What if AI datacenter demand outstrips the power grid's supply?
Datacenter load outrunning supply is the core bull case: power scarcity validates the AI-capex spend, lifting NVDA and the gas complex (TTF/Henry Hub) as gas becomes the swing firm fuel. Rhymes with the 2024 PJM capacity-auction spike that priced data-center demand and lifted IPP/gas names. Skeptical: connection moratoriums also cap near-term GPU deployment, so the unambiguous winners are gas generators and grid copper, not just chips. The +0.9 ai_capex/NG map is sensible.
how we built this number — every step
The class rate is measured from our dated, sourced event library (decade-normalized Poisson — the full table is public at base_rates.json). The variant’s share within its class is the analyst’s editorial call, published so you can audit it. A wider range means thinner precedent. Full recipe: methodology · scored at Reality Check.
The butterfly cascade
How this trigger trickles across markets, left → right — the root shock, its first‑order moves, then the ripple effects. Drag any node; tap a market for its real price history.
Resolution timeline — how this probability is moving
Our model's odds (gold) over time vs the crowd's (Polymarket, blue), from the past toward the 6–18 months horizon. Each dot is a real macro event that nudged the probability — green pushed it up, red pushed it down. Tap a dot for the source. The gold path is an illustrative reconstruction anchored to today's estimate — real dated events, not a live re-estimate history.
What it would mean
If this plays out, it is a mixed shock. Datacenter electricity demand outstrips new supply in key US regions, forcing connection moratoriums and price spikes. The trigger decomposes into signed root‑shocks — Natural gas ▲ · AI capex ▲ — which propagate through our causal graph to the markets below.
If it happens — the markets it would move
Biggest moves first. Projected moves are cascade-model priors; hist A–B% = what comparable past events actually did (measured abnormal returns), and model prior · unmeasured marks markets with no analogue backing yet. Tap any market for its price history.
| Market | Class | Projected move | |
|---|---|---|---|
| 1 | Nvidia NVDAon Hyperliquid 📈 chart | Equity | ▲ +1.7% hist -1.77–+3.29% · other way -3.19% (n=12) |
| 2 | Broadcom AVGOon Hyperliquid 📈 chart | Equity | ▲ +1.3% hist -0.55–+1.88% · other way -0.43% (n=12) |
| 3 | Micron MUon Hyperliquid 📈 chart | Equity | ▲ +1.2% hist -0.04–+1.82% · other way +0.87% (n=12) |
| 4 | Semiconductors SMHon Hyperliquid 📈 chart | Equity | ▲ +0.8% hist +0.13–+0.93% · other way -0.28% (n=12) |
| 5 | Natural gas NGon Hyperliquid 📈 chart | Commodity | ▲ +0.5% hist -3.37–+2.32% · other way +4.24% (n=12) |
| 6 | AMD AMDon Hyperliquid 📈 chart | Equity | ▲ +0.6% hist -0.91–+0.99% · other way -0.96% (n=12) |
| 7 | TSMC TSMon Hyperliquid 📈 chart | Equity | ▲ +0.6% hist -1.13–+2.62% · other way -1.8% (n=12) |
| 8 | Marvell MRVLon Hyperliquid 📈 chart | Equity | ▲ +0.6% hist -0.74–+1.2% · other way +0.5% (n=12) |
| 9 | ASML ASMLon Hyperliquid 📈 chart | Equity | ▲ +0.5% hist +0.02–+0.7% · other way -2.54% (n=12) |
| 10 | Qualcomm QCOMon Hyperliquid 📈 chart | Equity | ▲ +0.4% hist -0.93–+2.04% · other way -1.82% (n=12) |
| 11 | Intel INTCon Hyperliquid 📈 chart | Equity | ▲ +0.3% hist -0.83–+4.42% · other way -4.71% (n=12) |
Probable recommendation
Why we may diverge from history
Trust the cascade long on ASML/AMD: the small negatives are chip-specific noise (Nvidia valuation, TSMC, CHIPS Act) — an AI datacenter power shock confirms the compute-demand thesis driving these names; off-channel analogues.
Historical precedent — what analogous events actually did
Across 14 analogous events (overlap‑weighted), as abnormal returns — market beta stripped, so it's the event's own effect, not the market backdrop. Shown at 20 days (persistent) and 5 days (immediate); ↺ fades = the two horizons disagree. Confidence = consistency × sample × significance.
| Asset | History says | Abnormal (20d · 5d) | Hit | n | Confidence | vs cascade |
|---|---|---|---|---|---|---|
| INTC INTC | LONG | +3.7% · 5d +1.2% | 87% | 14 | 0.53 | ✓ matches cascade |
| AVGO AVGO | SHORT | -1.3% · 5d -1.8% | 67% | 8 | 0.27 | ⚠ differs |
| NG NG | SHORT | -3.8% · 5d -1.8% | 61% | 14 | 0.21 | ⚠ differs |
| AMD AMD | SHORT | -1.3% · 5d -2.0% | 61% | 14 | 0.20 | ⚠ differs |
| TSM TSM | LONG | +2.3% · 5d -0.3% ↺ fades | 58% | 14 | 0.15 | ✓ matches cascade |
| Gold XAU | LONG | +0.6% · 5d -0.1% ↺ fades | 58% | 14 | 0.15 | · |
| SMH SMH | LONG | +0.4% · 5d +0.0% | 58% | 14 | 0.14 | ✓ matches cascade |
| QCOM QCOM | LONG | +1.8% · 5d -0.2% ↺ fades | 58% | 14 | 0.12 | ✓ matches cascade |
| Volatility VIX | SHORT | -3.2% · 5d -3.0% | 55% | 14 | 0.07 | · |
| NVDA NVDA | SHORT | -3.0% · 5d -5.8% | 42% | 14 | 0.00 | ⚠ differs |
| MU MU | LONG | +0.9% · 5d -3.0% ↺ fades | 48% | 14 | 0.00 | ✓ matches cascade |
| MRVL MRVL | SHORT | -1.2% · 5d -0.1% | 35% | 14 | 0.00 | ⚠ differs |
| ASML ASML | LONG | +0.3% · 5d -1.7% ↺ fades | 45% | 14 | 0.00 | ✓ matches cascade |
| US dollar DXY | SHORT | -0.0% · 5d -0.4% | 48% | 14 | 0.00 | · |
Why this probability
AI datacenter demand outstripping grid supply already happening; moratoriums and spikes near-certain trend. A base‑rate‑anchored prior, continuously scored against what actually happens — not a forecast.