Federal Reserve Monetary Policy Strategy Review 2025: Uncertainty, Risk, and Central Bank Communication

What the Federal Reserve 2025 Strategy Review Reveals About Uncertainty

The Federal Reserve FEDS 2025-073 paper on accounting for uncertainty and risks in monetary policy represents a foundational document in the FOMC’s ongoing strategy review. Published as background research for the Federal Open Market Committee, this paper systematically examines how the Fed measures, assesses, and communicates the uncertainty that surrounds every monetary policy decision.

The paper’s central insight is both humbling and practical: no single measure of uncertainty is sufficient, the tools available to communicate uncertainty are imperfect, and the academic and policy communities have not yet reached consensus on best practices. As the authors note, practices for communicating risks and uncertainty are “evolving” and differ substantially across the world’s major central banks.

This matters profoundly for financial markets, businesses, and policymakers. The Federal Reserve’s monetary policy decisions affect everything from mortgage rates and corporate investment to global currency flows and emerging market stability. How the Fed communicates the uncertainty surrounding those decisions shapes expectations, influences market behavior, and ultimately affects economic outcomes. For professionals exploring how complex financial research drives decision-making, this review offers essential insights into the analytical frameworks that guide the world’s most influential central bank.

Three Types of Uncertainty Shaping Fed Monetary Policy Decisions

The FEDS paper establishes a rigorous taxonomy of the uncertainty that monetary policymakers face, organizing it into three broad categories that interact in complex ways:

Uncertainty about the state of the economy: Even the most basic economic data is uncertain. GDP figures are revised multiple times after initial release, inflation measurements involve methodological choices that affect results, and critical variables like the output gap (the difference between actual and potential GDP) and the natural rate of unemployment cannot be directly observed — they must be estimated using models that themselves carry uncertainty. The Bureau of Labor Statistics data revisions regularly change the picture of where the economy stands.

Uncertainty about the structure of the economy: Policymakers cannot be certain how the economy actually works. The slope of the Phillips curve — the relationship between unemployment and inflation — has changed dramatically over time, appearing to flatten between the 1970s and the end of the 2010s, with post-pandemic uncertainty about whether it has steepened again. How monetary policy transmits to economic activity, through which channels, and with what lags remains subject to genuine model uncertainty.

Uncertainty about expectations: How businesses and households form expectations about inflation and monetary policy fundamentally affects how the economy responds to policy changes. Measuring these expectations is itself challenging — survey data arrives infrequently, financial market prices embed risk premiums that obscure true expectations, and there exists a “hall of mirrors” effect where policymakers and markets simultaneously try to read each other’s signals.

How the Fed Measures Economic Uncertainty: VIX to Text Analysis

The Federal Reserve monetary policy strategy review 2025 paper catalogs four distinct categories of uncertainty measurement, each with specific advantages and limitations that make them complementary rather than substitutes:

Financial market indicators: The VIX volatility index and options-implied probability distributions provide continuous, high-frequency, forward-looking measures of uncertainty. Their key limitation is that risk premiums and liquidity conditions complicate interpretation — a higher VIX might reflect genuinely increased uncertainty or simply a change in risk appetite among market participants.

Survey-based measures: Monthly and quarterly surveys of professional forecasters, businesses, and consumers directly measure beliefs about future economic outcomes. However, they arrive at low frequency, may suffer from respondent attention issues, and can be slow to update in rapidly changing environments.

Textual analysis (Economic Policy Uncertainty): The Economic Policy Uncertainty (EPU) Index uses automated text analysis of newspaper articles to measure policy-related uncertainty. This approach is timely and can identify specific sources of uncertainty but cannot provide a probabilistic interpretation of potential outcomes.

Statistical models: Econometric models grounded in macroeconomic theory provide structured, quantitative uncertainty assessments. However, they may fail to capture novel risks — by definition, model-based uncertainty measures “will shift only after a shock appears in the data,” making them backward-looking precisely when forward-looking information is most needed.

The paper’s recommendation is clear: policymakers should monitor “a suite of different measures” rather than relying on any single indicator, since each captures different dimensions of the uncertainty landscape.

Why Fan Charts Are Falling Out of Favor at Central Banks

One of the most significant findings in the Federal Reserve monetary policy 2025 review is the declining use of fan charts — the probability-shaded forecast bands that have long been a staple of central bank communication. Multiple central banks have discontinued them, and the paper explains why.

