Bloomberg Terminal earns hundreds of millions annually through data relay, now 6 institutions have directly put data on the blockchain.

Author: Thejaswini M A

Translation: Deep Tide TechFlow

Deep Tide Guide: Fidelity, Euronext, Tradeweb, and 3 other major Wall Street institutions are starting to publish market data directly on the blockchain via Pyth, accessible for free by any developer. This breaks Bloomberg and other data middlemen’s 44-year monopoly—no more signing two-year contracts, no more paying $27k annually, no more dedicated keyboards. More importantly, this is a prerequisite for RWA (Real World Assets) to truly scale in DeFi: data must come on-chain before the assets do.

In 1981, Michael Bloomberg was fired from Salomon Brothers. He was 39 at the time, having worked there for 15 years, leaving with a $10 million severance, deeply dissatisfied with how Wall Street handled information. His reaction to being fired was, by any reasonable standard, crazy: he started showing up every morning with coffee at Merrill Lynch’s office, wandering the hallways, handing out coffee to strangers, and explaining he was building a computer that knew everything. Traders took the coffee but weren’t so sure about that computer.

44 years later, each of those computers costs $27k per year. There are 350k worldwide, and Bloomberg earns about $10 billion annually from this business. The genius of its structure is: he inserted himself between data-owning institutions and those who need data, collecting tolls on everything passing through. Data was never Bloomberg’s—Merrill has data, Goldman Sachs has data, every trading firm on Wall Street has data. Bloomberg just built a toll booth, convinced everyone that the toll booth was the destination, and raised prices every year because, what could you do—call your broker and switch?

This model has withstood every technological change over forty years because no one could think of a better distribution mechanism. Until last Wednesday.

On April 9, six institutions feeding data into the toll booth started publishing data elsewhere: Euronext, Fidelity, Tradeweb, OTC Markets Group, the Singapore Exchange’s Forex department, and Exchange Data International, directly on-chain via Pyth’s new data marketplace, accessible to any developer on the blockchain. No contracts, no two-year minimum commitments, no dedicated green-yellow button keyboards.

Remember this: the interesting part of building a monopoly on others’ data is that those others will eventually notice.

The financial data industry is worth about $30 billion annually, one of the least discussed monopolies in the world, perhaps because the only people paying attention are those already paying for it.

Bloomberg controls about 33% of the global financial data market, generating over $10 billion annually just from terminal subscriptions. Refinitiv, now acquired by the London Stock Exchange Group for $27 billion, holds about 20%. Market data revenue reported by ICE Data Services is $2.8 billion. Then come FactSet, S&P Global, Morningstar, and some regional players in niche services. The top four providers together control most of the flow of financial data from the producing institutions to the demand side.

All these companies operate similarly. Exchanges, trading firms, banks, asset managers—these institutions generate pricing data as a byproduct of their work. They sell or license this data to vendors. Vendors package and standardize the data, add analytical tools, and sell it at a significant markup, locking in long-term contracts and proprietary access methods to make switching painful. Bloomberg’s subscriptions lock you in for two years. Cancel early and you pay 50% of the remaining contract value. Everything about the Bloomberg experience is designed to make leaving feel harder than staying. Different keyboards. Different data formats. Even half of Wall Street’s messaging systems used for internal communication run on Bloomberg, meaning switching terminals also means losing your contacts.

This approach has lasted forty years because vendors solved a real hard problem: acquiring data from hundreds of sources, cleaning and standardizing it, and delivering it with low latency via global infrastructure. Bloomberg earned its position.

But blockchain is a better distribution mechanism. Maybe not for everything, and not yet fully scaled. But for connecting data-owning institutions with developers who want to build on that data, a programmable, public blockchain infrastructure is structurally superior to proprietary terminals with two-year contracts. By turning data into permissionless, self-service APIs, you provide any on-chain developer with access that has no switching costs. That’s what Pyth is doing.

Euronext, Exchange Data International, Fidelity Investments, OTC Markets Group, the Singapore Exchange’s Forex department, and Tradeweb are now publishing their proprietary market data directly on-chain via Pyth’s new data marketplace.

Euronext Forex: spot currencies and precious metals. FX rates used in actual global trading.

Fidelity: ETF valuation and fixed income data. Data used daily by institutions to mark portfolios to market.

Tradeweb: intraday ETF pricing. Real-time valuations from one of the largest electronic trading platforms.

OTC Markets Group: over-the-counter securities. A market that barely exists in DeFi data today.

Singapore Exchange Forex: Asian currency pairs. The largest FX market by volume but with minimal on-chain coverage.

Together, these six institutions cover a large portion of asset classes that DeFi has never reliably built, because the data feeding these assets isn’t institutional-grade.

Why Data Must Come Before Assets

Everyone in crypto has been talking about tokenization for two years: tokenized treasuries, tokenized bonds, tokenized stocks. The entire discussion assumes the hard part is putting assets on-chain.

