Warsh Doctrine: Greenspan’s Fed, Recast for the AI Age

June 23, 2026

Kevin Warsh has brought the Federal Reserve back to the 1990s. Alan Greenspan's passing days later makes the comparison unavoidable.

A short Fed statement does not create less interpretation. It creates more interpretation per word.

Days before Alan Greenspan died at 100, Warsh chaired his first Federal Open Market Committee meeting and issued the shortest, sparsest Fed statement in years. Reuters called it a return to "stripped-down, 1990s-style central banking." Markets called it something less nostalgic: a hawkish surprise.

The Warsh Fed held rates steady at 3.50% to 3.75%. But that was the least interesting part of the meeting. The real action was in what the Fed stopped saying. The statement stripped out explicit guidance about future rate moves. Warsh then told reporters that forward guidance was not well suited to the moment. He joked that the Fed's dots were written with "big erasers" and that policymakers did not feel bound by them.

Wall Street understood the message: the Fed is no longer promising to hold your hand.

Stocks fell. The two-year Treasury yield jumped. The dollar rallied. Within days, Bank of America moved to forecast three rate hikes in 2026 - September, October and December - after previously expecting no change. That was not just a forecast revision. It was the market's attempt to reconstruct a reaction function after the central bank deliberately blurred it.

This is the Warsh doctrine in its first working form: fewer promises, fewer verbal guardrails, more faith that markets should price the data themselves. It sounds new because it is a rebellion against the Bernanke-Powell Fed. But it also sounds old because Greenspan used to run the Fed this way.

The question is whether Warsh is restoring useful discipline or reviving a dangerous mystique.

1. Greenspan's long shadow

Greenspan deserves the word "maestro," even if posterity should use it with an asterisk.

He took over two months before the 1987 crash and helped stabilize markets. He presided over the long expansion of the 1990s. He saw, earlier than most, that rising productivity could let the economy run hotter without triggering the inflation models' alarms. That was not a trivial achievement. It was judgment, and judgment is exactly what central banking requires when the data are backward-looking and the economy is changing underneath you.

He also made central banking a market event. Traders parsed his words almost syllable by syllable because, in the Greenspan era, a slight change in phrasing could move the bond market. He spoke in fog because he believed fog had operational value. "Fedspeak" was not an accident. It was a policy instrument.

And yes, there was a price. Greenspan's faith in market self-correction later collided with the housing bubble, shadow banking, and the great financial crisis. His legacy is neither sainthood nor indictment. It is the more interesting thing: extraordinary tactical skill paired with one very large strategic blind spot.

Warsh appears to have studied both sides of that ledger. Whether he has learned the right lesson is the question markets are now trying to price.

2. A similarity: opacity as optionality

Greenspan's Fed was not mute, but it was deliberately noncommittal. The modern Fed tells markets how it sees the economy, what risks it is watching, where inflation is going, where unemployment is going, and where each policymaker thinks rates might go. The Greenspan Fed preferred to retain maximum tactical flexibility.

Consider the Fed's first post-meeting statement in February 1994. Today's investor, raised on dot plots and press conferences, might read it twice and still miss the rate hike. The FOMC said it had decided to "increase slightly the degree of pressure on reserve positions." That was the central bank's way of saying: we just tightened policy.

The statement was only 99 words. It did not include a quarterly SEP. It did not show 19 dots. It did not provide a forecast path. It barely said what had happened.

Warsh's first statement was longer, but not by much. It was short enough to be immediately compared with Greenspan's. It also did something very Greenspan-like: it removed the signposts investors had grown used to seeing. Warsh's view is that if the Fed talks too much, markets stop processing the economy and start processing the Fed.

There is a valid point here. Modern central banking can become a hall of mirrors. Investors do not ask, "What does the inflation data mean?" They ask, "How will the Fed say the inflation data should be interpreted?" Then the Fed watches markets to judge financial conditions, except those markets are partly reflecting the Fed's own prior language. A feedback loop becomes a policy framework.

Warsh wants to break that loop. He wants markets to react to incoming data, not to a Fed-generated script about incoming data. That is the strongest case for the Warsh doctrine.

It is also the most Greenspanian case.

3. A difference: Personal v. institutional "Fog"

Greenspan's opacity was partly a product of the man. He liked elliptical language. He liked obscure data series. He liked leaving doors open. He spent his career perfecting the art of saying something without quite saying it (the "fog").

Warsh's opacity is more programmatic. He is not merely being hard to read. He is trying to redesign the communications regime.

That is a meaningful distinction. Greenspan's Fed gradually became more transparent almost despite itself. In 1994, it started announcing policy changes. In 1999, it began issuing statements even when rates did not change. By 2003 and 2004, the Fed was using phrases like "considerable period," "patient," and "measured pace" - early versions of what later became formal forward guidance.

Warsh is moving in the opposite direction. He is starting from the Bernanke-Powell world of abundant communication and walking it back. That is harder than simply never building the system in the first place.

Markets that have never had a map do not complain when the map is missing. Markets that have spent 15 years trading off the map tend to panic when it is taken away.

That is why the June meeting mattered. The Fed did not raise rates. It merely became less predictable. And that was enough to move stocks, bonds, and currencies.

