Reuters and Bloomberg reported in April 2026 that Jane Street generated $39.6 billion of net trading revenue in 2025. Reuters also reported that Citadel Securities, founded by Ken Griffin, generated about $12.2 billion. The numbers look obscene because the machinery that produced them is mostly invisible.
The blunt story is that finance found another way to skim the productive economy. Sometimes that story is right. It is not precise enough to be useful.
The trade is the moral unit. Not the firm, not the sector, not the billionaire. A trading profit can come from manipulation, legal privilege, speed, regulatory capture, balance-sheet capacity, liquidity provision, risk transfer, or better analysis. These are different mechanisms. Collapsing them into "finance" hides the only question that matters: did the activity make price a better allocator of real resources, or did it make price worse while taxing everyone forced to use it?
Productive trading profit is payment for market cognition under cost exposure.
Price discovery is the work of making production legible to itself.
A stale price sends capital to the wrong project. A wide spread makes exit expensive. A thin market makes hedging unreliable. A missing buyer strands inventory in the wrong hands. A missing seller prevents a real user from obtaining the thing she needs. Allocation error is not symbolic waste. It is output that never happens.
The market maker's defensible function is to reduce that error. Quote both sides. Hold inventory. Absorb temporary imbalance. Update prices when order flow reveals information. Let the producer hedge, the investor exit, the buyer enter, the adjacent market update. The market maker gets paid because immediacy, inventory, and adverse-selection risk are costly.
The analyst-trader's defensible function is adjacent. See a change before the median participant sees it. Understand which variable matters. Notice which adoption curve is real, which regulation will bind, which supply constraint is mispriced, which consensus has gone stale. Express that view through a trade. If the trader is wrong, the loss arrives directly. If she is right, the price moves toward the world faster than it would have without her.
That is not nothing. It is cognition with a P&L.
A bigger economy does not need less market cognition. It needs more.
Every expansion of production creates new unsettled claims. AI compute creates claims on power, land, transformers, GPUs, fiber, water, permits, debt capacity, cloud margins, inference demand, and the speed of enterprise adoption. Stablecoins create claims on Treasury demand, payment rails, banking deposits, regulation, offshore dollar usage, and sovereign-currency competition. A Mars economy would create claims on launch cadence, oxygen, water, habitats, liability, insurance, mining rights, orbital congestion, medicine, radiation shielding, and law.
Each new claim is a local perturbation the global economy has to evaluate. No planner can hold all of it. No quarterly report can move fast enough. Prices are the compression format. Trading firms are among the actors paid to keep that compression current.
This is the non-cynical reason Jane Street can make more money when America has more opportunity. Opportunity increases the surface area of uncertainty. More firms are born. More technologies cross diffusion thresholds. More capital tries to move from old uses to new uses. More people with local knowledge can act. More instruments appear to express views and transfer risk. The market-cognition layer gets larger because the real economy is throwing off more things that need live pricing.
The claim is not that volatility is always good. A crisis can enrich market makers while destroying real wealth. The claim is that under genuine productive expansion, the value of live pricing rises because the number of things that need to be priced rises.
The useful test is not whether the trader made too much money. "Too much" is not a mechanism.
Ask four questions.
Did the trade add immediacy or depth? If others could enter, exit, hedge, or rebalance with lower cost because the trader stood ready, some service was supplied.
Did the trade improve price as an information object? If the activity moved price toward a better estimate of future cash flows, scarcity, risk, or demand, the profit came from information improvement.
Did the trader bear real risk? Inventory risk, adverse-selection risk, basis risk, and model risk are not moral theater. A party exposed to being wrong is different from a toll collector who gets paid either way.
Did the edge avoid coercion, deception, and illegal information advantage? The edge has to come from better synthesis, lawful local knowledge, speed, capital, or risk-bearing. If it comes from hiding material facts, manipulating settlement mechanics, capturing rules, or trading on forbidden information, the profit changes category.
The firm name cannot settle this question. Jane Street can provide real liquidity in one venue and still face serious allegations in another. India's securities regulator issued a 2025 interim order alleging manipulative index-derivatives activity by Jane Street entities. Allegation is not conviction, and the procedural status matters, but the case preserves the boundary. A trading firm is not purified by being useful somewhere. It is judged trade by trade, venue by venue, mechanism by mechanism.
The strongest critique of modern market making is not that traders get rich. It is that some profits may come from venue fragmentation, routing privilege, opacity, or rule capture while being narrated as liquidity. That critique is real. The answer is not to romanticize traders. The answer is to ask whether the price got better.
The worst response to billionaire traders is resentful mystification. The second-worst response is hero worship. The useful response is curiosity about edge.
Jane Street's edge is not portable to most people. Its balance sheet, hiring machine, latency stack, venue access, modeling culture, and risk systems are industrial machinery. But the general shape of edge is portable: know something the median price-setter does not know, lawfully, because of where you stand.
Hayek's old point about local knowledge still cuts. Some knowledge is not centralized or scientific. It belongs to the person close to the circumstance: the builder using a tool before adoption shows in revenue, the practitioner watching a workflow break before the market names the bottleneck, the domain expert who can tell which regulation matters in practice, the customer who feels demand before the analyst model updates.
That is not insider trading. Insider trading uses forbidden access to material nonpublic corporate information. Local-knowledge edge uses lawful observation, public material, lived contact, and better synthesis. The distinction is the line between contributing cognition to a market and corrupting the market's information structure.
The right instruction is not "become Jane Street." It is narrower and harder: find the smallest market where your lawful local knowledge beats the consensus, size the position so being wrong teaches rather than destroys, and keep score. The market is not only where rich people take money from everyone else. It is also where a real edge becomes a public bid that the current price is wrong.
The Pricing of Everything argues that intelligence makes more domains explicitly priced. If citation flow, compute, energy, carbon, water, attention, agent labor, and physical resources become priced at finer granularity, then price maintenance itself becomes more valuable. Some of that work will be productive. Some of it will be toll collection. The discriminating test has to move with the pricing layer.
The trader is not sacred. The billionaire is not proof. The trade either improves the economy's cognition under real exposure, or it does not.
The thesis is wrong if Jane Street-style profit rises while market quality deteriorates after controlling for volatility: wider effective spreads, worse execution, noisier settlement prices, less reliable hedging, lower depth, more rule capture, and more profits traceable to opaque venue position than to risk-bearing or information improvement. In that world, the trading layer is not the economy's cognition layer. It is a tax on everyone who needs prices.
The thesis is incomplete if the cognition is real but the surplus capture is socially excessive. A trader can provide a real service and still capture more of the surplus than healthy market structure should allow. Social value and distribution are separate questions.
Do not hate the billionaire traders because they are rich. Ask what mechanism made them rich. Then ask where, legally and honestly, your own position sees something the price has not learned.
Sources: Reuters via Investing.com on Jane Street's reported 2025 net trading revenue and Citadel Securities' reported 2025 trading revenue; Bloomberg on Jane Street's reported 2025 trading revenue; F. A. Hayek, "The Use of Knowledge in Society" (1945); Shen and Starr, "Market-makers' supply and pricing of financial market liquidity" (Federal Reserve Bank of Kansas City, 2000); Leach and Madhavan, "Price Experimentation and Security Market Structure" (Review of Financial Studies, 1993); SEC, "U.S. Equity Market Structure: Making Our Markets Work Better for Investors" (2015); SEBI interim order in the matter of index manipulation by Jane Street Group (2025).