Who Owns the Data When Markets Become Software?

On April 23, 2026, Nasdaq announced new “Plus” market-data products for its planned 23-hour trading day. Nasdaq Basic Plus will combine best bids and offers from three Nasdaq venues. TotalView Plus will combine full order-book data through one connection. The launch is planned ahead of extended trading, which Nasdaq expects on December 6, 2026, subject to approval and market-system readiness.

This is not a data upgrade. It is distribution capture. As markets run longer, firms need one clean stream that can feed risk engines, trading models, and client products without pause.

Nasdaq creates the market event, packages the event, and sells the easier route into software. The exchange keeps direct feeds for firms that want more control. It also sells a simpler feed to firms that want lower operating cost.

The hidden asset is not the quote. It is the right to turn the quote into action.

Data ownership has six layers: raw events, first access, clean format, reuse rights, derived signals, and final execution. Different firms can control each layer.

A broker may receive exchange data, normalize it in the cloud, combine it with client flow, and send a derived signal into an execution model. Each handoff can add cost, delay, or limits on use.

Legacy exchanges defend the stack through contracts. CME Group, for example, offers licenses for firms that use its market data to build curves, indexes, risk tools, exchange-traded products, and other derived works. The fee does not stop at seeing the market. It can extend into what software builds from the market.

This creates a second margin pool. The first comes from matching trades. The second comes from controlling the data those trades produce and the products built from it.

The SEC has tried to reduce this control point in U.S. equities. Its market-data rule expands the data that must enter the national system and supports competing data consolidators. That may reduce some latency and packaging power, but it does not remove the exchange as the source of the event.

The key split is clear: facts may be public, but useful access remains scarce. Speed, clean structure, legal reuse, and system uptime still carry a toll.

Legacy markets place data rights around the matching engine. Exchanges set feed terms. Vendors pay for access, clean the feed, and sell it again. Large firms buy several versions because small delays can change execution quality.

This model offers strong records, clear liability, and stable data labels. It also creates a tax on every new use. A firm may need separate rights for display, machine use, redistribution, history, and derived products.

Autonomous markets move part of the record onto shared ledgers. That lowers the cost of checking past events. It does not make live, usable data free.

The new gates sit elsewhere. RPC providers control reliable access. Indexers turn raw chain events into fast queries. Oracle networks decide which outside data enters contracts and how feeds update. Chainlink price feeds, for example, aggregate data from several sources through independent node operators.

Sequencers hold a stronger position. Arbitrum says most transactions pass through its sequencer, and its direct sequencer endpoint offers the lowest-latency path. Public RPC endpoints are rate-limited, while paid providers offer more speed and capacity. The chain is open, but the fast lane is still an owned interface.

Order flow adds another layer. Flashbots MEV-Share lets wallets and apps reveal selected transaction data to searchers, who bid for access to the value around that flow. Data is not merely observed after execution. It is sold before execution and used to divide the margin.

Autonomous systems therefore do not end data ownership. They split it across sequencers, RPC firms, indexers, oracle operators, wallets, builders, and users. The control surface becomes modular, but it remains commercial.

Capital will move toward data systems that reduce total execution cost. That cost includes feed fees, delay, errors, legal limits, cloud load, and the staff needed to keep the pipe alive. A cheap feed that cannot support fast action is not cheap.

Large firms will keep buying direct access where latency changes outcomes. Other firms will move toward consolidated feeds, hosted APIs, and managed data layers. They will trade some control for lower build cost and faster deployment.

On-chain capital will follow the same rule. It will cluster around networks and service layers with reliable state, fast queries, trusted oracles, and clear order-flow rules. Public data helps, but dependable machine access wins the flow.

The strongest data owner will sit between raw events and automated action. That position controls routing, product creation, and the terms of reuse. It can charge the market without taking balance-sheet risk.

When markets become software, data becomes part of the execution layer. Ownership is no longer a simple claim over a file. It is control over access, speed, format, reuse, and the right to act first.

Legacy exchanges capture this through licenses and direct feeds. Autonomous markets capture it through sequencers, RPC access, indexing, oracle design, and order-flow auctions. Both systems sell a path from market state to machine action.

The new financial reality is hard: whoever controls the machine-readable state controls the market’s next product, next route, and next fee. Open records do not end ownership. They move ownership to the interface.

-{{sendingAccountName}}

Sources: Nasdaq, U.S. Securities and Exchange Commission, CME Group, Chainlink, Arbitrum, Flashbots, The Graph