What if AI demand stalls and leaves data centers overbuilt?
If AI demand stalls and hyperscalers cancel leases, overbuilt data-center REITs face stranded-asset writedowns and the capex bellwethers (NVDA, AVGO, Micron) lead lower as the build-out thesis cracks. Rhymes with the Jan-2025 DeepSeek selloff that briefly torched AI-capex names, and the 2000-01 telecom dark-fiber glut. Forward angle: power-purchase and lease economics are the tell — watch hyperscaler capex guidance; the REIT/stranded-power channel is the novel leg vs. a pure semi cycle.
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 1–3 years 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 risk-off shock. AI demand stalls, hyperscalers cancel leases, and overbuilt data-center REITs face stranded-asset writedowns and falling power-purchase values. The trigger decomposes into signed root‑shocks — AI capex ▼ · Growth surprise ▼ · Risk appetite ▼ — 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 | ▼ -2.6% hist -2.42–+0.82% · other way +4.19% (n=12) |
| 2 | Broadcom AVGOon Hyperliquid 📈 chart | Equity | ▼ -2.0% hist -1.87–+1.19% · other way +5.73% (n=12) |
| 3 | Micron MUon Hyperliquid 📈 chart | Equity | ▼ -1.9% hist -3.87–+0.22% · other way -1.63% (n=12) |
| 4 | Semiconductors SMHon Hyperliquid 📈 chart | Equity | ▼ -1.5% hist -1.01–+0.64% · other way +1.37% (n=12) |
| 5 | Solana SOLon Hyperliquid 📈 chart | Crypto | ▼ -1.4% hist -10.88–+1.35% · other way -17.09% (n=11) |
| 6 | MicroStrategy MSTRon Hyperliquid 📈 chart | Equity | ▼ -1.4% hist -1.91–+2.62% · other way +10.05% (n=12) |
| 7 | Nasdaq 100 NDXon Hyperliquid 📈 chart | Index | ▼ -1.1% hist -0.76–-0.34% · other way +0.03% (n=12) |
| 8 | Hyperliquid (HYPE) HYPEon Hyperliquid | Crypto | ▼ -1.1% model prior · unmeasured |
| 9 | Ether ETHon Hyperliquid 📈 chart | Crypto | ▼ -1.0% hist -9.56–+2.15% · other way +0.11% (n=11) |
| 10 | AMD AMDon Hyperliquid 📈 chart | Equity | ▼ -1.0% hist -1.02–+0.48% · other way -3.31% (n=12) |
| 11 | TSMC TSMon Hyperliquid 📈 chart | Equity | ▼ -1.0% hist -1.13–+0.82% · other way +1.86% (n=12) |
| 12 | Marvell MRVLon Hyperliquid 📈 chart | Equity | ▼ -1.0% hist -1.06–+1.29% · other way +0.78% (n=12) |
| 13 | Bitcoin BTCon Hyperliquid 📈 chart | Crypto | ▼ -0.8% hist -4.17–+1.28% · other way +3.62% (n=12) |
| 14 | ASML ASMLon Hyperliquid 📈 chart | Equity | ▼ -0.9% hist -2.25–+0.53% · other way -0.21% (n=12) |
Probable recommendation
Why we may diverge from history
Trust the cascade's SHORT on XLF: the +2.9% history is anchored to the 2000 dot-com window (+23%) and isolated chip-specific dips, none of which test financials' exposure to a multi-year data-center REIT stranded-asset write-down.
Historical precedent — what analogous events actually did
Across 40 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 |
|---|---|---|---|---|---|---|
| SOL SOL | SHORT | -8.6% · 5d -5.8% | 73% | 37 | 0.36 | ✓ matches cascade |
| MU MU | SHORT | -2.7% · 5d -2.9% | 69% | 40 | 0.29 | ✓ matches cascade |
| ETH ETH | SHORT | -8.1% · 5d -5.0% | 66% | 38 | 0.25 | ✓ matches cascade |
| Gold XAU | LONG | +1.1% · 5d -0.1% ↺ fades | 65% | 40 | 0.25 | ✓ matches cascade |
| JPM JPM | LONG | +1.3% · 5d +0.3% | 65% | 40 | 0.23 | ⚠ differs |
| INTC INTC | SHORT | -4.7% · 5d -3.1% | 61% | 40 | 0.22 | ✓ matches cascade |
| SMH SMH | LONG | +1.2% · 5d -0.3% ↺ fades | 59% | 40 | 0.17 | ⚠ differs |
| XLF XLF | LONG | +0.3% · 5d +0.1% | 59% | 40 | 0.16 | ⚠ differs |
| US dollar DXY | SHORT | -0.2% · 5d +0.0% ↺ fades | 59% | 40 | 0.16 | · |
| MRVL MRVL | LONG | +1.7% · 5d -1.4% ↺ fades | 57% | 40 | 0.12 | ⚠ differs |
| MSTR MSTR | LONG | +3.2% · 5d -0.5% ↺ fades | 55% | 40 | 0.10 | ⚠ differs |
| Bitcoin BTC | SHORT | -3.5% · 5d -2.3% | 56% | 39 | 0.10 | ✓ matches cascade |
| QCOM QCOM | SHORT | -2.4% · 5d -2.4% | 57% | 40 | 0.10 | ✓ matches cascade |
| Volatility VIX | LONG | +1.0% · 5d +1.1% | 55% | 40 | 0.08 | ✓ matches cascade |
Why this probability
AI capex still strong mid-2026; demand stall and lease cancellations a real 1-3yr risk. A base‑rate‑anchored prior, continuously scored against what actually happens — not a forecast.