Nomura: AI-driven productivity gains should not be simply viewed as a disinflationary force
AI summary card
Nomura: AI-driven productivity gains should not be simply viewed as a disinflationary force
The report argues that the relationship between productivity and inflation is unstable both theoretically and historically, and that near-term demand shocks from AI capital expenditure could even push inflation higher, while still potentially supporting the Fed in staying on hold rather than hiking preemptively.
- There are already early signs that the spread of AI is boosting productivity, but its transmission to inflation and interest rates remains unclear.
- In the short term, the AI-related capex boom is pushing up prices of technology components and creating inflationary pressure in areas such as consumer electronics.
- Historical data show that the relationships between trend productivity and CPI, economic slack, and unit labor costs are unstable and have weakened significantly in recent decades.
- The report argues that strong productivity growth can make policymakers more tolerant of strong output and wage signals, but it cannot be used to directly predict additional disinflation.
- Nomura interprets Warsh's emphasis on productivity and supply expansion as an important signal of his dovish inclination.
Report interpretation
Overview
Nomura analyzes whether AI-driven productivity gains will lead to lower inflation. The core conclusion is that productivity is not inherently a disinflationary force: in theory, it may lower prices through cost reductions, but it may also raise aggregate demand through higher real income and incentives for corporate investment. The current AI capex boom has already put upward pressure on prices of technology equipment and related consumer goods, so expected future productivity gains cannot simply offset the demand-driven inflation that is already emerging.
Core views
The report's core views include: first, AI and technological progress may bring productivity gains, but they do not create a stable and predictable mechanism for lower inflation; second, productivity and inflation have at times been negatively correlated historically, but the relationship has been inconsistent and has clearly broken down since the global financial crisis; third, the unit labor cost framework cannot mechanically show that higher productivity must reduce inflation, because wages adjust upward with productivity and profit margins may also absorb part of the cost change; fourth, for the Fed, strong productivity growth is better used as a real-time reference for tolerating strong growth and wage strength, rather than as a sufficient reason to forecast future disinflation; fifth, Warsh's emphasis on the supply-side effects of productivity suggests he is unlikely to support preemptive rate hikes.
Analysis framework
The report evaluates the relationship between productivity and inflation from seven angles: theoretical mechanisms, historical correlations, real-world evidence from AI capital expenditure, economic slack, unit labor costs, macro inflation models, and the risk of policy misjudgment, and maps these analyses to implications for Federal Reserve policy.
Methodology notes
Productivity gains affect both supply and demand
Productivity gains may lower costs in individual industries and reduce prices, but they may also increase real income, stimulate corporate capital expenditure, and raise aggregate demand, so the overall direction of inflation depends on the policy regime, financial conditions, and demand response.
Unit labor cost equals the relationship of wages relative to labor productivity
Although unit labor costs have a strong contemporaneous correlation with inflation, productivity is an important long-term driver of wages and profit margins may also adjust, so this framework has limited value for forecasting future inflation.
Conditional interpretation of productivity variables in inflation models
Some inflation models assign a negative coefficient to productivity after controlling for demand indicators, but this only reflects a partial relationship under given demand conditions; if productivity is assessed from a forward-looking perspective, its effect on demand itself should also be considered.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- U.S. ratesThe productivity narrative affects the Fed's policy reaction function
- Strengths
- If productivity improvement is confirmed, policymakers may be more tolerant of strong growth and wage data.
- Weaknesses
- Productivity itself cannot reliably predict disinflation and cannot independently support a dovish view.
- Comparison
- Compared with viewing AI as a disinflationary shock, Nomura places greater emphasis on the inflation risks from demand and capital expenditure channels.
- Risks
- If policy places excessive faith in future supply expansion, it may underestimate current demand-driven inflation pressure.
- Technology capex chainAI investment demand is a source of near-term inflation pressure
- Strengths
- Demand for AI equipment is strong, companies are willing to increase capital expenditure, and related industry chains are supported by favorable conditions.
- Weaknesses
- Equipment supply and price constraints may raise investment costs and weaken actual investment efficiency.
- Comparison
- The report argues that current evidence is more supportive of AI investment being inflationary rather than immediately generating broad disinflation.
- Risks
- If capex is brought forward excessively or financial conditions change, the pace of demand may fluctuate.
- U.S. dollar and risk assetsAffected by the inflation path and expectations for Fed policy
- Strengths
- If inflation momentum peaks and policy remains on hold, risk appetite may receive temporary support.
- Weaknesses
- AI investment pushing inflation higher will limit room for faster easing.
- Comparison
- Compared with a pure disinflation narrative, this report is more cautious on the scope for policy easing.
- Risks
- If core inflation reaccelerates, market pricing for staying on hold or cutting rates may be revised.
Key data
- Report date2026-07-28Production Complete: 2026-07-28 18:55 UTC.
- Research institutionNomura Securities International, Inc.The report was issued by Nomura's U.S. entity.
- Research topicAI-driven productivity and U.S. inflation and rate policyThe title is "US: Will AI-driven productivity lead to disinflation?".
- Short-term inflation viewAI capital expenditure may be inflationaryThe report states that the AI capex boom is pushing up prices of technology components and creating inflationary pressure in consumer electronics.
- Historical evidenceThe negative correlation between productivity and inflation is unstableThe report says this relationship has performed poorly in recent decades and was especially ineffective in the post-financial-crisis period.
- Policy implicationStrong productivity growth can increase tolerance for strong growth and wagesThis differs from assuming that future productivity growth will directly bring additional disinflation.
Impact & implications
For asset markets and policy expectations, the report weakens the one-dimensional narrative that the AI productivity dividend will automatically lower inflation and support faster rate cuts. If AI investment continues to expand and financial conditions remain loose, it may strengthen near-term price pressure in equipment, technology components, and some consumer electronics. On the policy side, Warsh's belief in the supply-side effects of productivity may incline him to stay on hold and tolerate stronger growth signals, but Nomura does not see this as a basis for preemptive rate hikes.
Risks
- There is uncertainty around the magnitude, timing, and sectoral diffusion path of AI-driven productivity gains.
- AI capital expenditure may continue to push up prices of technology equipment and related consumer goods.
- The historical relationships among productivity, inflation, economic slack, and unit labor costs are unstable.
- The Federal Reserve may misjudge productivity trends or demand strength, resulting in policy errors.
- If the recent slowdown in inflation mainly reflects weaker tariff pass-through and seasonal factors, its durability still needs to be verified.
What to watch
- Whether AI-related capital expenditure continues to expand and spread to other industries.
- Whether prices of technology equipment, components, and consumer electronics continue to rise.
- The relative changes between U.S. trend labor productivity and real wage growth.
- Whether the transmission among unit labor costs, profit margins, and core inflation strengthens.
- Changes in how FOMC minutes and officials' remarks describe the inflationary effects of AI investment and the disinflationary effects of productivity.
- Whether Warsh continues to emphasize productivity and supply growth and thereby influence his policy stance.