selling. That would be a different tool — sometimes called 'active balance sheet reduction' or outright sales. They did some of that with MBS in 2022, but mostly they're letting maturities run off. The difference matters for market impact: passive runoff is more predictable, less disruptive. Luna: But there is a risk that the Fed undershoots and drains too much, like 2019. Lucas: Absolutely. And that's why they're slowing down. The lesson of 2019 is that reserve scarcity can explode overnight. So the Fed is being cautious, but they're also under pressure from the Treasury to stop, because QT raises borrowing costs. It's a delicate balance. Luna: And that's the kind of nuanced trade-off that really makes central banking fascinating — and why we love digging into this stuff. Lucas: Yeah, and if you find these conversations helpful, a couple of dollars a month genuinely makes a difference. Buy me a coffee dot com slash fexingo — that's buy me a coffee dot com slash fexingo — helps keep the show ad-free and going. Luna: It really does. We don't run ads, so listener support is how this stays sustainable. Even a small contribution goes a long way. Lucas: Alright, back to the plumbing. One underappreciated aspect of QT is its interaction with the debt ceiling. When the Treasury runs down its cash balance at the Fed — the Treasury General Account — it injects reserves into the banking system, effectively offsetting QT. That's been a big factor in 2025 and 2026. Luna: So the debt ceiling drama isn't just fiscal theater — it has direct monetary consequences. Lucas: Exactly. When the debt ceiling is hit and the Treasury can't issue new debt, it draws down its cash balance. That adds reserves, loosening conditions even as the Fed tries to tighten. So the actual pace of tightening depends on fiscal politics. Luna: Which makes it even harder to calibrate. Lucas: Right. The Fed has to guess how long the Treasury will be constrained. And that's a political variable, not an economic one. Some analysts argue that QT should be paused during debt ceiling standoffs to avoid muddying the signal. Luna: Did the Fed ever do that? Lucas: No, they haven't. In 2023, during the debt ceiling standoff, they kept QT running even as the Treasury drew down its account. The result was a gradual decline in reserves, but no crisis. But 2025 was different because ON RRP was already drained, so the Treasury's cash drawdown went straight to reserves, slowing the effective tightening. Luna: So the Fed got a de facto pause without having to announce one. Lucas: Exactly. And that's one of the challenges of communicating QT. The actual pace of tightening depends on factors outside the Fed's control. So the cap is a guideline, not a precise tool. Luna: Let's zoom out. How does QT fit into the broader toolkit? Is it a substitute for rate hikes or a complement? Lucas: The Fed's framework treats balance sheet policy as a complement. Rate hikes are the primary tool — they directly affect short-term borrowing costs. QT works through the term premium and liquidity channels — more indirect, but still tightening. In 2022-2023, the Fed did both: they raised rates aggressively and started QT. That was intentional — they wanted to tighten fast. Luna: But now rates are on hold, and QT is still running. So it's doing the heavy lifting. Lucas: In a sense, yes. The fed funds rate has been at four point five to four point seven five percent since late 2024. QT is the only tightening tool still active. So the marginal adjustment is all balance sheet. Luna: And yet, markets seem fairly calm. The S&P is near all-time highs. Credit spreads are tight. So either QT isn't that tight, or the economy is absorbing it well. Lucas: Both could be true. The economy has been resilient — GDP growth around two percent, unemployment at four point one. And the banking system has more capital than in 2019. So the plumbing is holding up. But the risk is that the last few hundred billion of runoff could be the ones that matter. Luna: Like pulling a tablecloth out from under a dish — it's fine until it's not. Lucas: Exactly. And that's why the Fed is feeling its way. They'd rather stop too early and restart later than overshoot and trigger a liquidity crisis. Luna: So what's the one thing you want listeners to take away about QT? Lucas: That it's not just QE in reverse. The dynamics are different because the starting conditions matter. When QE starts, you're adding reserves to a system that's often in crisis. When QT starts, you're removing reserves from a system that's flush. The challenge is knowing when flush becomes scarce. And that line moves. Luna: And to follow the data, not just the headlines. Lucas: Exactly. Watch the H.4.1, watch the ON RRP, watch the repo rate. That's where the real story lives.