Thoughts on the Market
2026-07-03 · Hosted by Mike Wilson · Morgan Stanley
Executive Summary
Morgan Stanley’s Serena Tang debriefed two weeks of investor meetings across Europe and Asia, where two themes dominated: the Fed’s policy path and AI CapEx. Morgan Stanley argues the hikes implied by the June SEP should be read with caution and expects the Fed to remain on hold through 2026 on a lower core-inflation path. On AI, the firm sees “chipflation” (memory prices up sixfold in a year) as more likely to reprice and ration AI infrastructure than derail the cycle, and forecasts massive AI-related debt issuance — supporting a constructive stance on risk assets.
Key Stories & Changes
1. Fed Policy: Hold Through 2026
Many investors interpreted Chair Kevin Warsh’s first (June) FOMC as “unambiguously hawkish,” with the SEP implying a hike in 2026 and pushed-out cuts
Morgan Stanley economists urge caution: the projections appear conditioned on elevated near-term inflation and may not capture disinflation from a Strait reopening
MS anticipates a lower core-inflation path (reversal in travel-related inflation + tariff payback), underpinning its call that the Fed stays on hold through 2026
2. AI: “Chipflation” and the CapEx Cycle
Investors “firmly” believe in the ongoing AI-CapEx cycle but feel unease as AI becomes an inflation story (macro) and a funding story (micro)
“Chipflation” is the new buzzword — memory prices up sixfold over the past year
MS view: chipflation is more likely to reprice and ration AI infrastructure than derail the cycle; demand scales across three layers (more memory per chip, more chips per system, more systems per cluster), with hyperscalers first in the allocation queue
Key risk flagged: CapEx efficiency, as memory becomes a larger share of AI system cost
3. AI Funding & Bond Issuance
The majority of corporate bond issuance year-to-date has funded data-center construction
Hyperscalers broadened their investor base via ~$25B of non-dollar debt (Euro, Swiss Franc, JPY) in May
Credit strategists forecast nearly $600B more AI-related global issuance in 2026; for US IG corporates alone, $1 trillion of net issuance — a reason MS expects the asset class to underperform
Hyperscaler cash CapEx projected to surpass $1 trillion in 2027, accelerating issuance
Trends Identified
1. Rates and AI as the Twin Investor Obsessions
Every investor meeting across regions returned to the same two uncertainties — the Fed’s path and AI CapEx — which are increasingly intertwined, as AI spending itself becomes an inflation input that shapes the rate outlook.
2. Chipflation as a Rationing Mechanism, Not a Cycle-Killer
Rather than viewing surging memory prices as a threat that ends the AI buildout, Morgan Stanley frames chipflation as a repricing force that rations infrastructure toward hyperscalers first, keeping the cycle intact while raising CapEx-efficiency risk.
3. AI’s Growing Footprint in Credit Markets
The scale of AI-related debt issuance — a trillion dollars of US IG net issuance and $600B globally in 2026 — positions data-center funding as a dominant driver of corporate bond supply and a reason to expect IG underperformance. —-
Sentiment Analysis
Overall Market Sentiment: Constructive
Despite investor unease around the Fed and chipflation, Morgan Stanley’s base case (lower inflation, Fed on hold, intact AI cycle) supports staying constructive on risk assets.
Risk Factors Highlighted
CapEx efficiency: Memory becoming a larger share of AI system cost pressures the economics of the buildout.
Chipflation: Memory prices up sixfold could ration AI infrastructure and squeeze non-hyperscaler buyers.
Hawkish Fed misread: If elevated near-term inflation persists, the June SEP’s implied 2026 hike could materialize.
IG credit oversupply: ~$1 trillion of US IG net issuance (much AI-related) is a headwind for the asset class.
AI funding acceleration: Hyperscaler cash CapEx surpassing $1 trillion in 2027 escalates issuance needs.
This episode was covered in today’s The Market Signal — 2026-07-03, a cross-source synthesis of multiple podcast reports.