Odd Lots2026.08.2947 min

Rethinking R-Star, AI Data Center Inflation, and Fed Reaction Functions

Original title · Austan Goolsbee Is Worried the Economy Is Overheating

用阿法問答深入這集 ✨

或自己問：

Evaluating monetary policy restrictiveness requires analyzing real rates alongside inflation trajectory rather than relying on abstract models. Chicago Fed President Austan Goolsbee signals concern over sticky service inflation and high AI capital expenditure that risks overheating the real economy.

Disentangling real rates from theoretical benchmark models

Policy restrictiveness cannot be evaluated using unobservable concepts like R-star, which function more like a "Sasquatch" than an actionable policy tool. With long-term landing targets around 3% nominal rates, 2% inflation, and 1% real rates, today's real rate remains tightly coupled to actual inflation dynamics. If inflation stays elevated near 3% or higher, current policy is substantially less restrictive than nominal yields suggest.

AI infrastructure investment creates real resource friction

Capital expenditure in AI data centers is moving beyond localized industry competition into direct demand for physical resources. Heavily bidding up land, construction labor, and electrical contractors creates localized price pressures that threaten to spill over into aggregate economic demand. When heavy business investment pushes output above potential, monetary policy risks becoming insufficiently restrictive, regardless of theoretical neutral rate calculations.

Shift from explicit forward guidance toward transparent reaction functions

Central banks are moving away from rigid forward guidance and explicit dot-plot forecasts toward clearer communication of their underlying reaction functions. Binding forward commitments introduce unnecessary market volatility and tie policymakers' hands during supply-shock environments. For traders, tracking how central bankers weigh sticky service inflation versus temporary commodity shocks offers a far clearer signal than multi-year interest rate projections.

Evaluating whether inflation persistence is driven by structural demand or transitory supply shocks remains the primary signal determining future rate adjustments.

Key questions

Why does Goolsbee argue against using theoretical benchmarks like R-star?→

Goolsbee contends that real rates—nominal rates adjusted for actual and expected inflation—are the only practical metric. Focusing on unobservable benchmarks like R-star obscures the true restrictiveness of policy when inflation remains sticky.

💡 詢問阿法，看詳細逐字稿分析→

How does AI infrastructure spending threaten to overheat the economy?→

Massive AI-related capital expenditure acts as an aggregate demand shock. If this investment exceeds the economy's potential output before productivity gains materialize, it risks triggering broader wage and price inflation across adjacent sectors.

💡 詢問阿法，看詳細逐字稿分析→

Why is the Federal Reserve moving away from explicit forward guidance?→

Forward guidance is best suited for zero-bound interest rate environments. In a normalized regime, binding commitments increase market volatility and threaten central bank credibility when macro forecasts deviate from actual economic data.

💡 詢問阿法，看詳細逐字稿分析→

Further research

Tickers and signals often linked to this episode's themes in public sources · AI-compiled, not investment advice

AI Infrastructure Supply Bottlenecks

Accelerating AI infrastructure spending is encountering severe physical shortages in high-voltage transformers, power grid interconnections, and specialized labor, creating cost inflation for developers while enriching equipment suppliers.

US stocks

ETN
EatonBenefitsEaton produces critical electrical switchgear and distribution transformers needed for data center power architectures, giving it strong pricing power and expanding multi-year backlogs amid equipment shortages.
GEV
GE VernovaBenefitsGE Vernova manufactures high-voltage grid transformers and power generation equipment required for substation upgrades, directly benefiting from multi-year supply shortages and utility backlog expansion.
PWR
Quanta ServicesBenefitsQuanta Services supplies specialized engineering and construction labor to build high-voltage transmission lines and substation interconnections for power-constrained data centers.
VRT
VertivBenefitsVertiv supplies high-density power distribution and liquid cooling equipment to data centers, capturing elevated demand and pricing power as rack power requirements surge.
DLR
Digital Realty TrustPressuredDigital Realty Trust faces escalating equipment costs and multi-year power grid interconnect delays that threaten to push back data center energization schedules.

Risks

A deceleration in hyperscaler CapEx spending or breakthroughs in energy-efficient AI model architectures could relieve grid capacity constraints and reduce equipment supplier pricing power.

Watch list

High-voltage transformer lead times and backlog figures reported by electrical equipment suppliers
Utility grid interconnection queue processing times across major data center hubs like Northern Virginia and Texas
Hyperscaler quarterly CapEx updates and commentary on data center energization timelines
Announcements of behind-the-meter on-site power generation and power purchase agreements (PPAs)

This section is AI-compiled from public sources, may be inaccurate or outdated, is for research reference only, and is not investment advice.

Want the full episode?