The Fed’s Inflation Forecasting Problem Is Now a Credibility Problem

The Fed’s Inflation Forecasting Problem Is Now a Credibility Problem

The chart is not just a picture of inflation forecasts. It is a picture of institutional drift. Core PCE inflation, the Federal Reserve’s preferred measure of underlying price pressure, has remained above the 2% target for nearly five years. The April 2026 reading was 3.3% year over year, according to the Bureau of Economic Analysis, and the June 2026 Summary of Economic Projections again showed officials marking their inflation path higher. That level matters, but the more important signal is the pattern. Across projections for 2022, 2023, 2024, 2025, 2026, and even later years, the same basic mistake kept appearing: inflation was expected to fall back toward target faster than it actually did.

Forecasting is always hard. Nobody should expect central bankers to predict every energy shock, fiscal impulse, supply-chain disruption, tariff change, labor-market shift, or behavioral response perfectly. But there is a difference between noisy error and one-sided error. When the forecast misses repeatedly in the same direction, the issue is no longer ordinary uncertainty. It becomes a sign that the model, the reaction function, or the institutional incentive structure is systematically underestimating inflation persistence. In that environment, monetary policy risks being calibrated not to the inflation the economy is delivering, but to the inflation policymakers expect to arrive soon.

That is the central problem. The Fed did not merely make optimistic forecasts. It allowed those forecasts to shape the policy stance for too long. If inflation is expected to fall quickly, then a restrictive stance can be described as sufficient, patience can be defended, and rate cuts can remain visible on the horizon. If inflation does not fall, the institution is left explaining why the destination keeps moving. Over time, the market stops asking whether the forecast is right and starts asking whether the 2% target is a binding objective or an aspiration deferred into the long run.

What the Chart Really Shows

The chart tracks fourth-quarter over fourth-quarter growth rates in core PCE projections across successive FOMC projection vintages. Each colored line is a year. The visual message is simple: the dots moved upward. The projected path for 2022 rose sharply as inflation proved more severe than initially expected. The 2023 path also moved higher before eventually easing. The 2024 and 2025 forecasts were lifted as the disinflation process became slower. The 2026 projection, which had once sat near the target-consistent zone, moved up materially by mid-2026. Even the later-year lines no longer look like a clean glide path back to 2%.

That pattern matters because the SEP is not a private research note. It is one of the Fed’s main communication tools. It tells households, firms, investors, and legislators how policymakers understand the economy. The SEP does not mechanically bind policy, but it shapes expectations. When the forecast repeatedly implies that inflation will soon behave, and inflation repeatedly refuses to behave, the public receives an implicit message: the institution is always close to success in theory, but not quite there in practice.

The forecast revisions also reveal how much policy depended on a benign baseline. The Fed’s policy stance is normally justified by its expected effect on future inflation and employment, not by spot inflation alone. If officials believe underlying inflation will decelerate, they can tolerate current overshoots more easily. But if that belief is wrong year after year, the real-time stance is too easy relative to the true state of inflation. A forecast error therefore becomes a policy error, even if the error was intellectually honest at the time.

This is why the chart should be read as a credibility diagnostic. It is not enough to say that inflation is above target. Markets already know that. The deeper issue is that the institution’s own projected path has repeatedly moved in the direction that made prior policy look less adequate. When an institution revises the destination after every leg of the journey, people begin to question whether the map is useful.

The Difference Between Random Error and Systematic Bias

Every forecast contains error. In a well-calibrated forecasting process, some misses should be too high and others too low. Over time, the errors should look roughly balanced unless the economy is repeatedly hit by shocks from one side. That is why one-sided forecast error is so important. It is evidence that the model may be missing a persistent force. In this case, the missing force appears to be inflation inertia.

Inflation inertia can come from several channels. Wages can adjust slowly after a price shock, preserving cost pressure even after goods prices normalize. Housing inflation can lag market rents. Services inflation can remain sticky because it is labor-intensive. Firms can become more comfortable passing through cost increases once consumers have learned to expect price changes. Fiscal deficits can sustain demand. Geopolitical shocks can keep energy and goods prices volatile. None of these channels alone explains everything, but together they make inflation less transitory than a simple supply-shock narrative would imply.