4. A 90s warning: direction is not enough

The Greenspan era offers a brutal lesson for Warsh: markets can correctly guess the direction of policy and still be badly wrong.

In 1994, the Fed was clearly moving toward tighter policy. The economy was strong. Inflation risks were building. Nobody paying attention could claim the Fed was dovish. But markets repeatedly mispriced the timing, size, and intent of the tightening cycle.

The February 1994 hike was only 25 basis points, but it broke a five-year period without rate increases. The Dow fell 96 points, then its biggest one-day point loss in more than two years. The bond market had been built around the assumption of stable or falling rates. The Fed's move detonated that assumption.

Then came August. The Fed raised rates by 50 basis points and said the action was expected to be sufficient "at least for a time." Markets treated that phrase as a kind of pause signal. That was the danger of primitive forward guidance: say too little, and markets fill in the blank with what they want to hear.

Three months later, the Fed hiked by 75 basis points.

That November 1994 move is the cleanest cautionary tale for Warsh. The market was not asleep. Investors expected a hike. The mismatch was size and reaction function. Greenspan worried that a 50-basis-point move would cause markets to immediately price another 50 in December. So the Fed delivered 75. In one move, it tried to change not just the rate level but the market's expectation of how serious the Fed was.

That is what BofA is doing now in reverse. After Warsh's first meeting, the bank looked at the dots, the stripped-down statement, and the chairman's rhetoric, and concluded the Fed's reaction function was more hawkish than previously thought. Three hikes became a plausible path.

In other words: Warsh said less, so Wall Street inferred more.

5. The productivity bet: AI is Warsh's internet boom

The most interesting similarity between Greenspan and Warsh is not communications. It is productivity.

Greenspan's finest hour may have been the late 1990s, when he resisted a mechanical interpretation of the Phillips curve. Traditional models said low unemployment should create inflation pressure. Greenspan suspected the internet, computing, logistics, and corporate restructuring were lifting productivity and allowing faster growth without the usual inflation penalty.

He was largely right, at least for a while.

Warsh's version of that argument is artificial intelligence. If AI raises output per worker, then the economy can grow faster without pushing up unit labor costs. That could, in theory, let the Fed run easier than old models would suggest. It is the supply-side dove case: more productivity, less inflation, less need for punitive rates.

The problem is timing.

Productivity miracles are usually obvious in retrospect and murky in real time. Firms spend years reorganizing before the gains show up in national accounts. Workers displaced by technology do not instantly become higher-productivity workers. And an AI investment boom can itself be inflationary if it sucks up electricity, chips, data centers, capital, and skilled labor faster than the economy can supply them.

This is where a Krugman-style objection bites: productivity is not magic pixie dust. It changes the economy's speed limit only when it actually arrives. Until then, it is a story investors tell themselves to justify valuations.

Greenspan's productivity judgment was brave because he acted against the models. Warsh's AI optimism will be brave only if the data eventually cooperate. Otherwise it risks becoming a more sophisticated version of "this time is different."

6. A bubble: Revisiting "irrational exuberance"

Greenspan gave us the phrase "irrational exuberance" in 1996. It remains one of the great central-banking phrases because it captured both the insight and the impotence of the Fed.

He saw something frothy in markets. He said so. Markets dipped, then went back to partying for years. By the time the dot-com bubble burst, the Fed had not prevented the excess. It had mostly watched it.

That is the dilemma Warsh inherits. AI valuations are not the dot-com bubble, at least not mechanically. Today's largest AI-linked companies have real earnings, real cash flows, and real infrastructure spending. But the market psychology is familiar: a productivity revolution, a valuation surge, and a central bank tempted to believe that technology has loosened old constraints.

Warsh's critique of the post-crisis Fed is that it became too dominant in markets. QE, forward guidance, and endless press-conference nuance turned investors into Fed addicts. Fine. But if he steps back too far, he may rediscover Greenspan's other mistake: trusting markets to discipline themselves precisely when narrative is overpowering price discovery.

A less chatty Fed is not necessarily a less powerful Fed. Silence can move markets too. Just ask anyone holding two-year Treasurys after Warsh's debut.

7. What markets should expect

The Warsh Fed is likely to produce three changes.

First, more volatility around data releases. If the Fed refuses to pre-interpret inflation and jobs reports, markets will do it themselves. Payrolls, CPI, PCE and ISM will become larger policy events.

Second, more dispersion among bank forecasts. In the Powell era, the Fed's own guidance compressed the range of plausible paths. In the Warsh era, the sell side has more room to build aggressive house views. BofA's three-hike call is an early example.

Third, a higher term premium. Investors who are less sure how the Fed will respond will demand more compensation to hold duration. That does not mean long yields must explode. But the direction of risk is clear: opacity has a price.

Bottom line

Warsh is not simply resurrecting Greenspan. He is trying to selectively revive Greenspan: the discipline, the humility, the productivity intuition, the reluctance to spoon-feed markets - without repeating the complacency about bubbles and financial excess.

Greenspan's great contribution was to remind economists that models are tools, not oracles. Warsh's doctrine will succeed if it restores uncertainty where markets had become lazy. It will fail if it restores opacity where markets need accountability.

The Maestro has left the stage. His music is back in the room.