AI Cycle
Indicator
Tracking where excess is building in the AI trade — across valuations, capex, memory, power, and market structure.
💡 The Core Thesis
The AI companies are cheap. The trades built around them are not. NVDA trades at 26× forward earnings with a PEG of 0.67 — Cisco peaked at 200–472× P/E. NVDA generated $96.7B in free cash flow on 71% gross margins. The risk isn’t in AI itself — it’s in the capex supply chain, the memory cycle, and the structural concentration. This indicator tracks where excess is building, so you can see it before the consensus does.
AI Cycle Composite
41
WARM
COOL (0–30)
WARM (30–55)
HEATED (55–75)
FROTHY (75–100)
Updated quarterly after earnings + monthly for market structure signals
1. Valuation Health
Weight: 25% · The NVDA vs Cisco reality check
HEALTHY
Forward P/E vs Growth
NVDA 26× fwd on 66% EPS growth
PEG 0.67 — undervalued on growth
Cisco at peak: 200–472× P/E with ~$5B FCF
Free Cash Flow
NVDA: $96.7B FCF · AVGO: $27B/yr
Cash machines, not cash burners
71% gross margins · $120B net income
Price vs EPS Growth Gap
NVDA price +64% 1yr vs EPS +67% 1yr
Price tracking earnings — no gap
Caution signal if price outpaces EPS by 40%+
2. Capex Conversion
Weight: 25% · $725B in spending — is the revenue following?
WATCH
Capex-to-AI Revenue
$725B capex vs ~$100B AI revenue
7.25:1 ratio — $7 per $1 of revenue
Track quarterly — improving or worsening?
Hyperscaler Borrowing
2026: ~$400B borrowed vs $165B in 2025
2.4× increase — telecom 2000 echo
Source: Morgan Stanley credit analysis
AI Revenue Growth Rate
MSFT AI: $37B ARR (+123% YoY)
Revenue IS growing fast
GOOG Cloud +63% YoY · AWS +28%
3. Memory & Semi Cycle
Weight: 15% · DRAM is the canary in the coal mine
LATE CYCLE
DRAM Contract vs Spot
Contracts: +58–63% QoQ in Q2
Spot DDR5 falling — leads by 1–2Q
Source: TrendForce · Q1 was +90–95%
SOX vs 200-DMA Spread
50%+ above 200-day moving average
Last seen: March 2000
Technical stretch, not valuation stretch
HBM Demand Dynamics
1 bit HBM = 3 bits conventional DRAM
All 3 majors expanding capacity
Analyst view: 2028–29 oversupply risk
4. Power & Infrastructure
Weight: 20% · Real demand with a multi-decade buildout ahead
EARLY INNINGS
Power Demand Reality
US grid +2–2.5% CAGR · AI data centers need 35–50 GW by 2030
Structural demand, not speculation
First grid growth in 20 years — buildout barely started
GEV Backlog & Pipeline
$163B backlog (Q1 2026) · Gas turbines sold out to 2029
Backlog supports earnings growth for years
Revenue visibility rare in industrials — this is it
Grid Capex Cycle
$1.4T+ utility investment through 2030 · T&D upgrades just starting
Multi-decade capex cycle with bipartisan support
IRA + AI demand + electrification = infrastructure supercycle
5. Market Structure
Weight: 15% · Concentration exceeds dot-com levels
ELEVATED
Mag 7 Concentration
33% of S&P 500 market weight
Dot-com peak was 27% — we exceed it
Structural fragility, not valuation excess
AI Leadership Breadth
NVDA +19% YTD · AMD ~+100% · SOXX +45%
Leadership rotating within the theme
Track # of AI names making new highs
Options Euphoria Gauge
Call/put skew on NVDA, SMH, SOXX
Sentiment gauge, not valuation gauge
Track retail vs institutional positioning
Where the Risk Actually Lives
CHEAP ON EARNINGS
NVDA · AVGO · QCOM
AI SaaS leaders
→
CAPEX UNCERTAINTY
$725B spend
$400B borrowed
→
LATE CYCLE SIGNALS
DRAM spot falling
SOX stretched
→
REAL BUILDOUT
Power & grid infra
Multi-decade demand
→
FRAGILITY
33% concentration
Narrow breadth
▲ Signals That Push Toward HEATED
DRAM spot inverts contracts
Capex-to-revenue ratio widens
Grid permitting bottlenecks stall buildout
Hyperscaler capex cuts announced
AI revenue growth decelerates
▼ Signals That Confirm COOL
AI revenue growth accelerates
NVDA PEG stays below 1.0
Market breadth expands
Vendor financing stabilizes
Grid buildout timelines accelerate
Important Disclaimer
M3I Research is for informational and educational purposes only. Nothing on this page constitutes financial, investment, legal, or tax advice. The composite score, layer assessments, and signal classifications are outputs of a proprietary research framework and should not be relied upon as the sole basis for any investment decision. References to historical patterns and valuations describe past observations and do not guarantee future results. Always consult a qualified financial advisor before making investment decisions.