A systematic bias can also come from institutional incentives. Central banks want to avoid unnecessary recessions. They also want to avoid appearing panicked. Forecasting a gradual return to target is often the least disruptive communication strategy. It reassures markets, gives policymakers optionality, and avoids a direct admission that policy may need to be tighter for longer. But if that communication becomes too comfortable, it can create a bias toward forecasted disinflation. The baseline becomes a tool of policy smoothing.

This is not a claim of bad faith. It is a claim about process. Institutions can be sincere and still biased. A model can be technically sophisticated and still miss the dominant regime shift. A committee can be full of experts and still converge on forecasts that underweight unpleasant outcomes. The question is not whether officials tried hard. The question is whether the process learned fast enough from repeated misses in the same direction.

Why Core PCE Matters More Than the Headline Debate

Core PCE matters because it is the measure the Fed uses to judge underlying inflation pressure. Headline inflation can be moved by food and energy prices, and those components often reverse. Core inflation is not perfect, but it is meant to filter out some of that noise. When core PCE remains above target for years, the inflation problem cannot be dismissed as a sequence of isolated commodity shocks.

The April 2026 reading of 3.3% is important because it sits far enough above 2% to matter economically, yet low enough to tempt complacency. This is not a 1970s-style double-digit inflation rate. It is a slow credibility bleed. A 3.3% core inflation rate can be described as elevated but manageable. That is exactly why it is dangerous. It encourages a policy language of patience while allowing the price level to keep compounding above the intended path.

The arithmetic is straightforward. If prices rise at 3.3% instead of 2% for five years, the cumulative price level ends up roughly 6% to 7% higher than it would have been under target-consistent inflation. For households, that is not an abstract rounding error. It changes rent burdens, wage negotiations, retirement budgets, insurance costs, and the political meaning of economic growth. For investors, it changes discount rates, real yields, equity multiples, credit spreads, and currency valuation.

Core PCE also matters because it is embedded in the Fed’s own credibility contract. The Fed has defined price stability around 2% inflation. If the preferred measure stays above that level for a long time and the forecast keeps promising convergence later, the public may conclude that 2% is not the operating target. It is the long-run story the institution tells while tolerating a higher realized regime.

Forecast-Based Policy and the Risk of Waiting Too Long

Monetary policy works with lags, so central banks must be forward-looking. That is unavoidable. A central bank that reacts only to current inflation will often move too late. The problem is not that the Fed uses forecasts. The problem is what happens when the forecast is persistently too optimistic.

If officials believe inflation will fall from 3.3% toward 2% without much additional restraint, then policy can be held steady or even eased as soon as labor-market softness appears. If the true inflation process is more persistent, that approach leaves real policy less restrictive than intended. The expected real rate may look high on paper, but the realized inflation process keeps eroding the stance. Markets then price the possibility that the Fed will tolerate more inflation to protect growth.

This creates a credibility wedge. The Fed says policy is data dependent, but the data are filtered through a forecast. If the forecast repeatedly leans dovish on inflation, then data dependence can become forecast dependence. Officials wait for inflation to behave because the model says it should. When it does not, they revise the forecast and repeat the exercise. The result is not a single dramatic policy mistake. It is a sequence of small delays that compound into years of above-target inflation.

This dynamic is familiar from the literature on real-time monetary policy. Athanasios Orphanides showed that policymakers can make poor decisions when they rely on real-time estimates that later prove wrong, especially around potential output and inflation pressure. The lesson is not that forecasts should be abandoned. It is that central banks must be humble about real-time uncertainty and must respond aggressively when the distribution of errors becomes one-sided.

Credibility Is an Asset, Not a Slogan

Central-bank credibility is often discussed as if it were a moral trait. It is better understood as an asset. It reduces the cost of disinflation because households and firms believe the central bank will defend the target. If wage setters, price setters, and bond investors trust the target, they do not need to build as much inflation risk into contracts. That trust lowers the sacrifice ratio: the amount of output and employment pain needed to bring inflation down.

