VAN R. HOISINGTON AND LACY H. HUNT: Until broad money growth falls back toward a 3% benchmark, inflation is likely to stay above the Fed's 2% target

The Gist

Hoisington and Hunt say inflation will keep running above 2 percent until the Fed slows money growth to about 3 percent a year, because the economy cannot grow fast enough to absorb more money, and inflation will not fix itself. This is a steelman: Van R. Hoisington and Lacy H. Hunt's arguments are reconstructed in their strongest fair form, not endorsed.

Conclusion

Until broad money growth slows toward a benchmark of about 3 percent a year, inflation is likely to stay above the Fed's 2 percent target: money is growing roughly twice as fast as the economy's slower potential growth and rising velocity can absorb, demographics and productivity keep that benchmark low, and inflation does not fall back on its own without an active, credible policy response. Temporary factors may lower inflation for a time without changing this.

Premises

The non-inflationary rate of broad money growth today is about 3 percent a year (range 2.75 to 3.25 percent): roughly 2 percent potential real growth plus the 2 percent inflation target, less about 1 point for rising velocity.
Broad money growing at about 6.5 percent a year, roughly double the 3 percent benchmark, is a monetary excess similar in scale to 1979 and risks nominal spending outrunning productive capacity, keeping inflation and inflation expectations above target.
Inflation does not reliably fall back to target on its own; it takes an active, credible monetary regime, and the Fed's delayed response in 2021 and 2022 is why inflation has stayed above 2 percent for 66 straight months.
Slowing population and labor-force growth (roughly zero now) lowers potential real growth and therefore the rate of money growth the economy can absorb without inflation.
Aggregate productivity growth is likely to stay near its long-run 2 percent pace despite AI, because a large government sector, a services-heavy job mix and lost economies of scale limit how far gains spread.

Assumptions

Broad money means the Federal Reserve's broad monetary aggregate (written M-2 here), the measure the letter uses throughout.
The relation between money growth, velocity and inflation holds over a horizon of a few years.
Financial interest: Hoisington Investment Management manages U.S. Treasury portfolios and for decades was known for holding very long-duration Treasuries on a falling-inflation, falling-rates view. Press reports on its 2026 letters say the firm cut effective duration from about 20.9 years at the end of September 2025 to under one year by June 30, 2026, so this letter's forecast of higher long-term yields is consistent with the firm's current short-duration positioning rather than with its earlier long-duration stance. Sources: "Decades long bull turns bearish. Cites structural deficits…

Analysis

Argument type: Inductive.

Potential Fallacies

False precision (Premise 1 (and 'twice the benchmark' in Premise 2)) — The 2.75–3.25% band is built from three uncertain estimates: potential growth, the inflation target and the velocity trend. Their errors compound, so the true credible range is much wider than ±0.25 points. A modest change in the velocity assumption moves the benchmark by a point or more, and with it the 'double the benchmark' framing.
Post hoc / single-cause attribution (Premise 3) — The Fed's delayed 2021–22 response is presented as 'why' inflation has stayed above 2% for 66 months. Supply disruptions, fiscal stimulus and deficits, energy prices, shelter lags and tariffs are plausible co-contributors, and no counterfactual is offered to isolate the Fed's share. A late Fed may be one cause among several, not the sufficient explanation.
Weak or selective historical analogy (Premise 2) — Comparing today's money growth to 1979 treats similarity in one variable as similarity in the whole situation. Starting inflation, expectations anchoring, velocity behavior, financial structure and Fed credibility all differ. Episodes where money growth ran well above a benchmark without sustained inflation, such as the post-2008 years, are not discussed.
Unfalsifiable hedging (Conclusion (final sentence)) — The clause that temporary factors may lower inflation 'without changing this' lets falling inflation be read as transitory and persistent inflation as confirmation. With no stated time horizon, index or disconfirming observation, it is unclear what evidence would count against the thesis.
Partial circularity (benchmark built on the target) (Premises 1–2 and Conclusion) — The benchmark includes the 2% target as an input, and the conclusion says inflation stays above target when money growth exceeds the benchmark. This follows largely from the quantity-theory identity. The informative content depends on the unstated assumption that velocity and potential growth take the stipulated values.
Overgeneralization from a limited sample (Premise 3) — The claim that inflation is not mean reverting rests on regime-specific history. Other episodes, including the 2022–24 fall from about 9% to roughly 3% without a recession, show substantial disinflation without a dramatic regime change. The claim is overstated as a general rule.
Speculative projection presented as likely fact (Premise 5) — Productivity near 2% despite AI is a forecast about a nascent general-purpose technology. It is supported by plausible but unquantified structural drags (government size, services mix, scale). Historical technologies diffused with long and variable lags, so confidence in a specific figure is higher than the evidence supports.

