Executive Summary

In January 2025 the President of the United States ordered the Department of Defense to design a national missile shield, first called the Iron Dome for America and soon renamed Golden Dome for America. Four months later the White House put a price on it, $175 billion, and a date, before the end of the presidential term in early 2029. In July 2025 Congress appropriated about $24.4 billion of mandatory money for integrated air and missile defence through the budget reconciliation law. By March 2026 the programme's director, General Michael Guetlein, had raised the estimate for the "objective architecture" to $185 billion and moved the full architecture's horizon to around 2035. By mid-August 2026 about 90% of the reconciliation money had been obligated. And by October 2026 the programme had no settled money for fiscal year 2027 at all: the administration had requested about $17.5 billion, of which only about $400 million sat in the regular defence budget.

This study asks a narrow question with large consequences. where will defence-space money flow from 2026 to 2035, which layers of the architecture are funded and on contract rather than announced, and which suppliers sit on the critical path? It is written for institutional readers who need to understand the shape and durability of this demand: infrastructure and industrial investors, lenders to the supply chain, government and allied planners, and companies deciding whether to build for it. Figures are as of 21 October 2026 and are sourced in the text. Where a number is Setcoin's own reconstruction or estimate, it is labelled as such and the arithmetic is shown.

The answer in brief

Golden Dome is, so far, a programme that has been funded to see and to talk, but not yet to shoot.

The sensing and data-transport layers are on contract. The Space Development Agency has awarded about $3.5 billion for 72 Tranche 3 tracking satellites and a further $1.75 billion for 36 accelerated missile-defence tracking satellites. The Space Force has awarded SpaceX $4.16 billion for the first satellites that track aircraft and cruise missiles from space and $2.29 billion for the backbone of a military space data network. Those awards run on firm, fixed-price or prototype agreements with launch dates in 2027 to 2029.

The interception layer is not decided. Space-based interceptors are the most ambitious and most expensive part of the concept. Twelve companies hold prototype agreements with a combined ceiling of $3.2 billion, but the programme's own director has told Congress that boost-phase interceptors may be too expensive for a $185 billion plan and may not make the final architecture. The Congressional Budget Office's estimate of about $1.2 trillion over 20 years is high precisely because it includes a constellation of several thousand interceptors.

The money is the weakest link, not the technology. The first $24.4 billion came through a one-off reconciliation law. The fiscal 2027 request depends on a second reconciliation bill for about 97% of its value, and because Golden Dome had no fiscal 2026 appropriation, a continuing resolution gives it nothing to roll forward. The programme's director has said plainly that without new legislation "there is no Golden Dome because there is no funding."

Nine findings

• Three price tags measure three different things. The Department's $185 billion is the acquisition cost of its objective architecture through about 2035. The CBO's $1.191 trillion is the 20-year cost to develop, deploy, replace and operate an architecture built from the executive order, including a large interceptor constellation. AEI's $3.6 trillion is the most expensive of six illustrative architectures over 20 years. They are not competing estimates of the same system; they are estimates of different systems over different periods.
• The interceptor layer is the swing factor. In the CBO estimate, space-based interceptors are about 70% of acquisition and 60% of total cost. Remove or shrink that layer and the estimates converge toward the hundreds of billions; keep it at the scale needed to counter a large salvo and they diverge toward the trillions.
• Sensors and data networks are the durable demand. Tracking satellites, moving-target indication, data transport and launch are useful to the armed forces with or without interceptors in orbit. They were funded before Golden Dome existed (the Space Development Agency's architecture dates from 2019) and they continue in every budget scenario this study models.
• The money is structurally fragile. About $24.4 billion of enacted mandatory money, a fiscal 2027 request that is about 97% contingent on reconciliation, no fiscal 2026 base, and a Space Force topline that itself leans on $12.1 billion of reconciliation in its $71.3 billion fiscal 2027 request. A sustainable programme needs regular appropriations; it does not yet have them.
• One supplier holds a large share of every funded layer. SpaceX holds the data-network backbone award, the first moving-target-indication award, one of the twelve interceptor prototype slots, the largest share of national security launch orders, and launches most Space Development Agency tranches. That concentration lowers near-term cost and raises long-term dependency risk.
• Schedules slip; the threat timeline does not. Tranche 1 transport launches paused for several months in 2026 after on-orbit checkout problems; the first Next-Generation OPIR missile-warning satellite slipped from December 2025 to no earlier than October 2026; Tranche 3 tracking satellites launch in fiscal 2029, which is why a separate accelerated tranche was bought for 2028.
• The wars have exposed the magazine, not the sensors. Ukraine has faced more than 6,000 missiles and drones in a single month, and in the 2026 Iran war the United States and its partners intercepted more than 1,500 ballistic missiles. The CBO estimates that the United States used between one-half and two-thirds of its inventory of certain missile-defence interceptors and that rebuilding them could take at least five years. Replenishing Patriot, THAAD and Standard Missile stocks now competes with Golden Dome for the same budgets and the same solid-rocket-motor suppliers.
• Europe is building its own layers, later and smaller, and still has no missile warning of its own. The EU's IRIS² constellation was expanded in August 2026 to 348 satellites, including a 66-satellite defence and security layer, at an estimated €15.6 billion; Germany has committed about €35 billion to space security by 2030 with plans for up to 1,200 satellites; France is adding €4.2 billion for patrol satellites and counterspace; the United Kingdom plans £7.8 billion of space spending to 2030. The Franco-German JEWEL warning system targets 2030 for its space segment, and a European hypersonic interceptor is not expected before about 2035. In Asia, Japan is the ally most likely to add sensors to the US architecture: it co-develops the Glide Phase Interceptor, hosts US surveillance sensors on its navigation satellites and plans glide-vehicle tracking satellites.
• The adversaries are answering with more missiles and with weapons aimed at the sensors. China is on course for more than 1,000 nuclear warheads by 2030, launched 91 orbital missions in 2025 and is deploying two megaconstellations; Russia has fielded weapons designed to evade missile defence, parks inspector satellites beside US reconnaissance satellites and is assessed to be developing a nuclear anti-satellite weapon. New START expired in February 2026. Both powers call a space-based interceptor layer destabilising, and both are likely to respond to it by building more of what they already build.