Kydland and Prescott, and later Barro and Gordon, formalized the time-inconsistency problem. A central bank may have an incentive to promise low inflation and then tolerate higher inflation to support employment or reduce real debt burdens. Rational agents understand that incentive, so they demand higher inflation compensation unless the institution has a credible commitment mechanism. The 2% target is supposed to be that mechanism.

But commitment is revealed in hard cases, not easy ones. When inflation is near target and growth is stable, every central bank looks credible. The test comes when inflation is above target for a long time and political pressure favors easier policy. If the public sees repeated upward revisions to inflation forecasts and a policy stance that appears calibrated to hoped-for disinflation, the commitment mechanism weakens. The target begins to look conditional.

Credibility loss is nonlinear. It can appear stable for years and then reprice quickly. Long-term inflation expectations may look anchored until they are not. Bond term premia may remain contained until investors demand compensation for regime uncertainty. Currency markets may tolerate policy ambiguity until a shock forces a reassessment. That is why the Fed should not treat credibility as something it can repair later at low cost.

The Market Implications

For markets, the issue is not simply whether the Fed cuts, holds, or hikes at the next meeting. The bigger issue is whether investors should assign a higher probability to a structurally higher inflation regime. If the Fed’s forecast errors are systematic, then the distribution of future rates, real yields, and risk premia should be wider than a benign soft-landing narrative implies.

Equities are affected through both earnings and discount rates. Moderate inflation can support nominal revenue, but persistent inflation eventually pressures margins, raises wage costs, increases working-capital needs, and lifts discount rates. Long-duration growth stocks are especially sensitive because more of their value comes from cash flows far in the future. If the market concludes that inflation persistence forces a higher real-rate floor, valuation multiples should be lower than in the pre-2020 regime.

Credit markets face a different tradeoff. Higher nominal growth can help borrowers in the short run, but persistent inflation can also keep policy restrictive, raise refinancing costs, and expose weak balance sheets. Private credit and leveraged borrowers are particularly sensitive to the difference between expected and realized rates. If policy remains tighter for longer because inflation does not converge, interest coverage becomes the transmission channel.

The Treasury market is the cleanest expression of the credibility question. A credible Fed can keep long-run inflation risk contained even when near-term inflation is high. A less credible Fed requires a higher term premium. That term premium is not only about inflation. It is also about fiscal deficits, debt supply, policy uncertainty, and the risk that the central bank is slower to react than investors previously assumed. Repeated forecast misses therefore matter for the entire curve.

The Taylor Rule Lens

A simple Taylor-rule framework helps clarify why forecast errors matter. A stylized rule says the policy rate should rise when inflation is above target and when output is above potential. The exact coefficients are debatable, but the intuition is robust: higher inflation should call for a higher nominal policy rate, especially when the labor market is not collapsing.

The problem is that the rule depends on the inflation input. If policymakers use expected inflation that is too low, the implied policy rate is too low. Suppose core inflation is 3.3%, the target is 2%, and the neutral real rate is around 1%. A simple prescription might require a nominal rate meaningfully above neutral to generate restraint. But if the forecast says inflation will soon fall toward 2.3% or 2.1%, the required stance looks less restrictive. The forecast lowers the apparent urgency.

This is how an optimistic forecast can substitute for actual tightening. The committee can say policy is restrictive because expected inflation is falling. Markets can believe cuts are approaching because the forecast path shows convergence. Financial conditions can ease in anticipation. That easing can then support demand and make inflation more persistent. In this way, forecast-based policy can become self-defeating when the forecast is too benign.

A rule-based perspective does not mean policy should be mechanical. The Fed must account for financial stability, labor-market risk, supply shocks, and uncertainty. But rules are useful because they discipline discretion. They force policymakers to explain why repeated inflation overshoots do not require a stronger response. Without that discipline, the institution can drift into a pattern of explaining away every miss as temporary.