Counterarguments

These challenges are not ranked.

Premise 1 (High impact) — Velocity is unstable and endogenous to rates and expectations. A sensitivity check with velocity moving ±2 points yields benchmarks from about 1% to 5%, so '6.5% is double the benchmark' holds only under one velocity assumption. The 2010s showed large M2 expansion with below-target inflation, and 2022–23 showed M2 contraction with disinflation and no recession.
Conclusion (High impact) — The M2–inflation link is too loose at multi-year horizons to anchor a forecast. Most central banks dropped monetary targeting after the 1980s–90s for this reason. Inflation in 2021–26 is better explained by supply shocks, fiscal impulse, tariffs, shelter lags and expectations, with M2 as an incidental or lagging indicator.
Premise 2 (Medium impact) — M2 is partly endogenous to credit demand, deficits and portfolio reallocation, so growth of 6.5% may partly reflect asset shifts, or a rebound from the 2022–23 contraction, and not 1979-style excess. In 1979 inflation was already near double digits and expectations were unanchored, so equal ratios do not imply equal risk.
Premise 3 (High impact) — Inflation fell from about 9% to about 3% without a recession or 1979-style regime shift, which suggests inflation can recede as shocks fade and restrictive real rates work. The 66-month streak is also consistent with continuing supply and fiscal shocks, not with absent mean reversion.
Premise 5 (Medium impact) — If AI or capital deepening, spurred by labor scarcity, lifts productivity to 3% or more, or immigration rebounds, potential growth and the benchmark rise and 6.5% money growth looks much less excessive. This tension with the demographic premise is not addressed.
Conclusion (Medium impact) — The authors manage Treasury portfolios, are positioned in very short duration, and reversed a decades-long deflation thesis within months. A higher-inflation, higher-yield call aligns with that positioning, and the earlier failed forecasts raise doubts about the framework's reliability. This lowers the forecast's weight but does not refute it.
Conclusion (Low impact) — Forcing M2 growth down by roughly half would risk recession and financial stress, and the Fed controls interest rates, not M2. Whether a 3% money rule is a workable or desirable policy goal is not examined.

Suggested Improvements

Velocity sensitivity — Show how the benchmark changes under alternative velocity trends (for example flat, -1, +1, +2) and give an explicit uncertainty band instead of a ±0.25 range. The benchmark rests on this single judgment-based input, and a sensitivity analysis would show how much of the 'double the benchmark' claim survives.
Falsifiability and horizon — Specify the inflation index (core PCE or CPI), a time horizon, and the inflation path or M2 outcome that would count against the thesis, and bound the 'temporary factors' clause. This turns an accommodating forecast into a testable one and lets readers judge it fairly.
Causal attribution — Decompose recent inflation into monetary, fiscal, supply and tariff contributions, or soften the claim that the Fed's delay 'is why' inflation stayed high. The single-cause claim goes beyond what the evidence isolates, and competing drivers are prominent in current discourse.
Engagement with the 2022–24 episode — Explain why M2 contraction coincided with disinflation without recession and why the current M2 rebound should now produce inflation, for example through lags or velocity. This is the most obvious challenge to a mechanical money-inflation link and is currently unaddressed.
Historical analogy — Compare benchmark, velocity, inflation starting point and expectations in 1979 and today, and include episodes of high money growth with low inflation. A like-for-like comparison would replace rhetorical force with evidence, and a fair reference class would show the base rate of persistent inflation after excess money.
Productivity scenarios — Present a range of productivity paths, including an AI upside, and show the resulting benchmark under each. The point forecast is highly uncertain, and scenario ranges would show how robust the conclusion is.
Disclosure and track record — Explain the shift from the earlier deflation thesis and report the out-of-sample record of the benchmark framework. This helps readers judge whether the forecast is an independent analytical conclusion or an after-the-fact fit to positioning.

