Supply-shock inflation persistence: why central banks stay tighter than demand suggests

One-line summary: A sequence of compounding negative supply shocks (Covid → Ukraine → trade war / Brexit → Iran war) has shifted household and firm price-setting behavior in a way that creates persistent inflation pressure despite weak demand. Central banks therefore stay tighter on policy than the demand picture alone would suggest. Implication for this project's chains: real financing costs for AI capex stay elevated, commodity-side inflation stays sticky, and the "rate-cut to relieve buildout pressure" path is less available than the consensus prices.

The framing

Three behavioral shifts compound shock-by-shock:

1. Household attention threshold has moved down. People used to ignore inflation below 4%; it now matters to households at 3–3.5%, and they are more attentive to upside surprises than downside. Once inflation crosses the attention threshold, expectations become harder to anchor.
2. Firms have shifted from schedule-based to state-dependent pricing. Pre-pandemic, prices reset on calendar cadence regardless of inflation. Post-pandemic, firms reset when inflation moves, which means second-round pass-through happens faster.
3. Second-round effects on wages. Earlier shocks (Covid, Ukraine) left wage-growth expectations stuck above what historical relationships would predict, even before the next shock landed. Each new shock pushes wage-growth expectations up rather than starting from a clean slate.

The net result: a series of "transitory" shocks that should each fade behaves more like a single persistent inflation regime, because the behavioral substrate has changed.

Why it matters to stock-market

• AI capex chain. ai-capex-to-power-and-materials-cascade runs on $725B–$830B of hyperscaler 2026 capex financed at the prevailing real rate. If central banks stay tighter for longer, the cost-of-capital input to those capex models stays higher than the consensus prices, and the marginal data-center build becomes more sensitive to financing costs. Beneficiaries with structural cash flow (e.g. EWS via SpaceX, hyperscalers with already-funded capex) widen their lead vs. dependent challengers.
• Commodity chains. Iran-war-driven oil supply dislocations are a current supply shock layered on top of the existing sequence — the inflation impulse from oil pricing flows into headline CPI even when core stays subdued, and the household-attention threshold makes that politically salient. copper-supercycle-ai-data-centers and us-critical-mineral-independence both have a commodity-price input that is now harder to fade.
• Tariff regime. Trade-war / Brexit-style supply shocks are explicitly cited as part of the compounding sequence — i.e., tariffs are not just political artifacts, they re-enter the chain as another negative supply shock that pushes the inflation-persistence dial up. us-industrial-policy-tariff-shield takes the political-economy view; this page takes the macro-monetary-policy view of the same phenomenon.

Evidence

• From 2026-05-11-odd-lots-the-bank-of-englands-megan-greene-on-monetary-policy (Megan Greene, BoE MPC, on threshold behavior): "we've done a lot of research showing that households and businesses are possibly just more attentive to inflation and, particularly once inflation comes within a certain band. So there's a threshold. It used to be no one paid attention to inflation. If it was below 4%, that band has actually shifted down. So in the UK, we think that if inflation is somewhere between 3.3.5%, that's the threshold at which people actually notice it a lot more. So if you then have another negative supply shock and inflation go up, people will be much more sensitive to that. We also have done research showing that people are more sensitive to upside surprises in inflation than downside surprises in inflation."

• From 2026-05-11-odd-lots-the-bank-of-englands-megan-greene-on-monetary-policy (Greene, on firm-side pricing shift): "for firms, when they're looking at inflation. It used to be before we had these supply shocks that firms kind of set prices on a schedule and since the pandemic they've shifted much more to state dependent pricing. So when inflation gets higher they set prices more often and pass through their higher costs through the end user in the form of higher prices and that's remained."

• From 2026-05-11-odd-lots-the-bank-of-englands-megan-greene-on-monetary-policy (Greene, on second-round persistence pre-Iran): "I was already worried about some of this inflation persistence and some of the second round effects from the last couple of negative supply shocks even before Iran was invaded. And of course, since then, we've now had a negative supply shock, an energy shock, and that stands to push inflation up and growth down, which is a terrible situation for a central banker to be in."

• From 2026-05-11-odd-lots-the-bank-of-englands-megan-greene-on-monetary-policy (Greene, on UK posture despite weak growth): UK economy at 0.2% quarterly growth, "supply side is also incredibly weak" — even with weak demand, she voted to hold rates because supply weakness makes any demand inflationary. Stagflationary asymmetry in the policy choice.

• From 2026-05-11-odd-lots-the-bank-of-englands-megan-greene-on-monetary-policy (Christopher Waller framing cited by host Tracy Alloway): "Governor Christopher Waller in the US, he had a good speech recently called one transitory shock after another" — the language convergence at the Fed level on supply-shock compounding matters for the US rate path.

• From 2026-05-14-odd-lots-martin-wolf-on-the-terrifying-superpower-the-us-wields (Wolf, on the Iran-oil-prices dimension feeding the same sequence): Wolf frames the strange market response — oil prices rising on Iran war but equities at record levels — as a tension between commodity-side inflation impulse and the "great deal of ruin in the world economy" being absorbed by growth. The wiki should track whether equity strength holds when the second-round effects Greene describes start landing in 2026 H2 CPI prints.

