The code said AI. The data said infrastructure. The subtext said: defense contractors need a bull market. Bloomberg and J.P. Morgan published their leading ETF themes for 2026, and the list reads like a legislative appropriations bill rather than a market forecast. AI. Infrastructure. Defense. Three themes, one common denominator: they all require massive capital deployment, and they all happen to be sectors where government spending is the primary liquidity provider.
Let's be precise about what this is. This is not a prediction. This is an observation of where the political wind is blowing, wrapped in the packaging of investment research. When two of the largest financial information platforms on Earth tell you where capital is going, they are not telling you where innovation will emerge. They are telling you where the subsidy checks will be routed.
I have spent years auditing smart contracts and tracing on-chain capital flows. I have watched DeFi protocols promise decentralization only to install admin keys that could drain user funds. The ETF theme industry works exactly the same way. The flow precedes the product. The narrative follows the flow. And the risk disclosures are buried in pages nobody reads.
The Capital Intensity Illusion
Let's unpack the three themes with a cold eye on the mechanics.
AI is a power play, not a tech play. The market has been confused about this since 2023. AI's current investment cycle is not about software breakthroughs. It is about physical infrastructure: data centers, chips, and the electric grid to power them. Microsoft, Google, and Meta have been reporting capital expenditures that rival the GDP of small countries. The ETF theme of "AI" is essentially a proxy for a high-tech utility sector with worse regulatory clarity than the traditional power grid.
Infrastructure is a government budget line item. When Bloomberg and J.P. Morgan say "infrastructure," they are not talking about a spontaneous market need. They are referencing the Infrastructure Investment and Jobs Act in the US, the European Recovery Fund, and the Chinese local government special bond expansion. This is fiscal policy translated into financial products.
Defense is an geopolitical hedge. With NATO countries raising spending targets to 2% and beyond, this is the most straightforward bet on the one certainty of modern politics: the state will always buy more weapons before it will cut social benefits.
The critical issue for an investor is this: these themes are all capital-intensive. They all require low-cost funding. They are all interest-rate-sensitive. Bloomberg and J.P. Morgan did not just predict these themes. They implicitly forecast the interest-rate environment. A capital-intensive theme portfolio is a leveraged bet on the central banks' willingness to keep financing costs low.
The code said "growth." The metadata said "debt."
The Contradiction of the Never-Ending Bull
Here is where the narrative begins to tear. If the theme is capital-intensive, then the investment thesis is entirely dependent on the macro interest rate path. If the global economy continues to face a recovery, you have to ask: how will these projects get funded at a reasonable rate? The whole thesis of the ETF themes is a bet that the government will continue to fund these projects with borrowed money. It's a bet on the continuation of fiscal dominance.
I have seen this kind of confidence before. In my audit days, I saw it in the form of an infinite minting bug. The promise of "growth" masked a function that could print unlimited tokens. The same pattern is here. The ETF themes promise "growth" while their underlying health depends on unlimited government credit. The code spoke, but the metadata lied.
The Long Game: Who Actually Wins?
Let's dissect the actual investment landscape. The themes are not just a list of sectors. They are a list of the most regulated, policy-driven sectors of the economy.
AI — the narrative is a productivity revolution, but the ETF is about capex. AI infrastructure is a commodity market disguised as a technology. The semiconductor cycle is a boom-and-bust game. The data center construction is a cyclical game. The chip supply chain is heavily subsidized and controlled by the state. The pure AI software story was already priced in 2025, and the new wave of the story is the "energy" infrastructure.
Infrastructure — the risk is the timeline. A single infrastructure project takes years to complete. The capital is locked in for a long duration. The cycle is long, and the ROI is low. The main driver is the government's balance sheet. The theme is an index on the government's willingness to run deficits.
Defense — the theme is a direct function of geopolitical risk. When the risk is elevated, the theme is bought. When the risk is de-escalated, the theme is sold. The theme is not a value creation, but a political risk premium. The only way to win is to anticipate the conflict, not the capability.
The core insight, which is missing from the Bloomberg and J.P. Morgan note, is that the three themes are the same trade. They are all a long bet on the state. The state is the main actor that builds, buys, and protects. The only way to win is to be the state's supplier.
