What if the US bars exports of frontier AI model weights?
Capping frontier model-weight and training-compute exports above a FLOP threshold to tier-three nations directly dents Nvidia/Broadcom data-center TAM at the margin and pressures the AI-capex trade. Rhymes with the Jan-2025 AI Diffusion Rule that hit semis on TAM-cap fears before partial rollback. The targets are Gulf/SEA compute buyers routing to China; the forward angle is weight-export limits are hard to enforce on open models, so the rule may be more bark than bite - the ai_capex -0.4 root is the right primary driver.
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 0–6 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 risk-off shock. US bars export of frontier model weights and training compute above a FLOP threshold to tier-three nations. The trigger decomposes into signed root‑shocks — AI capex ▼ · Trade tension ▲ — 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 | ▼ -2.4% hist -1.65–-0.62% · other way +4.19% (n=12) |
| 2 | Semiconductors SMHon Hyperliquid 📈 chart | Equity | ▼ -2.4% hist -1.42–-0.85% · other way +1.37% (n=12) |
| 3 | Nasdaq 100 NDXon Hyperliquid 📈 chart | Index | ▼ -2.1% hist -1.47–-0.42% · other way +0.03% (n=12) |
| 4 | Broadcom AVGOon Hyperliquid 📈 chart | Equity | ▼ -1.9% hist -1.44–+1.0% · other way +5.73% (n=12) |
| 5 | Micron MUon Hyperliquid 📈 chart | Equity | ▼ -1.9% hist -3.47–+0.26% · other way -1.63% (n=12) |
| 6 | TSMC TSMon Hyperliquid 📈 chart | Equity | ▼ -1.9% hist -1.75–-0.39% · other way +1.86% (n=12) |
| 7 | AMD AMDon Hyperliquid 📈 chart | Equity | ▼ -1.6% hist -1.33–+0.21% · other way -3.31% (n=12) |
| 8 | Marvell MRVLon Hyperliquid 📈 chart | Equity | ▼ -1.6% hist -1.07–-0.25% · other way +0.78% (n=12) |
| 9 | ASML ASMLon Hyperliquid 📈 chart | Equity | ▼ -1.5% hist -3.96–+0.99% · other way -0.21% (n=12) |
| 10 | Alibaba BABAon Hyperliquid 📈 chart | Equity | ▼ -1.3% hist -1.07–-0.33% · other way -0.24% (n=12) |
| 11 | Tech sector XLK 📈 chart | Equity | ▼ -1.4% hist -0.9–-0.48% · other way +0.4% (n=12) |
| 12 | Chinese yuan CNY 📈 chart | FX | ▼ -1.0% hist -0.89–-0.16% · other way -0.01% (n=12) |
| 13 | Qualcomm QCOMon Hyperliquid 📈 chart | Equity | ▼ -1.1% hist -3.94–+0.82% · other way -2.95% (n=12) |
| 14 | Intel INTCon Hyperliquid 📈 chart | Equity | ▼ -0.9% hist -6.61–+1.18% · other way -2.72% (n=12) |
Probable recommendation
Why we may diverge from history
Trust the cascade long on USDJPY: the -1.8% history is built on 2024 chip/AI-capex selloffs, not a model-weight export rule — today's compute-control shock is dollar-bid, unlike those megacap-led drawdowns.
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 |
|---|---|---|---|---|---|---|
| INTC INTC | SHORT | -5.4% · 5d -3.2% | 67% | 39 | 0.34 | ✓ matches cascade |
| ETH ETH | SHORT | -7.3% · 5d -5.2% | 72% | 38 | 0.34 | ✓ matches cascade |
| XCU XCU | SHORT | -1.2% · 5d -1.0% | 67% | 39 | 0.29 | ✓ matches cascade |
| KWEB KWEB | SHORT | -3.0% · 5d -1.3% | 67% | 39 | 0.29 | ✓ matches cascade |
| Gold XAU | LONG | +1.3% · 5d -0.0% ↺ fades | 67% | 39 | 0.29 | · |
| FCX FCX | SHORT | -2.0% · 5d -1.9% | 67% | 39 | 0.28 | ✓ matches cascade |
| SOL SOL | SHORT | -8.5% · 5d -6.8% | 70% | 36 | 0.27 | ✓ matches cascade |
| TRY TRY | SHORT | -1.5% · 5d +0.8% ↺ fades | 65% | 39 | 0.24 | ✓ matches cascade |
| KRW KRW | LONG | +0.5% · 5d -0.1% ↺ fades | 63% | 39 | 0.24 | ⚠ differs |
| NDX NDX | SHORT | -0.5% · 5d -1.3% | 65% | 39 | 0.21 | ✓ matches cascade |
| MU MU | SHORT | -2.3% · 5d -3.4% | 63% | 39 | 0.21 | ✓ matches cascade |
| CNY CNY | SHORT | -0.4% · 5d -0.3% | 60% | 39 | 0.19 | ✓ matches cascade |
| GBPUSD GBPUSD | LONG | +0.2% · 5d -0.1% ↺ fades | 62% | 39 | 0.19 | ⚠ differs |
| QCOM QCOM | SHORT | -3.1% · 5d -2.9% | 62% | 39 | 0.17 | ✓ matches cascade |
Why this probability
AI model-weight/compute export rules actively drafted; tiered FLOP-threshold framework already advancing. A base‑rate‑anchored prior, continuously scored against what actually happens — not a forecast.