Neutral Interest Rate

What Is the Neutral Rate?

The neutral interest rate, written as r* (pronounced "r-star"), is the most important number in finance that nobody can directly observe. It is the short-term real interest rate at which monetary policy is neither stimulating nor restricting the economy: output is growing at potential, unemployment is at its natural rate, and inflation is stable at the central bank's target.

Every monetary policy decision the Federal Reserve makes is implicitly a judgment about r*. When the Fed raises rates, it is raising them above r* to cool the economy. When it cuts, it is lowering them below r* to stimulate. The entire framework of "tight" versus "loose" policy depends on where r* actually is, and that is one of the most fiercely debated questions in economics.

For traders, r* is not an abstraction. It determines the terminal rate of every cutting cycle, the anchor for the long end of the yield curve, the justified P/E ratio for equities, and the equilibrium level of the dollar. Getting r* wrong by even 50 basis points means mispricing virtually every financial asset on the planet.

Why r* Matters: The Plumbing of Everything

The Policy Stance Calculator

The "stance" of monetary policy, whether it is tight, neutral, or loose, is calculated as:

Policy Stance = Real Fed Funds Rate − r*

Where the real fed funds rate = nominal fed funds rate − core inflation (usually core PCE).

The same nominal rate can be either very tight or quite loose depending on where r* sits. This is why the r* debate is not academic, it directly determines how many rate cuts are "appropriate" and how long rates will stay elevated.

The Great r* Debate: Low vs. High

Two camps dominate:

Team Low r (secular stagnation)*:

• Proponents: Some academic economists
• Estimate: Real r* = 0.5-1.0% → nominal neutral ≈ 2.5-3.0%
• Reasoning: Ageing demographics suppress investment demand; technology is deflationary; inequality channels income to high savers; the savings glut persists
• Implication: The Fed at 5.5% was extremely restrictive. Many rate cuts are justified. Long-duration assets are cheap.

Team High r (fiscal/structural shift)*:

• Proponents: Larry Summers, some Fed officials, bond market "higher-for-longer" traders
• Estimate: Real r* = 1.5-2.5% → nominal neutral ≈ 3.5-4.5%
• Reasoning: Massive fiscal deficits crowd out private investment; deglobalisation raises costs; energy transition requires enormous capex; defence spending increasing globally; AI capex boom absorbs savings
• Implication: The Fed at 5.5% was only mildly restrictive. Few rate cuts are justified. Long-duration assets are expensive.

The financial market difference between these two views is staggering. If Team Low is right, the 10-year Treasury yield should trade at 3.0-3.5%, the S&P 500 deserves 22-25x forward earnings, and 6-8 rate cuts are coming. If Team High is right, the 10Y should trade at 4.5-5.0%, the S&P deserves 16-18x, and only 2-3 cuts are coming. Every major cross-asset trade is ultimately a bet on r.*

How r* Is Estimated

The Laubach-Williams Model

The most cited r* estimate comes from the Laubach-Williams (LW) model, developed by Fed economists Thomas Laubach and John Williams (now president of the NY Fed). The model uses a statistical filtering technique (Kalman filter) to extract the unobservable neutral rate from observed data on GDP growth, inflation, and interest rates.

Key characteristics:

• Updates quarterly
• Estimated real r* declined from ~3.5% in the 1960s to ~0.5% by 2019
• Has very wide confidence intervals (±200bps or more)
• Responds slowly to regime changes

The Holston-Laubach-Williams (HLW) Model

An extension that incorporates international data. Available on the NY Fed's website. The HLW model shows r* rising in 2023-2025 from its post-GFC trough, but the estimates lag real-time economic conditions significantly.

Market-Implied r*

Traders can extract the market's r* estimate from:

• FOMC longer-run dot: Currently ~3.0% nominal, implying ~1.0% real r*
• 5-year forward 5-year real yield (from TIPS): The real yield the market expects 5-10 years from now, a proxy for where the market thinks real r* will settle
• Fed funds futures terminal rate: The rate at which the futures curve flattens (implying no further cuts expected) approximates nominal r*

The "Flickering Light" Problem

Fed Chair Powell famously described r* as a star that is "only visible by the light it casts on other variables." You cannot observe r* directly, only infer it from how the economy responds to the current rate. If the economy is growing strongly and inflation is rising despite a 5% fed funds rate, r* must be higher than previously thought. If the economy is weakening rapidly at 5%, r* must be lower.

This creates a fundamental challenge: you can only know r with confidence in hindsight.* By the time you've confirmed that the current rate is above or below r*, the economic data has already moved.

The Evolution of r*: A Historical Perspective

The Secular Decline (1980-2019)

The Post-COVID Reassessment (2020-Present)

The pandemic triggered a structural reassessment:

1. Fiscal explosion: US federal debt jumped from $23T to $35T+. Annual deficits of 5-7% of GDP during full employment are historically unprecedented. This debt must be financed, requiring higher real rates to attract buyers.

2. Deglobalisation: Supply chain reshoring, friend-shoring, and tariff escalation reverse the deflationary force that suppressed r* for two decades. Producing goods domestically costs more, sustaining higher inflation and requiring higher rates.

3. Energy transition capex: Achieving net-zero emissions requires $3-5 trillion annually in global investment (IEA estimates). This massive capital demand competes with other investment, pushing up the equilibrium rate.

4. AI infrastructure: Data centre construction, semiconductor fabs, and power generation for AI are absorbing hundreds of billions in annual investment, a new demand source for capital that didn't exist pre-2020.

