Song Xuetao: Walsh’s “Correct View of Political Achievements”

In recent years, the Federal Reserve has grown accustomed to being data-dependent, and the market has also become used to acting according to the Fed’s (expected) feedback. This is the root reason for the frequent episodes of “bad news is good news.” However, Walsh’s first press conference unfolded with an entirely new approach: the statement was shortened, forward guidance was abandoned, he didn’t submit dot plot forecasts himself, promoted five major reform directions in the form of task forces, and made it clear that the financial markets should better reflect real economic data, rather than constantly responding to Fed statements.

The Fed’s job is to maintain its dual mandate of full employment and price stability; the market’s job is to price in growth, supply shocks, AI factors, tail risks, and policy uncertainty. The new Walsh Fed hands the answers back to the data and leaves volatility to the market.

Walsh has demonstrated a unique and explicit new understanding of performance: What are the facts? Is the policy effective? Is the framework applicable? For the market, this means less certainty in the short term; but for the Fed, this is a more appropriate starting point in the context of the new era.

First, seeking truth from facts: maintain a neutral stance on inflation and interest rates, without taking sides.

He takes a fact-based approach to both US inflation dynamics and the restrictiveness of interest rates.

US inflation has exceeded the 2% target for five consecutive years; in the SEP released at this meeting, this year’s core PCE was further revised up to 3.3%. But Walsh did not simply attribute the problem to “overheated demand.” Instead, he repeatedly talked about supply shocks, energy prices, and second-round effects. He believes the Fed cannot control single prices directly: oil, eggs, milk are not things monetary policy can directly influence. The real risk is these price shocks spreading and becoming broader-based inflation.

When asked by reporters whether interest rates are currently restrictive, his response was straightforward: It’s uneven—one can see restrictiveness in the housing market, but not as obviously in financial markets, capital expenditure, and AI investments—overly loose financial conditions can co-exist with tight real sector conditions.

Walsh does not take a “black-or-white” stance: One side says, “Inflation is all oil prices, we should look through it and keep cutting rates”; the other insists, “Inflation is still high, the economy is strong, we need to hike rates immediately.” His position is: oil shocks can explain part of it, but shouldn’t be a reason to allow inflation to spread; rates are restrictive, but the effects are uneven, so you can’t end the discussion with “Already tight enough”—these are Walsh’s embodiment of seeking truth from facts.

Second, refrain from speculative comments: keep up with the times, admit uncertainties, focus more on action than words.

Walsh acknowledges there are many “known unknowns” in economic data. He said businesses are already making decisions using real-time data, while official data still “echoes history.” It’s not to disparage official statistics, but to make the Fed’s data system (and America’s overall official statistics: BLS, BEA, etc.) catch up with the economy’s structure in 2026.

Because of this, Walsh explicitly stated that forward guidance is no longer suitable under current conditions and must be abandoned. He said that committee members use pencils with big erasers when submitting forecasts, and can revise them anytime—which was becoming increasingly apparent in the later Powell era, since the SEP never really helped policy that much. On most questions from reporters fishing for forward guidance at the press conference, he hardly took the bait, instead quoting the statement directly several times.

More importantly, Walsh didn’t rush to soothe market volatility either. A reporter asked whether the rapid rise in two-year Treasury yields signaled markets saw the need for more tightening; he replied directly that he would not comment on market moves over the past 30 or 60 minutes (not even swings within a week matter much).

Short-term market moves are not Walsh’s KPI; what he wants to give the market is a new Fed: one that asks tougher questions, rather than providing more pleasant-sounding answers; light on forecasts, heavy on response.

Third, in-depth research: reform traditional structure, establish five major task forces, plan before acting.

The task force model launched by Walsh was the focus of this press conference.

Among the five task forces: the communications group studies how the Fed can convey information more effectively; the balance sheet group studies ample reserves and the path for quantitative tightening; the data group looks at how to reform legacy data collection that is too slow and outdated; the productivity and employment group researches supply-demand structures in the AI era; the inflation framework group studies inflation drivers and policy responsibilities.

While the formation of these task forces is not about outsourcing decision-making, advisory roles will be broadened to include more market participants, which changes the market’s place: from reacting to policy to becoming part of policy formulation (similar to the New York Fed’s Advisory Group).

Notably, setting up these task forces reflects Walsh’s “long-termism,” aimed at better aligning the Fed’s monetary policy goals. Because of this, until the task forces are fully implemented and some preliminary conclusions or proposals emerge, monetary policy is unlikely to change—especially rate hikes and quantitative tightening. Walsh needs to fully articulate things under the new framework, and the current task force system is still under construction.

Fourth, maintain technological neutrality: neither overhype nor underestimate AI; if productivity lifts, rate cuts are possible, if not, keep watching.

Compared to previous full embrace of AI, Walsh, as Fed chair, was very restrained in his remarks on AI.

He acknowledged that AI might be the most important general-purpose technology in decades, and that America would benefit in the long run. But when asked by a reporter whether AI is currently boosting demand or supply, he thought that from a macro perspective, “demand is easier to count”: the construction of data centers, CAPEX expenditure, etc. are already in GDP statistics; “supply can only be inferred”: exactly when and to what extent productivity gains materialize is uncertain. In the short term, it’s a “race between supply and demand” for the effect on inflation.

Although Walsh did not use AI as an excuse to push for monetary easing—being more cautious than the market expected, which is another manifestation of “seeking truth from facts”—he nonetheless showed a clear preference: The last time there was a mention of productivity relevance in the monetary policy statement was twenty years ago, so this reflects a certain inclination towards rate cuts.

Overall, Walsh’s correct approach to performance is to clarify three things: What are the facts? Is the policy effective? Is the framework applicable? On this basis, do not act rashly just for the sake of being seen to act, do not speak carelessly just to appease the market, and do not preemptively cut rates just to embrace AI (nor the other way around). Maintain a wide perspective and a narrow focus in action; let the Fed do its job, and let the market do its job.