Goldman Sachs interprets the Federal Reserve: It will be difficult for Waller to gain support for “rate cuts” through this

Author: Dong Jing; Source: Zhui Feng Trading Desk

Goldman Sachs believes that the recommendations of the U.S. Federal Reserve’s “five working groups” are not binding on the FOMC. Federal Reserve Chair Kevin Warsh’s aggressive stances across multiple areas—including cutting forward guidance, shrinking the balance sheet, and pushing for easing under the banner of an “AI disinflation” thesis—struggle to gain majority support within the FOMC. The end result will likely be a series of compromise changes that look “major for Warsh but have limited impact on other officials.”

According to a report from Zhui Feng Trading Desk, Federal Reserve Chair Warsh has recently announced the formation of five monetary policy “advancing” working groups, attempting to reshape the Fed across five dimensions: communication mechanisms, the balance sheet, data collection, AI and productivity, and the inflation framework. However, in its July 20 research note, Goldman Sachs offered a cold reality check: if investors bet that Warsh can use the narrative that “AI has a structural disinflationary effect” to push the current “rate cuts” (a dovish monetary policy), they are destined to be disappointed.

Goldman Sachs said that although Warsh, as chair, holds enormous power in communication channels such as press conferences, his radical proposals (such as a major balance-sheet contraction and dovish policy based on AI expectations) are in significant conflict with the institutional inertia of the Federal Open Market Committee (FOMC).

Therefore, the firm believes the Fed’s policy path will not undergo an abrupt, radical turn. It is extremely difficult for the FOMC to agree to adopt easing today based on forecasts of future productivity. On the balance sheet and forward guidance, the final implementation is likely to be a compromise along the lines of “appearing to satisfy Warsh on the surface, with limited real impact” (e.g., stopping publication of the median dot plot, or tweaking the composition of asset purchases). The Treasury will also offset market shocks arising from changes to the Fed’s asset structure by adjusting its debt-issuance strategy.

Goldman Sachs maintains its forecast for the federal funds rate: the full-year target range for 2026 remains 3.50%-3.75%, and for 2027 it gradually declines to 3.00%-3.25%.

Working Group 1: Communication mechanisms—“dot plot” may be sidelined, but it won’t disappear

Warsh’s position: He advocates a large reduction in forward guidance and, to date, has offered little comment even on his own economic assessments.

Working group leaders: economists Peter Fisher (University of Washington), former Brazilian central bank governor Arminio Fraga (Arminio Fraga), and former Bank of England governor Mervyn King (Mervyn King). Their common view is that the central bank should clearly state its reaction function while acknowledging uncertainty in forecasts.

Core dispute: whether to modify the “Summary of Economic Projections” (SEP), especially the interest-rate forecast portion commonly referred to as the “dot plot.”

Goldman Sachs points out that the FOMC discussed communication reforms last year but failed to reach consensus, making it even harder in the near term to push through major changes again. Warsh has hinted that the dot plot could be abolished, and a minority of FOMC members have also expressed reservations about the current approach. But Goldman Sachs believes that fully eliminating the dot plot would be too large a step backward in transparency for most officials.

The most likely compromise: adopt a recommendation from former vice chair Don Kohn—stop publishing the median forecast in the SEP to prevent outsiders from interpreting it as an endorsement of the FOMC. For Warsh, this change carries substantial symbolic significance, but investors can still calculate the median themselves, with limited actual information loss.

In addition, Goldman Sachs proposes two options to enhance transparency of the reaction function: first, link each committee member’s economic forecasts to interest-rate forecasts; second, publish scenario analyses by Federal Reserve staff. The eight-page uncertainty quantification material attached to the SEP is hardly noticed in the market.

Working Group 2: Balance sheet—“ample reserves” framework is hard to move

Warsh’s position: He has long criticized quantitative easing (QE) and the Fed’s large balance sheet, calling for a review of the “ample reserves” mechanism and the structure of asset holdings. But he has also recently admitted, “I’m not naive enough to think we can return to the state we were in when I joined the Fed.”

Working group leaders: Karen Dynan (Karen Dynan), an economics professor at Harvard University; Raghuram Rajan (Raghuram Rajan), a professor at the University of Chicago and former governor of the Reserve Bank of India; and Jeremy Stein (Jeremy Stein), an economics professor at Harvard University and former Federal Reserve governor.

The three have differing views. Stein argues in a Jackson Hole conference paper that a large balance sheet helps financial stability because ample reserves reduce incentives for financial intermediaries to rely on short-term, runnable funding; Rajan warns that balance-sheet expansion has “ratchet effects”—during QE, increased bank deposits lead to changes in business models that are difficult to completely reverse when shrinking the balance sheet.

Goldman Sachs’ assessment: within the FOMC, there is virtually no support for abandoning the ample reserves framework, and the room to compress bank reserve needs via regulatory means—and thereby shrink the balance sheet—is also very limited. The ratchet effect is more often seen as a reason to raise the threshold for QE, rather than as a major issue at present.

