Graph Research Tools Method About Login Get Pro →

The knowledge graph that maps where capital is forced to go — before markets price it.

A causal graph of physical dependencies, supply chain chokepoints, and forced capital flows across semiconductors, defense, energy, and critical minerals. When the world changes, the graph traces who is exposed and what must move next.

Supply chain cascade

SiC wafer demand inflects with AI and EV buildout → WOLF is the dominant Western SiC substrate supplier → ON Semi and Infineon sit 2 hops downstream → Defense and hyperscaler procurement converge on the same node → CPO laser supply chain mapped Mar 16

WOLF: +224% since Mar 16, 2026

In the graph since Mar 16, dependencies wired and price monitored since. Now rated a near-monopoly chokepoint.

Free tier includes weekly intelligence. Pro: Full graph access + real-time data + deep analysis →

Explore the live graph →
4686
Nodes
20582
Edges
34
Sources
8
Sectors
The Intelligence Layer

Map What Physically Cannot Be Substituted

Physical dependency chains with severity ratings, typed relationships, and zero-alternative flags across the global supply chain.

AXTI SIVE AAOI TSMC LMT RTX MP AI Defense Energy Minerals Robotics

When disruptions hit Ras Laffan and the Strait of Hormuz in early 2026, headlines focused on oil prices.

The graph traced 71 downstream dependencies most analysts missed.

Qatar produces 25–33% of global helium — a critical gas for semiconductor fabrication. The cascade: helium supply → TSMC and Samsung fabs (leak detection, EUV source cooling) → ASML lithography systems → advanced chip production worldwide. South Korea sources 64.7% of its helium from Qatar. Taiwan sources 69% from the Gulf.

The Hormuz closure compounded it — 71 nodes reachable within 3 hops, from crude oil and LNG flows to Korean fab energy supplies to Japanese naphtha used in chip substrates. Each path was already mapped in the graph before the first tanker rerouted.

Every edge is typed, directional, and severity-rated. Every cascade path is computable in milliseconds.

Explore the Graph →
4686
Nodes mapped
across 42 sectors
20582
Supply chain edges
with typed relationships
670×
Helium asymmetry
$6B market gates $4T+ in chips
Live Graph Query

Supply Shock Simulator

Select a disruption. Watch the cascade propagate through real supply chain dependencies.

HBM3E memory is the binding constraint on AI GPU production. SK Hynix controls ~50% of supply. A bottleneck cascades into every major AI company.
0%
1-2 Multiple suppliers
3-4 Few alternatives
5 Sole source
34 Active

Intelligence Network

Automated collection across government, markets, and policy — feeding every data point into the knowledge graph.

Government
Congressional trades, lobbying, DOD contracts, activist stakes, bills, SEC filings
Markets
Options, insiders, 13F, technicals, buybacks, dilution, earnings, hyperscaler, buyback execution, short interest, FTD, debt maturity, treasury auctions, FedWatch
Policy
Federal policy, trade policy, prediction markets, BIS/IMF papers
Research
Transcript sentiment, SEC letters, COT, jobs, ETF flows, credit stress, ecosystem

Automated collection feeds every data point into the knowledge graph to detect chain-level convergences

Published Research & Theses

Supply chain deep dives and investment theses generated from the graph

SUPPLY CHAIN
Japan: $23.8T Downstream Exposure
48 companies. 15 materials with no alternative. Six domains. First computational mapping of cross-domain dependency.
830 Nodes 6 Domains Mar 2026
SUPPLY CHAIN
Korea: The Memory Chokepoint Behind Every AI Chip
~79% of global HBM production. Two companies. Zero US production. Three-layer allied bottleneck with no redundancy.
HBM $54.6B Market Mar 2026
THESIS
Carpenter Technology: Specialty Alloy Tollbooth
500+ patents, 90% jet engine cert coverage. EU rearmament = structural demand floor. Graph-identified chokepoint.
CRS Defense Feb 2026
THESIS
Amazon: AI Infrastructure Capex Cycle
Multi-source convergence across congressional, institutional, and supply chain layers. $100B+ AWS AI build-out.
AMZN AI Infra Feb 2026
CASEFILES

Open investigations

Live, auditable investigations into physical-economy cascades. Every branch tested, every monitoring trigger, every kill condition stated in advance. Published before consensus prices them.

Casefiles load dynamically. Browse the archive →
View all casefiles →
AM
Ahmed Mir

I spotted a convergence in silver — congressional trades, a technical breakout, and institutional accumulation all hitting one after another. I bought calls and made 12x. But I sold at $48.

It went to $95.

The thesis was right. The entry was right. What I lacked was systematic convergence scoring — a way to know when enough structurally distinct data points had aligned to hold, not just enter.

Seven years inside Amazon's constraint systems — regulatory deadlines that couldn't slip, manufacturing floors that forced procurement, compliance rules that triggered real economic action regardless of anyone's intent. ForcedAlpha applies the same thinking: find the structural constraints that force capital to move, then map the chain before the market prices it.

Every data source is public. The structural intelligence that connects them is ours. Every alert is logged at detection, not hindsight. No cherry-picking, no silent deletions.

Our thesis: when governments act, institutions file, and capital shifts — every downstream chain is trackable before the market prices it. Read the framework →

Start with the Graph

Free tier — weekly intelligence from 4686 nodes and 34 sources

Upgrade to Pro anytime — or integrate directly for enterprise API access.