Why the world’s biggest traders are betting on blockchain data

All that could become reality thanks to a project backed by some of the world’s biggest quantitative and high-speed trading firms.

Pyth, the brainchild of Chicago-based Jump Trading Group, is a blockchain-based technology service that seeks to provide free, real-time data for a variety of crypto projects. It is loosely overseen by the Switzerland-based Pyth Data Association, whose members include heavyweights of electronic trading such as Jump, Jane Street Capital LLC, Susquehanna International Group LLP and Virtu Financial Inc. So far, they aren’t running Pyth to make a profit.

Collectively, the firms behind Pyth trade hundreds of billions of dollars of assets daily. Earlier this year, they began streaming price data from some of their trades in stocks, currencies, commodities and cryptocurrencies into Pyth’s network.

Pyth’s data is publicly available and developers from around the world can simply write code to connect to it. They can use the data to create new applications in the fast-growing area known as decentralized finance. DeFi, for short, encompasses efforts to take traditional financial activities—like trading and lending—and automate them with software, with no central intermediary overseeing transactions.

For DeFi to succeed, it needs data—and that’s where services like Pyth come in. The creators of Pyth hope to unseat rival services to become the leading source of financial data for DeFi projects. The idea is to provide fast, accurate data that could enable innovative mashups of traditional and crypto markets.

Proponents say DeFi will cut out middlemen and benefit the world’s unbanked population—1.7 billion people who lack access to financial institutions, according to the World Bank—by providing banklike services to anyone with a mobile phone. For instance, people seeking loans could turn to DeFi lending platforms. Like banks, such platforms take money from depositors and lend it out to borrowers. But instead of bankers approving loans, they have “smart contracts,” or computer code that authorizes transactions.

“You can have the whole world go on strike or all the banks could shut down, and smart contracts would still do what they’re programmed to do,” says Mike Cahill, head of European business development for digital assets at Jump, and a director of the Pyth Data Association.

DeFi has enjoyed extraordinary growth: Crypto investors have posted $250 billion worth of assets as collateral in various DeFi projects, up from $18 billion a year ago, according to data provider DeFi Llama. Still, there are many impediments to broader adoption. For one thing, DeFi is risky. Losses from thefts and fraud in DeFi have exceeded $10 billion this year, according to Elliptic, a blockchain analytics firm. Since DeFi deposits aren’t backstopped by the government—in contrast to conventional bank accounts—investors have little recourse if they lose money to criminals.

Regulators have yet to establish clear policies on DeFi. But some officials are pushing for tighter oversight. Securities and Exchange Commission Chairman Gary Gensler has said DeFi isn’t immune to regulation and warned that U.S. securities laws apply to digital tokens that track stock prices. That could inhibit the creation of markets based on Pyth’s stock data.

Here’s how Pyth data gets used today. Consider a DeFi lending platform. Borrowers on such platforms often want loans in stablecoins—digital coins pegged to traditional currencies like the dollar. To get the loan, the borrowers must post collateral. They often post the collateral in a more volatile cryptocurrency, like bitcoin, in a pool of funds overseen by a smart contract. If the price of bitcoin crashes, the smart contract automatically liquidates the collateral. This ensures that depositors—the people who provided the stablecoins in the first place—are kept whole.

To make all that work, the smart contract must monitor the price of bitcoin using a third-party source of data called an “oracle.” Oracles allow smart contracts to get external information needed for transactions. Right now oracles are mainly used to stream crypto prices, but they could provide information on almost anything, from stock prices to weather data to election results.

A number of oracle networks aggregate data from multiple sources and provide it to DeFi projects. The biggest is Chainlink, whose creators say it supports over $80 billion worth of assets tied up in various smart contracts. Its data sources include the Associated Press, which agreed in October to provide data on the economy, sports and elections through Chainlink.

Pyth is undercutting rivals with free data—unlike Chainlink, for instance, whose oracles cost money to use. Pyth’s backers say it has better technology too. It’s based on the speedy Solana blockchain, which allows Pyth to update prices approximately every four-tenths of a second; the bigger Ethereum blockchain, home to some rival networks, takes up to 15 seconds to update. Pyth supporters also say its price data will be more reliable because it comes from seasoned trading firms with decades of market experience.

Still, Pyth has had embarrassing stumbles since going live in August. In September, its underlying Solana blockchain suffered a 17-hour outage. Days later, a glitch caused the price of bitcoin reported by Pyth to briefly crash by around 90%.

Fallout from the incident was limited because Pyth is still in its early stages, but such snafus can potentially have painful consequences. “If the oracle doesn’t do what it’s supposed to do, people lose hundreds of millions of dollars in seconds,” says Chainlink co-founder Sergey Nazarov.

Some trading firms are backing Pyth because they want to shake up the traditional market-data business. The NYSE and Nasdaq Inc. make money by aggregating price quotes and trades from their markets and selling that data to brokers and traders. Vendors such as Bloomberg LP also profit by reselling that data to Wall Street. The firms behind Pyth are major buyers of such data. They have long complained that it’s too expensive, and that they shouldn’t be stuck with steep costs for market data when their own trading activity generated the data in the first place.

“Each of us are frustrated by this market-data complex,” says Virtu Chief Executive Douglas Cifu. He predicts that Pyth could become a low-cost source of financial data. “There are market-data repackagers and resellers that have made a lot of money for a long time that are going to feel pricing pressure from Pyth and similar networks,” Mr. Cifu says.

Firms like Virtu are currently feeding data to Pyth for free. But the service could eventually shift to a model where the firms get rewarded—potentially in a new Pyth digital currency—for their data.

Other traders say they are backing Pyth to support innovation. DeFi will be as transformative to markets as the shift from floor trading to electronic trading in the 1990s and 2000s, predicts Ari Rubenstein, chief executive of GTS, a trading firm that contributes data to Pyth.

“Pyth is just a building block,” Mr. Rubenstein says. “We don’t know the direction it’s going to go, but it’s definitely going to sprout many different things.”