A plain-English guide to r-star, the neutral rate.

Every speech a Fed Chair gives that touches the "longer run" references one number: r-star, the real interest rate consistent with full employment and stable inflation. The longer-run dot in the SEP is essentially the Committee's estimate of r-star plus 2 percent. Markets watch this dot more carefully than any other in the plot because, while the year-end dots move with the business cycle, the longer-run dot reflects a structural view of where the economy is headed. With Warsh's first FOMC tomorrow, the longer-run dot is where the most-watched cross-asset implications sit. This is the explainer.

What r-star actually is

In words, r-star is the real interest rate the Federal Reserve would set if the economy were at full employment, inflation were at target, and no shocks were present. It is the rate that is neither stimulative nor restrictive. In equations, it sits in the Taylor rule as the constant the policy rate is benchmarked against:

i = r* + π + 0.5(π − 2%) + 0.5(y − y*)

where i is the nominal policy rate, π is realised inflation, and y − y* is the output gap. When the policy rate equals r* + π, the rule is neutral. When it exceeds that, policy is restrictive; when it falls below, policy is accommodative. The published "longer-run" dot is essentially the Committee's estimate of r* + 2 percent (the target inflation rate).

Why r-star is hard to estimate

Three structural problems make r-star one of the most contested objects in macroeconomics.

1. It is unobservable. No data series exists that directly measures the neutral rate. It has to be backed out of a model.
2. It is time-varying. r-star moves with structural factors: demographics, productivity growth, the global saving glut, the safe-asset shortage. Any model has to allow it to drift, which means more parameters to estimate from limited data.
3. The error bands are wide. Published r-star estimates carry standard errors of 100 to 150 basis points around the point estimate. The published number is best understood as "approximately X, with high uncertainty" rather than a precise value.

The three published models

• Laubach-Williams (LW). The original published US estimate, from a 2003 paper. Uses a Kalman-filter-based approach to extract r-star from observed GDP, inflation, and the policy rate. Published by the San Francisco Fed.
• Holston-Laubach-Williams (HLW). A 2017 update that adjusts for trend GDP growth more directly. Tends to produce slightly higher r-star estimates than the original LW in low-growth regimes. Published quarterly with a one-quarter lag.
• Lubik-Matthes (LM). A 2015 alternative, published by the Richmond Fed. Uses a different model class that incorporates financial-market data more directly. Tends to produce systematically higher r-star estimates than LW or HLW; the gap has been roughly 50 basis points through the post-2010 sample.

All three are useful; none is the "right" answer. The Committee members read all three plus their own internal estimates, and the median longer-run dot is the result.

The three estimates over time

Why the longer-run dot has been drifting higher

Through 2024-2025 the median longer-run SEP dot drifted from 2.50 percent to 3.00 percent, the largest cumulative move in the post-2012 SEP history. The drift reflects what each model is telling Committee members:

• Productivity growth has stabilised higher. Post-COVID labour-force re-allocation appears to have produced sustained productivity gains, which lifts both trend GDP and r-star.
• Treasury supply is higher. Persistent fiscal deficits raise the safe-asset supply, which absorbs some of the structural saving glut that depressed r-star through 2009-2019.
• Demographics are slower-moving but turning. The retirement of the boomer cohort has been a tailwind to lower r-star (savers exit the workforce); but the cohort effects are starting to roll off, and the millennial consumption profile favors more spending than saving.

Each of these three pushes the LM model harder than LW and HLW. That is why the LM estimate has risen to roughly 2.0 percent while LW and HLW sit closer to 1.1 percent. The Committee's median longer-run dot at 3.00 percent (r* of 1.00 percent plus 2 percent inflation) is currently anchored to LW and HLW; Warsh has written that he favours LM, which would imply a 4.00 percent longer-run dot.

What a higher r-star changes

The downstream implications of a 50 basis-point shift in the Committee's r-star estimate are wider than most analysts realise.

1. Every Taylor-rule-derived policy benchmark shifts. A 50 basis-point higher r-star means the policy rate is, mechanically, 50 basis points less restrictive at any given level of inflation. Today's 3.625 percent policy rate is restrictive against a 2.5 percent inflation target plus 1.1 percent r-star (LW), but essentially neutral against the same inflation target plus 2.0 percent r-star (LM).
2. The OIS curve's terminal rate rises. Every long-end fixed income trade is referenced against the Committee's implicit r-star. A 50 basis-point r-star shift moves the terminal rate that markets price by roughly 30 basis points.
3. The dollar smile's right-edge gets steeper. Higher US r-star against unchanged G10 r-star means wider real-yield differentials, which support DXY through the rate-channel. Half the dollar's run from 95 to 100 over the past month can be attributed to this re-pricing in r-star expectations.
4. The term-premium story changes. A higher r-star means a higher neutral 10-year nominal yield. If r-star is 2 percent and target inflation is 2 percent, the neutral 10-year is around 5 percent. The 10-year at 4.5 percent is then accommodating relative to neutral, not restrictive.

What to watch at the SEP tomorrow

• The median longer-run dot. A move from 3.00 percent to 3.125 percent is the most likely direction. A move to 3.25 percent would be the cleanest signal that Warsh's framing is shifting the Committee.
• The dispersion in the longer-run column. A wider central tendency in the longer-run dots than in March signals that the Committee is genuinely divided on r-star, which is what would happen if Warsh is leaning LM-style higher while the rest of the Committee stays anchored to LW.
• Press conference framing. A direct reference by Warsh to Lubik-Matthes or to "higher structural neutral rate" is the most actionable observation, even more so than the dot itself.

What r-star does not tell you

1. The current rate. r-star is the long-run anchor; today's actual short rate is set by the Committee based on current inflation and slack.
2. The path. r-star tells you where the policy rate will average over the long run, not when it will get there. The next two years' dots in the SEP capture the path.
3. Whether the model is right. All three published models have wide error bands. The Committee's published longer-run dot has a roughly 50 basis-point one-standard-deviation interval implied by the dispersion in the SEP.

Related reading

• Today's analysis applies this directly: FOMC eve: what's priced and what isn't.
• The SEP and dot-plot framework: A plain-English guide to the SEP and the dot plot
• How r-star transmits through term premium: A plain-English guide to the term premium
• The dollar-smile channel: A plain-English guide to the dollar smile