The Bank of Canada stopped publishing fan charts in 2015. Norges Bank followed in 2020, and Sweden’s Riksbank reassessed their use after the COVID-19 pandemic revealed fundamental limitations. The core problem is multifaceted: the public overwhelmingly focuses on the central forecast line and ignores the probabilistic bands that are supposed to convey uncertainty. Research shows that even sophisticated market participants extract primarily the point forecast from fan chart presentations.

The methodological issues are equally concerning. Fan charts typically rely on a rolling 20-year window of historical forecast errors. The current window now includes both the Global Financial Crisis (2007-2009) and the COVID-19 pandemic — two events that dramatically increased forecast error magnitudes. As a result, fan charts today show much wider confidence bands than they did a generation ago, potentially overstating normal-times uncertainty while still potentially understating the risks from genuinely novel events that historical samples cannot capture.

When central banks attempt to adjust fan chart skewness to reflect their risk assessment — making the bands asymmetric to show, say, greater downside risk to growth — interpretation becomes more complicated rather than clearer. The paper concludes that while fan charts remain useful internal analytical tools, their effectiveness as public communication instruments is limited.

How Forecast Errors Have Doubled Since the Great Moderation

The FEDS paper presents striking data on the deterioration of forecast accuracy over recent decades. Four-quarter-ahead forecast errors from 2004-2024 are approximately twice as large as those from 1983-2003 — a period known as the Great Moderation for its relative macroeconomic stability.

The specific magnitudes are significant for anyone relying on economic forecasts for business or investment decisions:

• Real GDP growth: Standard deviation of forecast errors increased from 1.6 percentage points (1983-2003) to 2.9 percentage points (2004-2024)
• CPI inflation: Error standard deviation rose from 1.1 to 1.9 percentage points
• Unemployment rate: Error standard deviation increased from 0.7 to 1.6 percentage points

This doubling matters practically because both the Fed’s Tealbook and Summary of Economic Projections (SEP) use 20-year rolling windows for constructing confidence intervals. A GDP growth forecast of 2.0% that would have carried a 70% confidence interval of roughly 0.4% to 3.6% during the Great Moderation now carries a band of approximately -0.9% to 4.9% — a dramatically wider range that encompasses recession and boom scenarios simultaneously.

The paper emphasizes that this widening is driven primarily by the inclusion of two extreme events in the sample, raising the question of whether these wider bands accurately represent future uncertainty or merely reflect the mechanical inclusion of historically unusual episodes.

Scenario Analysis: The New Frontier in Risk Communication

As fan charts decline, scenario analysis is emerging as the preferred alternative for communicating monetary policy uncertainty. The Federal Reserve monetary policy strategy review documents the rapid adoption of scenario-based approaches across major central banks, though it cautions that “it is too early to evaluate the advantages and disadvantages of this tool.”

Sweden’s Riksbank has been the most aggressive adopter, publishing alternative scenarios in every quarterly monetary policy report since 2023. Critically, the Riksbank is the only central bank that includes expected monetary policy response paths within its scenarios — showing how interest rates would change under each scenario, not just economic variables.

The Bank of Canada took perhaps the most radical approach in 2025, entirely abandoning its baseline forecast during a period of elevated trade policy uncertainty and communicating only through alternative scenarios. This unprecedented step acknowledged that identifying a single most likely outcome was less informative than presenting the range of plausible paths the economy might take.

The Bank of England began including scenarios in its May 2025 monetary policy report following recommendations from Ben Bernanke’s 2024 independent review of its forecasting practices. This review was itself prompted by the Bank’s forecast performance during the post-COVID inflation surge.

The paper identifies key design choices that shape the effectiveness of scenario analysis: which risks to highlight, which economic models to use, whether to include policy response paths, and how to present scenarios without creating the impression that one is more likely than another. These choices have significant implications for how markets and the public interpret the information. Analysts studying how central bank research shapes market expectations will find this evolution particularly important.

Growth-at-Risk Models for Predicting Economic Tail Events

The FEDS paper examines the “outlook-at-risk” approach — an econometric methodology that links financial and economic indicators to the entire probability distribution of future outcomes, not just the central tendency. This approach has gained significant traction since Adrian, Boyarchenko, and Giannone’s influential 2019 work on growth-at-risk.

The key insight is that financial conditions affect not just the expected level of economic activity but the shape of its entire probability distribution. When stock market volatility increases to two standard deviations above its mean, growth-at-risk models show that the GDP growth distribution becomes more skewed toward lower outcomes — the left tail fattens, indicating elevated recession probability — even though the modal (most likely) outcome may change relatively little.

For inflation, elevated volatility produces more dispersed distributions with greater weight in both tails, suggesting that periods of financial stress increase both deflation and inflation risks simultaneously. This finding is particularly relevant for central banks trying to balance their dual mandate during periods of financial turbulence.