5. Defence spending: NATO countries ramping from 1.5% to 2-3% of GDP in defence budgets represents hundreds of billions in additional government borrowing globally.

6. Demographic shift in China: China's working-age population is declining, reducing the "savings glut" that flowed into US Treasuries for two decades. Less foreign demand for US bonds means higher yields to clear the market.

Each factor individually might push r* up by 25-75bps. Cumulatively, they suggest a real r* 100-200bps higher than the post-GFC consensus, one of the most consequential macroeconomic regime shifts in decades.

Trading r*: Practical Applications

The Terminal Rate Trade

The most direct r* trade is positioning for the terminal rate, where the cutting cycle ends:

• If you believe r is low (2.5-3.0% nominal)*: Buy 2-year Treasuries when the fed funds rate is well above this level. Each rate cut toward neutral is capital appreciation. This was the winning trade from late 2023 through mid-2024 as the market priced in aggressive cuts.
• If you believe r is high (3.5-4.5% nominal): Short front-end rate expectations (via SOFR futures) when the market is pricing cuts to below your r estimate. The market will be disappointed by a shallower cutting cycle.

The Curve Shape Trade

r* beliefs drive the yield curve shape:

• Low r world: Curve should be steep (short rates headed much lower; long rates reflect r + term premium)
• High r world*: Curve is flat or inverted at current rate levels (short rates aren't going much lower; long rates reflect a higher floor)

When the consensus r* estimate shifts, the curve reprices:

• r* rising → curve flattens (long rates anchored; short rate cuts priced out)
• r* falling → curve steepens (more cuts expected; long rates fall)

The Equity Valuation Framework

For equity investors, r* determines the justified P/E ratio:

A 100bps increase in r* compresses the justified P/E by roughly 1.5-2.0 turns, representing a 10-15% reduction in fair value for the S&P 500 without any change in earnings expectations.

This is why the r* debate is not just for bond traders. Every equity investor who uses a DCF model is making an implicit r* assumption.

The Dollar and Cross-Border Trade

The US dollar's long-run equilibrium level is influenced by the r* differential between the US and other major economies. If US r* has risen while European and Japanese r* remain low, the real rate differential widens, attracting capital to the US and supporting a structurally stronger dollar.

This has profound implications:

• EM assets: A higher US r* means a stronger structural dollar, which is a persistent headwind for emerging market debt and equities
• Gold: Higher real r* raises the opportunity cost of holding gold (which yields nothing), creating a structural headwind for gold prices, unless inflation expectations rise even faster
• Commodities: A stronger dollar from higher r* depresses dollar-denominated commodity prices, all else equal

Historical Episodes Where r* Mattered

The Greenspan "Conundrum" (2004-2006)

In 2004-2006, Greenspan raised the fed funds rate from 1.0% to 5.25%, but long-term yields barely moved, the 10Y stayed around 4.5-5.0%. Greenspan called this a "conundrum." The explanation: global r* had fallen (Asian savings glut, demographics), so even as the Fed tightened, long-term rate expectations didn't rise. The conundrum was the market telling the Fed that r* was lower than it thought, a message the Fed didn't fully heed, contributing to the housing bubble.

The 2018-2019 Overtightening

In 2018, the Fed raised rates to 2.25-2.50%, believing r* was around 3.0% nominal. But the economy slowed sharply, equity markets sold off 20% in Q4 2018, and the repo market froze in September 2019. The Fed reversed course and cut three times in 2019. The lesson: r* was lower than the Fed estimated, and tightening into a low-r* world caused financial instability well before reaching what the Fed considered "neutral."

The 2022-2023 Resilience Puzzle

The Fed raised rates from 0% to 5.25-5.50%, the most aggressive tightening in 40 years. Yet the economy barely slowed: GDP grew above trend, unemployment stayed near 3.5%, and the stock market rallied. The explanation: r* had risen. What appeared to be a severely restrictive stance was actually only mildly restrictive because r* had shifted upward by 100-200bps from its post-GFC level.

Limitations

1. Unobservable: r* cannot be measured directly, only estimated with models that carry enormous uncertainty (±200bps confidence intervals)
2. Time-varying: r* changes as structural factors evolve, but models update slowly, always fighting the last war
3. Model-dependent: Different models produce very different r* estimates. The Laubach-Williams model shows r* near 1% real; equilibrium models using fiscal projections suggest 2%+ real
4. Circular reasoning risk: If the Fed believes r* is low and keeps rates low, the resulting asset price inflation may raise r* through wealth effects, the observation changes the quantity being measured
5. Multiple equilibria: There may not be a single stable r*, different policy regimes may produce different equilibrium rates

What to Watch

1. FOMC longer-run dot: The Committee's r* proxy. Track shifts across SEP releases, even 12.5bps moves are significant.
2. 5Y5Y real yield (TIPS): The market's medium-term r* estimate. A breakout above 2.0% real signals the market is accepting the higher-r* thesis.
3. Economy's response to current rates: If growth stays resilient and unemployment stays low at 4.5-5.0% fed funds, r* is higher than you think. If growth slows sharply, r* is lower.
4. Federal deficit trajectory: Widening deficits push r* higher; fiscal consolidation lowers it. Watch the CBO's annual long-term projections.
5. Global savings flows: China's current account surplus, Japanese foreign bond purchases, and OPEC petrodollar recycling all affect the global supply of savings and thus r*. Shifts in these flows reprice r* with a lag.