The line is old and it is correct: control the flow and you control the system. The flow does not begin at a contractor or a fund. It begins when a government borrows or a central bank expands its balance sheet in dollars. Banks and dealers intermediate the debt. Foundations and asset managers warehouse the claims. MIC, TIC, and FIC convert those claims into contracts that redistribute value, capability, and control. Some sequences match the classic pump-and-dump shape: accumulate, catalyze, extract, lock. The product is three proofs: intent from the record, actions that moved the flow, and a bounded future that expires.
01The three complexes are stacked, not parallel
The object of study is three worldwide systems that provide services to the entire planet and recognize no sovereign boundaries: the Financial, Military, and Technology Industrial Complexes — FIC, MIC, TIC. The naive question is "who are the top five in each." The real question is the one underneath.
Treat them as three side-by-side buckets and you miss the structure entirely. They are layered. The FIC — specifically the handful of giant asset managers — owns the equity of the MIC and TIC primes. The TIC increasingly supplies the compute substrate that the MIC's targeting systems and the FIC's trading systems both run on. The FIC's own apex is a network of central banks coordinating through a private-law body in Basel that sets the price of money itself.
The sharpest structural point falls out of that picture: Tier 1 is public institutions coordinating privately; Tier 2 is private institutions with public-scale power. Both have escaped sovereign boundaries — but from opposite directions.
02Watch the scale move
Before the mechanism, the magnitudes — as they moved, not as a snapshot. Four series from the record, against a world economy of roughly $112 trillion: the military burden (SIPRI/World Bank), the semiconductor valve (WSTS), the stock of global financial claims (Allianz), and the combined holdings of the three largest asset managers. Two honesty rules govern the chart. The panels never share a y-axis — the FIC's 283% would flatten the chip valve's 0.6% into a floor. And they never sum: the tiers are stacked, so the same dollars recur up the stack.
data table
| year | MIC % GDP | chips % GDP | FIC % GDP | Big Three $tn |
|---|
The motion is the finding. The military burden halved after 1991 and has been climbing again since 2014. The chip valve's share cycles but ratchets upward. The stock of financial claims hovers near three times the world economy. And the Big Three line barely exists in 2008, then goes near-vertical: from under $4 trillion to over $26 trillion. Size and control run in opposite directions: the smallest line on this chart is the tightest valve on the planet, and the towering ones route through it.
03One dollar, three ledgers
So where is the total? Not in a sum — in a circuit. A dollar enters a fund at Tier 2, buys the equity of a prime, is spent by the prime on chips, and returns to the same managers as revenue, fees, and buybacks. Add the three complexes together and you have counted that dollar three times. The honest total is reach, not addition — the share of world flow that crosses at least one valve. That union is the focusing ratio φ.
04A valve has three properties, all at once
A control valve is not just a big node. It qualifies only when three conditions hold simultaneously. Where all three hold you have control; where only one or two hold you have influence.
In the FIC: dollar-clearing, messaging (SWIFT), the Basel reserve standard, US Treasuries as collateral. In the MIC: jet-engine hot sections, guidance-grade chips, specialty materials, and state procurement monopsony. In the TIC: ASML lithography, TSMC fabrication, Nvidia/CUDA lock-in, cloud, and export control as a meta-valve.
05The valves are not where the myth says
The tightest real valves are physical and technological, not financial-dynastic — lithography, fabrication, engine hot sections, dollar clearing. And the second layer of control over those valves is increasingly the state, not private families. The twentieth-century story of banking dynasties at the monetary valve has partly inverted.
06Is the flow being focused?
A static map of valves is not the point. There is constant redistribution of almost everything. The dynamical question: are the chokepoints being used to focus that redistribution — value, capability, and control — toward the few who hold the valves?
And concentration is arithmetic. Collapse holders that are not independent into blocs; sum squared shares for a Herfindahl index; invert for the effective number of independent holders:
07Demonstrated, not alleged
You cannot subpoena the world. You can read the public record — SEC filings, monetary and fiscal series, contract disbursements, proxy statements, fund vote disclosures — and assert a feature only when multiple independent record-streams co-locate on the same transfer. That is concordance across bearings. Nothing is a photograph; everything is a reconstruction with stated uncertainty.
Ownership at this concentration is not coordination, it is determination — readable from filings, votes, and disbursements. The claim is ungoverned, determinative concentration: a small set of actors holds gating leverage over the conditions of nearly every life. It is provable because it is shown, node by node. Where a deliberation is kept off the record, the method quantifies its effect — not a transcript of what was said.
08The hidden lever
The decisive mechanism is unrecorded: private engagement between asset managers and boards that resolves governance issues before a recorded vote. You cannot see inside it. You can characterize it by its distortion of the signals you can measure.
Four fingerprints, queryable over public timing data: pre-emptive non-conflict; window-clustered concessions; withdrawn proposals; engaged-vs-unengaged divergence. Two hard limits: the channel is adaptive (estimates decay), and it yields magnitude, not content.
09What is being built
Global Control Structures is a money-flow and deal system across FIC/MIC/TIC with three proof surfaces — intent, actions, future — and an immersive instrument that makes those proofs legible. Product first: deal ledger, lineage, technique signatures, math, perishable forecasts. Backend later: point-in-time storage and live multi-clock sources when fixtures are no longer enough. Engineering contracts stay as rails — they are not the product.