But the hard part is the data. Before you can trade tokenized treasuries in DeFi protocols, you need to know what they’re worth right now, accurate to the second, with the same precision as when Goldman Sachs’ trading desk prices this tool. Before you can build lending protocols around real-world assets, you need a continuous source of prices that you can run reliably—coming from actual market-making institutions, not scraped from websites and updated every few minutes.

DeFi protocols need accurate, real-time traditional financial data to create derivatives, loans, and structured products, but historically they’ve relied on limited or slow data sources. That’s why, from its inception to today, DeFi has mostly been crypto-to-crypto. The data feeding these products hasn’t been reliable enough, fast enough, or coming from institutions credible enough in compliance conversations.

Pyth Pro is Pyth’s institutional subscription tier launched in September 2025, providing 1-millisecond latency price feeds across more than 2,200 tools. Polymarket integrated Pyth Pro in April 2026 to settle new markets for traditional assets including major stock indices, commodities, and U.S. equities, replacing manual or exchange-specific data with standardized data sources aggregated from more than 125 trading firms. Hyperliquid now uses Pyth’s price feeds to run perpetual contracts for oil and gold. Data quality is reaching a level where you can build serious financial products on top of it without having to apologize.

The tokenization wave needs this layer to operate at scale. Without a reliable fixed income price feed, you can’t reliably build fixed income products on-chain.

Oracle Wars

The original oracle problem in crypto was simple: smart contracts live on-chain, prices live off-chain. Something is needed to connect the two. Chainlink has been the dominant oracle for most of DeFi’s history. It solves this problem by running a large independent node network: these nodes pull prices from third-party sources (exchanges, aggregators, data APIs) and submit them on-chain. Many independent sources, many independent nodes, reasonable decentralization, acceptable latency.

From the start, Pyth took a different approach: it directly went to institutions that are actually trading. Today, more than 120 institutions publish data via Pyth, including global exchanges, trading firms, and market makers. Jane Street isn’t describing Bitcoin prices to Pyth secondhand—it’s becoming a publisher. The data comes from the source, not from someone describing the source.

This is faster, more accurate, and more directly bound to real market activity than aggregated price feeds. In structural terms, it’s also more centralized: a smaller club of publishers who know each other and verify their own data. Pyth has staking and slashing mechanisms designed to create economic incentives around accuracy. But a better way to put it is that Pyth chose speed and data quality instead of maximizing decentralization. For institutional finance, that may be the right trade-off.

The Cost of Centralization

Pyth’s creation involved significant participation from Jump Crypto, which played an important role in the events in 2022—events that most people in crypto are reluctant to revisit. The publisher network is a small club: institutions basically know each other and verify each other’s data. Staking and slashing mechanisms create economic incentives for accuracy, but Pyth is faster and higher quality than what came before, and also more centralized than what marketing suggests. You’re not replacing a commons for a monopoly. You’re replacing one centralized system with another—one that just happens to run on the blockchain.

The PYTH token reached an all-time high of $1.20 in March 2024. It now trades around $0.046, down about 96% from the peak. The obvious reason is: using Pyth data doesn’t require holding or buying PYTH. The network can grow significantly while the token stays range-bound. This is a known issue, and Pyth’s reserve plan is trying to address it by allocating a portion of protocol revenue to buy back PYTH in the open market.

The End of the Tollbooth

Getting data from the producing institutions to the desks of those who need it requires hardware, proprietary networks, sales relationships, and ongoing support. Bloomberg solved all of that and charged accordingly. Since data producers didn’t have an alternative distribution mechanism, they sold data to middlemen, and the middlemen kept the profits. Blockchain removes that particular friction—not analytics, not workflows, not keyboards. It’s simply that someone must move the data from one place to another and charge for that privilege.

But Bloomberg sells workflows. Terminals, keyboards, messaging systems, analytics, support teams. Traders build entire careers around it. Pyth sells none of that. It’s a data layer inserted into the protocol. The only overlap is the underlying data itself—and that part has just shifted.

This is important because if Fidelity publishes its ETF valuations on-chain, any developer anywhere can read that data without negotiating licensing agreements, without paying $32k annually, and without waiting for standardized formats from vendors. Data becomes programmable infrastructure rather than proprietary products. Institutions retain control over what they publish and retain attribution rights. The middlemen’s job—moving data from source to user—becomes unnecessary.

These six institutions are choosing Pyth as their primary distribution channel, which is a commitment different from just a pilot. Pilots get shut down when the people advocating for them change jobs. The main distribution channel becomes an operational dependency.

Tokenized bonds, tokenized stocks, tokenized everything. Most of these are still months or years away from meaningful scale. But the raw materials that make real-world asset products possible in DeFi are now available without contracts, without terminals, and without two-year minimum commitments.

Michael Bloomberg spent months delivering free coffee in the hallways of Merrill Lynch because the data he needed was locked inside institutions he had no reason to give it to. He built his entire business on that friction.

The tollbooth won’t disappear overnight. Every monopoly in data distribution ends the same way—not with fights, laws, or revolutions, but because someone somewhere asks: why should I pay for something I already own?