• javier-blas in 2026-05-18-odd-lots-why-the-price-of-oil-beef-electricity-and (the transport-cost transmission channel of the 2026 oil shock): "Oil is very important because it permeates everywhere... if we have a food price in the next few months, it's going to be more because everything is more expensive to transport." This specifies how the 2026 shock transmits — via transport-cost pass-through into food and goods, not via the direct electricity-bill destruction that made 2022 so acutely inflationary. See energy-shock-2026-vs-2022 for the full 2026-vs-2022 distinction. The transport channel is slower and more diffuse than the 2022 power-bill channel, which is one reason the headline impulse so far is milder than the IEA's "worse than 1973, 79 and 2022 combined" warning.

• jack-farley in 2026-05-29-podcast-forward-guidance-how-to-trade-the-ai-productivity-boom-weekly (Taylor rule prescribes hiking; consumer income turning negative): "if you just look at something like a Taylor rule function here...Fed funds rate is below policy rule prescriptions up to the last round of insurance cuts. Even with Fed's R Star estimate...it would be a large gap under DB's R star estimate. So basically, if you follow these Taylor world traditional models, we should be hiking here." Simultaneous data: "personal incomes are trending negative...incomes, they're accelerating to the downside now...if you just look at the other previous historical moments where that also occurred, it's during times of either growth shocks or outright recessions." — Corroborates the stagflationary bind: Taylor rule says tighten, but consumer income data says recession-adjacent. The two signals are pulling in opposite directions, exactly the dynamic Greene identified for BoE.

• jim-bianco in 2026-05-28-podcast-macro-voices-macrovoices-534-dr-pippa-malmgren-superpower-war (Fed impossible position under oil-driven CPI above 4%): "the market is not there for the Fed to cut rates and it could actually backfire on them...most of the estimates are that the CPI is going to be over 4% when the main numbers come in. If it's already at 4% and you're stimulated, you're okay with 4%. I'm not okay owning your bonds." — Bianco's mechanism: political pressure to cut (White House vs. new Fed chair Kevin Warsh) → if Fed cuts while CPI >4% → bond market punishes (yields rise further), real rates compress, inflationary spiral in oil scenario. The "fix is more supply, not financial engineering" formulation complements Megan Greene's supply-shock-inflation-persistence framework: both identify that demand-side stimulus (rate cuts, subsidies) backfires on supply-side inflation shocks. New Powell (Warsh) dynamic: Bianco notes Warsh's "last public speech was last year" where he wanted rate cuts — but market is pricing hikes. The new Fed chair's policy view is the key unknown.

Contradictions / tensions

• Strong equity markets vs. tightening monetary regime. Wolf's "stunning and utterly strange" observation: equities at record levels while oil rises and supply shocks compound. Either the market is pricing through the shocks correctly (transitory after all), or it's mispricing the persistence Greene describes. Resolved by 2026 H2-H3 CPI prints — if core inflation stays sticky above 2.5%, the central-bank-stays-tight thesis is validated; if it fades fast, the market call was right.
• Greene's UK-specific view vs. universalizability. Her behavioral threshold numbers (3–3.5% in UK) are local research; the US equivalent may be different. The qualitative behavior (state-dependent firm pricing, asymmetric household attention) is more likely to generalize than the specific numbers.
• Oil shocks historically make the Fed LESS likely to hike (Chisholm, Fidelity quantitative data). denise-chisholm in 2026-05-29-podcast-the-compound-and-friends-what-if-it-s-still-early-with-denise-chisholm: "There's a negative pattern. The bigger the shock in oil, the lower the probability that the Fed hike...even when oil is up 60, what you see is only a 25% chance that the Federal Reserve hikes. Why? Because most of the time since the 1980s, you don't see a pass through to core inflation. And by most I mean 48%." This is a direct quantitative challenge to the thesis's central-bank-stays-tighter narrative as applied to the current 2026 oil shock specifically. Chisholm's mechanism: oil "acts like a tax hike on the consumer" — central banks recognize they can't fight a supply shock with demand tools; historically, the bigger the oil spike, the less likely the Fed is to respond with hikes. If this pattern holds in 2026, the central-bank-stays-tight thesis applies to tariff/structural shocks but not oil shocks. Corroborating: core CPI ex-shelter = 2.3% (Chisholm data, May 2026) — shelter is acting as a supply shock (locked-in mortgages, not demand pressure), further undercutting the case for tightening.
• Counter-data point. Sax in 2026-05-08-all-in-podcast-elon-s-anthropic-deal-the-next-ai-monopoly makes the opposite micro-argument for AI economics: AI is deflationary at the unit level because "coding tokens basically 10xs or 100xs, the value of that market." If unit-level AI deflation outpaces commodity-side supply-shock inflation in aggregate, the macro picture flips. The wiki holds both: AI deflation (Sax) and supply-shock inflation persistence (Greene) are not mutually exclusive — different baskets, different time horizons — but the policy regime is set by the central bank's read of headline + core.

Open questions

• What's the US-equivalent attention threshold (Greene gives 3–3.5% for UK)? Is the Fed's behavior consistent with a similar attention-threshold model?
• Does Greene's state-dependent-pricing shift fade as the post-Covid generation of price-setters cycles out, or is it now structural?
• Does this regime privilege real-asset equity (commodities, industrials, real estate) over financial-asset equity (long-duration tech) in a way that's stable enough to position around?

Related

• ai-capex-to-power-and-materials-cascade — financing cost input
• copper-supercycle-ai-data-centers — commodity-side inflation flow-through
• us-critical-mineral-independence — supply-shock geography
• us-industrial-policy-tariff-shield — political-economy view of the same shock sequence
• energy-shock-2026-vs-2022 — the transport-vs-power transmission distinction for the current oil shock