The Role of Expectations

Inflation is partly backward-looking and partly forward-looking. Past inflation affects wages, contracts, and pricing norms. Expected future inflation affects current behavior. The Fed’s credibility problem sits at the intersection of those forces. If people believe the Fed will restore 2%, they have less reason to demand inflation protection. If they believe the Fed will tolerate 3% for several years, behavior adapts.

Expectations do not need to become unanchored dramatically to create trouble. A small upward drift can matter. Wage negotiations can move from 2% cost-of-living assumptions to 3% or 4%. Firms can raise prices more frequently because customers are less surprised. Bond investors can demand more inflation compensation. Households can pull consumption forward when they expect prices to rise. These micro-level adjustments make disinflation harder.

The danger is that the Fed may focus too much on survey measures and market breakevens that still look broadly anchored. Those measures are useful, but they are not complete. Inflation expectations are embedded in behavior as well as surveys. If the economy keeps producing above-target core inflation, then revealed expectations may be less anchored than measured expectations suggest.

This is why the next Fed chair’s communication challenge is so difficult. Restoring confidence does not require theatrical hawkishness. It requires convincing the public that the institution will not use optimistic forecasts as an excuse to postpone the target. That means explaining the conditions under which policy would become more restrictive, not only the conditions under which it would ease.

Why the Next Chair Inherits a Harder Job

The next Fed chair inherits more than a macro forecast. He inherits a credibility balance sheet. On one side is the institution’s long history of post-Volcker inflation-fighting credibility. On the other side is a recent record of above-target inflation, forecast revisions, and public frustration with the cost of living. The question is which side markets will weight more heavily in the next shock.

The challenge is not simply to sound tough. Overly aggressive rhetoric can damage credibility if it is not matched by action, and excessive tightening can create unnecessary recession risk. The harder task is to rebuild a reaction function that people can understand. The public needs to know that 2% is not a ceremonial number. Markets need to know that the Fed will not validate a higher inflation regime because returning to target is inconvenient.

A credible chair should probably reduce the institution’s dependence on precise point forecasts. The Fed can still publish projections, but it should emphasize forecast uncertainty, alternative scenarios, and reaction-function thresholds. Instead of repeatedly saying inflation is expected to fall, it should explain what it will do if inflation does not fall. That shift would move communication from prediction to commitment.

The chair also needs to confront the political economy of inflation. Above-target inflation redistributes wealth. It hurts cash holders, wage earners with weak bargaining power, and retirees on fixed incomes. It benefits some borrowers and nominal asset holders. It creates political anger even when unemployment is low. A central bank that appears too patient with inflation risks losing legitimacy beyond financial markets.

What Would Restore Confidence

Restoring confidence starts with acknowledging the pattern. The Fed does not need to apologize for every forecast miss, but it does need to recognize that repeated one-sided errors are informative. A serious institution learns from the direction of its mistakes. If inflation has been more persistent than expected for several years, the default assumption should shift. The burden of proof should move from “inflation will stay high” to “inflation will actually return to target.”

Second, the Fed should communicate policy in terms of realized progress, not only projected progress. It can say that cuts require sustained evidence of core inflation moving toward 2%, not merely forecasts that it will do so. That does not mean waiting for inflation to hit exactly 2% before easing. It means requiring enough realized evidence to regain confidence in the model.

Third, the Fed should make the target symmetric in practice. For years after the global financial crisis, undershooting 2% shaped the institutional memory. After the pandemic inflation shock, the relevant risk changed. A flexible average inflation framework can make sense in theory, but if the public interprets flexibility as tolerance for persistent overshoots, the framework becomes a liability. Symmetry means overshoots deserve the same seriousness that undershoots once received.

Fourth, the Fed should resist using labor-market softness as an automatic reason to declare victory over inflation. The dual mandate matters. But if inflation persistence remains high, easing too soon can create a worse tradeoff later. A delayed disinflation often requires more pain than an earlier one. Credibility is valuable precisely because it reduces that pain.