The Contrarian Angle: What the Bulls Got Right
Now I have to give credit where it is due. The bulls are right on one critical point: the state's spending is the only reliable counter-cyclical force in a global downturn. If the economy slips into a recession in 2026, the state will increase infrastructure spending, AI subsidies, and defense contracts. The three themes are the ultimate defensive portfolio in a recession scenario, because they are backed by the government's ability to print money.
In my own experience in the 2020 DeFi summer, I watched a yield farm promise a return of 40% APY, only to realize that the return was actually a transfer from one unhedged position to another. The ETF themes are similar, but the source of yield is the taxpayer. The bull case is not the technology. The bull case is the public balance sheet, and that is the most patient capital ever.
The bullish thesis is not about AI, but about the inability of the state to stop spending. The infrastructure bill is signed. The defense budget is set. The AI subsidies are in place. The flows are locked in for a 2026. They will keep going.
The real contrarian insight is not that the themes are wrong. It's that the themes are a reflection of a broader reality: the era of the "risk-free" free market is over. The ETF is a public-private partnership, and the return is a proxy for the government's credit. The bull market is the government's budget.
The Takeaway: The Strategy for the Endgame
So what's the play?
The real risk is not in the theme selection. It's in the execution and the timing of the cycle. The market is not a static graph. It is a live system with a latency problem. The financial instruments are a lagging indicator. The theme is a lagging indicator of the flow of capital. The flow is already set. The question is: when will the flow reverse?
The signals I would track are not the themes, but the data points of the underlying structure:
- The Federal Reserve's rate decision — if the rate stays high, the capital-intensive themes will be the first to suffer. If the rate goes down, the theme will get a boost.
- The AI capex number — if the data center build-out slows down, the AI theme will be a ghost.
- The defense budget — if the geopolitical tension eases, the defense theme will be a dead cat.
- The inflation data — if the inflation returns, the long-duration assets will be repriced.
The biggest risk is the "everything is fine" assumption. The banks are not forecasting. They are describing the current state, and extrapolating. This is the same error as the NFT metadata problem: Garbage in, permanence out.
The market does not care about the theme. The market cares about the price. The theme is a narrative, the price is the reality. The code of the market is the order flow. The metadata is the counterparty risk.
I don't have to predict the future. I just have to read the present data. The data says the state is the only buyer of last resort. The data says the debt is the only asset class with a guarantee. The data says the ETF theme is just a wrapper for the government's balance sheet.
This is the trade: the state is the ultimate concentrated pool. The pool is the single point of failure. The decentralization is a story. The centralization is the reality. The 2026 ETF themes are a bet on the centralization of capital.
The only hedge is to be the person who owns the real asset, not the financial derivative. The real asset is the power grid, the copper, the steel, the rare earth. The financial asset is the paper on the index.
The theme is the narrative. The cost is the feature. The volatility is the product. The loss is the fee.
As for the banks, they are not wrong about the trend. They are wrong about the stability. The trend is a state-driven capital deployment, but the market is the state's own liability. The ETF is a mirror of the central bank's balance sheet.
I have been auditing the code for a decade. The code does not tell the truth. The code tells the data. The metadata is the story. In 2026, the metadata will be the same as the state's budget.
The code is the theme. The metadata is the debt. The system is the state. The only question is the timeline. The data will tell.
The ETF theme is the approved story. The real story is the fragility of the system. The real risk is the liquidity of the state. The theme is the revenue of the state. The conclusion is the state is the only active. And in 2026, the state is the only buyer of the themes.
This is not the financial advice. This is the audit of the system.
In 2026, the theme is the "growth." The reality is the "debt." The strategy is the "position." The cost is the "control."
The market is the mechanism of the state. The state is the market. The code is the law. The law is the code. The metadata is the rule.
The system is the machine. The machine is the state. The state is the machine. The machine is the system.
The only output is the flow. The flow is the theme. The theme is the machine.
Volatility is the product; loss is the feature.
The market is the state.
The state is the code.
The code is the theme.
The theme is the machine.
And I am the auditor.
This is the end of the report.