Scenario Tests

M2 growth stays near 6–7% for two more years while core inflation stays well above 2.5%. (Supports) — This would be consistent with the thesis, although fiscal, tariff and supply explanations would also fit, so it would not by itself prove the money mechanism.
M2 growth stays near 6.5% but inflation drifts down to about 2% as shelter and tariff effects fade. (Challenges) — The 'temporary factors' clause would be invoked to preserve the thesis, but a sustained decline over several years would show that M2 is not the controlling variable. This exposes the unfalsifiability problem.
Velocity falls (for example because higher rates pull money into deposits and money funds) while M2 grows 6.5%. (Challenges) — Nominal spending growth would not match money growth, the benchmark would effectively rise, and the excess-money inference would weaken or vanish.
AI-driven investment lifts aggregate productivity growth toward 3% or more. (Challenges) — Potential growth and the benchmark would rise by roughly the same amount, shrinking the excess of 6.5% money growth. P5 would be directly refuted.
The Fed tightens until M2 growth falls toward 3% and inflation converges to 2% with a lag. (Supports) — This would be consistent with a lagged money-inflation link, though the cost in growth and employment is not addressed, and rate-driven tightening would have its own effects on inflation.
Immigration policy shifts raise labor-force growth materially. (Challenges) — Potential growth and the absorbable money growth rate would rise, weakening P4. Because the demographic input is policy-sensitive, the benchmark is less stable than presented.
A recession or credit shock pushes inflation to 2% without any change in regime. (Challenges) — This would count against the claim that inflation needs an active, credible regime to fall. The authors' hedge might still be used to preserve the thesis.

Coherence & Relevance

The argument is a coherent convergent structure. The benchmark premise (P1) and the measured excess (P2) form the core, demographics and productivity (P4, P5) justify the low benchmark, and P3 supports the claim that no self-correction should be expected. The conclusion is appropriately hedged ('likely') and follows if the quantity-theory framework and A2 hold. But the premises are not independent confirmations; they share one framework and one source, and the benchmark bears most of the load. Its velocity input is unsupported, rival drivers of inflation are not addressed, and the 'temporary factors' clause limits testability. The reasoning is plausible but not well enough supported to warrant the confidence the numerical precision implies.

The non-inflationary rate of broad money growth today is about 3 percent a year (range 2.75 to 3.25 percent): roughly 2 percent potential real growth plus the 2 percent inflation target, less about 1 point for rising velocity. (Strong) — This is the foundation of the argument, and without it the 'double the benchmark' comparison cannot be made. The velocity term is asserted and not derived or supported, and the benchmark is partly stipulative because it includes the 2% target as an input.
Broad money growing at about 6.5 percent a year, roughly double the 3 percent benchmark, is a monetary excess similar in scale to 1979 and risks nominal spending outrunning productive capacity, keeping inflation and inflation expectations above target. (Strong) — This applies the benchmark to current data and bears directly on the conclusion. The gap is the step from monetary excess to realized inflation, which needs stable velocity and a transmission mechanism that the argument does not show. It also does not distinguish spendable money from portfolio balances.
Inflation does not reliably fall back to target on its own; it takes an active, credible monetary regime, and the Fed's delayed response in 2021 and 2022 is why inflation has stayed above 2 percent for 66 straight months. (Moderate) — This supports the 'no self-correction' clause of the conclusion. But it adds a causal claim about the Fed that the conclusion does not need and the evidence does not isolate, and the 'credible regime' concept is not operationalized.
Slowing population and labor-force growth (roughly zero now) lowers potential real growth and therefore the rate of money growth the economy can absorb without inflation. (Strong) — This directly supports the potential-growth input in P1. It does not show how the demographic and productivity premises combine to the 2% figure, and it ignores supply-side responses such as capital deepening.
Aggregate productivity growth is likely to stay near its long-run 2 percent pace despite AI, because a large government sector, a services-heavy job mix and lost economies of scale limit how far gains spread. (Moderate) — This supports the 2% potential-growth input and rebuts the main upside scenario. It rests on qualitative mechanisms and does not engage the strongest case for slow-then-fast AI diffusion.

View this argument on LogicFirst.ai