Investment Framework for a Sticky-Inflation Fed

Investors should treat the chart as a regime-risk indicator. It does not say that inflation will accelerate forever. It says the probability of slow convergence is higher than the Fed’s prior forecasts implied. That affects portfolio construction. A portfolio built for quick disinflation, falling policy rates, and a return to the 2010s discount-rate regime is exposed if inflation remains sticky.

The first implication is duration discipline. Long-duration bonds and long-duration equities both benefit when inflation falls and real rates decline. If disinflation is slower, duration is less attractive unless yields already compensate for the risk. Investors do not need to avoid duration entirely, but they should understand that the hedge properties of long Treasuries are weaker when inflation credibility is in question.

The second implication is quality. Companies with pricing power, low leverage, high free cash flow, and limited refinancing needs are better positioned in a sticky-inflation regime. Businesses that require cheap capital, long payback periods, or constant refinancing are more vulnerable. Inflation persistence separates nominal growth from real value creation.

The third implication is diversification across inflation outcomes. Some assets benefit from disinflation and lower rates. Others benefit from nominal growth, commodities, real assets, or pricing power. The right mix depends on valuation, but the key is to avoid a single macro bet that the Fed’s optimistic inflation path will finally be right.

The Balance Sheet Between Growth and Inflation

A persistent forecasting bias also changes how investors should read the growth-inflation tradeoff. In a normal cyclical slowdown, weaker demand should reduce inflation pressure and give the Fed room to ease. In a sticky-inflation regime, the same slowdown is less comforting. Growth can soften while inflation remains too high because the remaining inflation is concentrated in services, wages, housing, insurance, healthcare, or other categories that do not respond quickly to lower goods demand. That makes the policy frontier worse.

This is why the repeated upward revisions matter beyond the inflation data itself. They imply that the Fed’s perceived tradeoff was often too favorable. If officials expected inflation to fall with limited labor-market damage, they could describe the economy as moving toward a soft landing. But if inflation persistence is higher, then the same policy path may deliver neither clean price stability nor durable growth. The institution ends up spending credibility while waiting for a better tradeoff to appear.

There is also a fiscal dimension. Persistent inflation can reduce the real burden of nominal debt, which creates political temptation, but it also raises nominal interest costs as debt rolls over. In a high-debt economy, the central bank’s credibility becomes even more valuable because it prevents inflation risk from being capitalized into the entire Treasury curve. If investors begin to suspect that the central bank is too tolerant of above-target inflation, the fiscal cost of that suspicion can show up through higher term premia and higher debt-service costs.

The Fed therefore faces a more complex balance sheet than the simple dual-mandate language suggests. On the asset side, it has credibility, institutional independence, and the ability to coordinate expectations. On the liability side, it has repeated forecast errors, political pressure, and a public that has already absorbed a large cumulative price-level increase. The next policy regime will be judged by whether it rebuilds the asset side faster than the liability side compounds.

Conclusion: The Target Must Become Operational Again

The chart’s message is uncomfortable because it turns a technical forecasting issue into a governance issue. Core PCE inflation has remained above target for nearly five years, and the Fed’s own projections have repeatedly moved higher as reality failed to match the expected glide path. Forecasting difficulty is understandable. Repeated one-sided optimism is more serious.

The risk is that policy becomes anchored to hoped-for disinflation rather than delivered disinflation. That approach can look reasonable meeting by meeting, but over time it weakens the credibility of the 2% target. The public does not experience inflation as a forecast path. It experiences inflation as a cumulative price level, a rent bill, a grocery bill, a wage negotiation, and a savings problem.

For the next Fed chair, the most important task may be restoring the operational meaning of 2%. Not as a long-run aspiration, not as a number that appears in projection tables, but as a policy objective that shapes decisions even when the tradeoff is uncomfortable. The Fed does not need perfect forecasts to regain credibility. It needs a reaction function that learns from its mistakes and treats the target as binding.

That is the deeper lesson of the chart. Inflation persistence was underestimated. Policy was too often calibrated to the inflation officials expected rather than the inflation the economy delivered. The next stage is not about producing a more elegant forecast. It is about convincing markets and households that if the forecast is wrong again, the target will still win.