The layers at a glance

What this means for different readers

For infrastructure and industrial investors, the investable demand is the part of the architecture that is already on contract and that every scenario keeps: satellite buses at production rates, infrared and radar payloads, optical terminals, ground stations and launch. The interceptor layer is an option, not a forecast.

For lenders and suppliers, the risk is not that the programme disappears but that the money arrives late and unevenly. Firm, fixed-price and prototype agreements shift cost risk to suppliers; reconciliation money arrives in one-off tranches; and a continuing resolution can stop a programme that has no base-budget line.

For allied governments, the American architecture sets the technical standards (optical terminal standards, data links, sensor formats) that allied constellations will need to meet if they want to share data with it. Europe's choice is whether IRIS² and national constellations are built to connect to it or to stand apart from it.

For policy readers, the strategic question is less whether the United States can build a shield than how large a shield it can sustain, and what an interceptor layer in orbit does to the stability of deterrence and the debris environment.

How to read this study

Chapter 2 sets out the threat and Chapter 3 the lessons of the wars in Ukraine and with Iran. Chapters 4 to 7 cover the programme, the money, the three cost estimates and forty years of history. Chapters 8 to 16 work through the budget and the architecture layer by layer: the Space Force budget, the Space Development Agency's constellation, the unit economics of defence satellites, missile warning and tracking, moving-target indication, the data layer, the kill chain in seconds, space-based interceptors and the ground and sea layers. Chapters 17 to 21 cover launch, commercial integration, the supplier map, the 2026–2035 demand map (with an interactive map of where the demand lands) and industrial bottlenecks. Chapters 22 to 24 cover Europe: the EU and NATO layer, the national programmes, and Europe's missile-warning and interceptor gap. Chapters 25 and 26 cover Japan and the wider Indo-Pacific; Chapters 27 and 28 China and Russia; Chapter 29 stability, debris and arms control. Chapter 30 sets out what the findings mean for different readers, Chapter 31 the budget scenarios to 2035, and Chapter 32 the risk register, the signals to watch and the method.

Five interactive models accompany the study. the budget flow, a reconciler for the three cost estimates, a space-based interceptor constellation sizer, a tranche scheduler and a supplier exposure model. Each opens with the figures used in the text, and every input can be changed.

What would change this view

This study's central reading, that sensing and data layers are durable while the interceptor layer is optional and the money fragile, would change if any of the following happened: Congress funds Golden Dome in the regular fiscal 2027 appropriations at or near the requested level; the Department publishes an objective architecture that fixes the size of the interceptor layer; a prototype space-based interceptor demonstrates an intercept in 2027 or 2028; or a major adversary test demonstrates that the sensing layers cannot provide fire-control quality tracks against hypersonic glide vehicles.