The truly unresolved question: what kind of assets should the Fed hold for the long term? Two options are: purchasing Treasuries in proportion to the Treasury’s issuance (respecting the Treasury’s debt management function), or mainly holding short-term Treasury bills (matching asset-liability duration and reducing profit volatility). Goldman Sachs believes that regardless of which option the Fed chooses, the Treasury can adjust its issuance strategy to hedge, resulting in only limited net impact on interest rates.

Working Group 3: Data quality—private data is a supplement, not a substitute

Warsh’s position: He criticizes the shortcomings of “old-style survey methods” and official statistical data that are subject to revisions, and calls for the Fed and statistical agencies to use more private data, especially new ways of measuring inflation.

Working group leaders: Raj Chetty, an economics professor at Harvard University; Kevin Murphy, an economics professor at the University of Chicago; and Doug McMillon, former CEO of Walmart. During the pandemic, Chetty’s “Opportunity Insights” lab was among the first to systematically use private data from credit card processors, payroll service providers, and others to track employment and consumption in real time.

Goldman Sachs’ evaluation: efforts to use private data have been underway for years at the Fed and among statistical agencies, and there is no disagreement on the direction. However, it faces increasingly stringent budget constraints.

The key challenge is that private data often struggles to meet the three core requirements for high-quality economic statistics: representativeness, accurate seasonal adjustment, and continuous availability. Taking the employment data from “Opportunity Insights” as an example, the deviation from nonfarm employment data is already quite significant. More seriously, some firms that began providing data during the pandemic later stopped supplying it, while official statistical series must remain continuous and comparable across decades.

Goldman Sachs judges that private data is more likely to serve as a supplement to official data rather than a replacement, and that raw data still needs to be processed by Fed staff or statistical agencies before it can be used for policy judgment.

Working Group 4: AI and productivity—“future disinflation thesis” is insufficient to support current rate cuts

Warsh’s position: He believes AI will produce a “structural disinflation” effect, and that its impact may not be “on the same scale” as past technological progress.

Working group leaders: Marc Andreessen (Marc Andreessen), co-founder of Andreessen Horowitz; Asha Sharma (Asha Sharma), CEO of Microsoft Xbox; and Charles Jones (Charles Jones), an economics professor at Stanford University (currently on leave from Anthropic).

In a recent NBER working paper, Jones concludes that AI will ultimately significantly boost productivity, but because parts of the production process will still require human involvement—making humans a bottleneck—the full impact will take a substantial amount of time to arrive. This also leaves an adjustment window for the labor market.

Goldman Sachs’ historical research finds that during periods when technological progress accelerates, on average it leads to slightly higher job substitution and unemployment rates and slightly lower inflation—thereby allowing the Fed to modestly cut rates to support the labor market transition.

But Goldman Sachs clearly states that this logic is insufficient to support the current dovish stance for two reasons. First, productivity forecasts have historically been extremely difficult to get right; even at the end of the previous economic cycle, forecasts were still quite pessimistic. Second, multiple FOMC members have emphasized near-term inflation pressures driven by AI demand, which conflicts with Warsh’s downplaying.

Goldman Sachs’ conclusion: most FOMC members will be skeptical of the argument that “future AI productivity gains” can justify supporting today’s easing. Warsh’s citation of a precedent from the Greenspan era—when strong productivity growth did not lead to rate hikes despite strong GDP—may receive some recognition. But using expectations of future productivity to drive today’s rate cuts will be hard to win support.

Working Group 5: Inflation framework—limited return of monetarism; consensus on responding to supply shocks

Working group leaders: Greg Mankiw (Greg Mankiw), an economics professor at Harvard University and former chair of the White House Council of Economic Advisers; Thomas Sargent (Thomas Sargent), an economics professor at New York University and a Nobel laureate in economics; and William White (William White), a senior research fellow at the C.D. Howe Institute (William White), an economic adviser to the Bank for International Settlements.

Goldman Sachs’ take on three major topics:

① Inflation target wording: Both Warsh and Mankiw argue that inflation targets should be treated as “2%” rather than “2.0%,” to avoid excessive self-criticism for small deviations. Goldman Sachs believes this view is uncontroversial in the current environment.

② Money quantity: Warsh argues for renewed focus on money supply indicators, but in congressional testimony he explicitly said, “I’m not a monetarist.” Mankiw also said, “Maybe it’s time to reassess the practice of ignoring money-supply quantities.” In its economic forecasts, Goldman Sachs uses price-based financial conditions indicators rather than quantity indicators such as M2. Fed staff may doubt the usefulness of traditional money supply quantities, but they are open to exploring whether alternative money supply indicators such as the Divisia index could improve inflation forecasting models.

It is worth noting that Warsh has included M2 in the latest Monetary Policy Report, but Fed staff added a mild disclaimer: “In a modern economy, it is difficult to measure the stock of money precisely.”

③ Response to supply shocks: Since 2020, the frequency of supply shocks has risen significantly. Warsh’s stance—ensuring that the initial price shocks “don’t spread and proliferate”—is expected to gain broad agreement among other Fed officials. Goldman Sachs assesses this risk by tracking inflation breadth indicators, but it notes that judging how long supply shocks will last remains the core challenge, and neither economic theory nor the working group can provide a simple answer.