The Norges Bank and the European Central Bank already incorporate outlook-at-risk models in their monetary policy reports. However, the FEDS paper notes that the efficacy of these models as communication tools — as opposed to internal analytical instruments — remains uncertain. The complexity of distributional analysis may be difficult to convey to the public in a way that genuinely improves understanding of economic risks.

The Hall of Mirrors: When Markets and the Fed Try to Read Each Other

One of the most intellectually fascinating sections of the Federal Reserve monetary policy 2025 review addresses what the authors call the “hall of mirrors” problem in expectations measurement. This phenomenon occurs when financial market participants and Federal Reserve policymakers simultaneously attempt to infer each other’s expectations, creating potentially self-referential feedback loops.

Consider the mechanism: the Fed observes market-implied probabilities of future interest rate paths to gauge expectations. Market participants, knowing the Fed watches these indicators, adjust their positions based on what they believe the Fed will conclude from observing those very prices. The Fed then observes the adjusted prices, potentially drawing conclusions about expectations that are partly artifacts of its own communication rather than independent assessments.

The paper warns that in such a hall of mirrors, “financial market prices may become less informative, adding further uncertainty about the public’s true expectations.” This is not merely a theoretical concern — the authors document how the Survey of Market Expectations showed significant probability distribution shifts between March and May 2025, with distributions moving toward lower GDP growth and higher inflation in just two months.

The practical implication is that the Fed must be cautious about over-relying on market-based signals, since those signals are partly endogenous to the Fed’s own behavior. Survey data provides a partial corrective, but surveys arrive at lower frequency and may not capture the rapid shifts that market prices can reflect. The challenge is to extract genuine signal from a noisy, self-referential information environment.

How Nine Major Central Banks Communicate Economic Risks

The FEDS paper surveys communication practices across nine major central banks: the Federal Reserve, European Central Bank, Bank of England, Bank of Japan, Bank of Canada, Norges Bank, Reserve Bank of Australia, Reserve Bank of New Zealand, and Sweden’s Riksbank. The comparison reveals substantial variation in approaches and no convergence toward a single best practice.

Policy rate forecasts: Only three central banks publish their own interest rate projections: the Riksbank, Norges Bank, and the Reserve Bank of New Zealand. The Fed publishes the dot plot — individual FOMC members’ rate projections — which is a different construct that shows the range of committee views rather than a single institutional forecast.

Fan charts: Use has declined significantly. Several banks have discontinued them, while others maintain them primarily as supplementary material rather than the centerpiece of forecast communication.

Scenario analysis: Adoption is accelerating, with the Riksbank, Bank of England, and Bank of Canada all implementing some form of published scenario analysis since 2023. However, approaches differ substantially — only the Riksbank includes policy response paths in its scenarios.

Qualitative communication: All nine central banks rely heavily on policy statements, press conferences, minutes, and public speeches as their primary uncertainty communication tools. The paper concludes that these narrative approaches are likely “the most used and most effective tools” for conveying the nuances of the economic outlook and the balance of risks — a finding that may surprise those who expect quantitative tools to dominate. Those following how monetary policy research evolves globally will find this comparative analysis essential.

What Fed Tealbook Scenarios Got Right and Wrong About Crises

The paper provides a revealing assessment of the Fed’s own internal forecasting track record through the Tealbook — the confidential briefing document prepared by Fed staff before each FOMC meeting. The Tealbook has included alternative scenarios alongside its baseline forecast for decades, and the paper evaluates how well these scenarios captured the risks that actually materialized.

The verdict is mixed. The Tealbook’s alternative scenarios encompassed the actual GDP and inflation forecast errors “most of the time” — their coverage was reasonable under normal conditions. However, the scenarios notably failed to sufficiently characterize downside risks before the Global Financial Crisis in 2007-2008. The scenarios presented in the years leading up to the crisis did not adequately capture the possibility of a systemic financial meltdown of the magnitude that actually occurred.

Similarly, and perhaps more understandably, the risks to output and inflation from COVID-19 “were not foreseen in the Tealbooks produced in 2019.” This is not surprising — pandemic risks of that magnitude are inherently difficult to incorporate into standard economic scenario planning — but it does illustrate the fundamental limitation of any forecasting approach grounded primarily in historical experience.

The implication for the FOMC’s current strategy review is clear: while scenario analysis is a valuable tool, it must be supplemented by explicit consideration of risks that fall outside the range of recent historical experience. Models and judgment must work together, particularly during periods when novel risks — whether geopolitical, technological, or environmental — may not be well-represented in existing frameworks.