Tokenized Money Needs a Price Layer

What a new wave of independent research reveals about the next market data supply chain

Tokenization is moving from financial market experiment to institutional infrastructure. That is the central signal running through Celent's recent research, including its report Tokenization Reaches the Real World, which examines the progress of real world asset tokenization through the lens of an asset manager, with money market funds identified as the most mature use case so far.

Read that work alongside Celent's research on tokenized money and treasury services and its study of investment data ecosystems, and a pattern emerges. The next phase of tokenization will be defined less by whether an asset can become digital, and more by what happens after it does. Treasury teams still need to manage liquidity. Asset managers still need to value portfolios. Risk systems still need current market inputs. Institutions still need data that can move across systems, venues, and jurisdictions without friction.

Put simply: tokenized money needs a price layer. This is where Pyth comes in, and it is worth walking through why.

Tokenization is becoming an operating model

The first wave of tokenization focused on the asset itself: how to represent a fund, a deposit, a security, or a payment instrument on programmable infrastructure. Institutional adoption depends on something broader than that.

A tokenized money market fund still needs valuation, liquidity monitoring, portfolio reporting, and risk controls. A tokenized deposit still needs to interact with foreign exchange markets, collateral systems, treasury platforms, and payment workflows. Turning an asset digital creates flexibility. Whether that flexibility produces real value depends on the financial processes wrapped around it.

That is why Celent's focus on money market funds carries weight. It places tokenization inside an existing institutional workflow, rather than treating it as a standalone technology project. The firm's related work on programmable money points in the same direction, identifying liquidity optimization, real time cash mobility, and FX risk management as the treasury opportunities that matter most, and arguing that programmable payments executed through smart contracts could become a genuinely decisive bank capability within five years.

The practical takeaway: tokenization earns its place when it improves how institutions manage money, assets, and risk, not before.

Programmable money needs programmable prices

A treasury system can move money automatically. It still needs to know when and why to move it.

Picture a treasury engine that converts currencies the moment a liquidity threshold is crossed, a collateral system that rebalances when an asset's value shifts, or a payment instruction that fires when a defined set of market conditions is met. Every one of those workflows runs on a current price. Programmable money supplies the movement of value. Market data supplies the judgment that makes the movement intelligent, FX rates, interest rates, equity prices, commodity prices, and the other inputs that let software evaluate risk and act inside defined limits.

That creates a real design requirement for financial infrastructure. The data layer underneath programmable money has to be available in real time, usable across asset classes, transparent about its sources and confidence, compatible with automated systems, and available wherever the workflow actually runs, not just during the hours a traditional exchange happens to be open.

This is the connective tissue between tokenized money and market data infrastructure that often gets skipped over. The future of treasury isn't only about faster payment rails. It's about pairing programmable value with continuously available information about the markets surrounding that value. For banks, asset managers, custodians, and fintech providers, the opportunity is building systems that respond to market conditions as they change, and that requires a price layer built for software driven finance rather than adapted to it after the fact.

Data distribution is becoming a strategic choice

Celent's work on investment data ecosystems makes a related point worth sitting with: financial data is now widely treated as a strategic asset, yet firms face genuinely complex data value chains, and no single provider model suits every institution.

That reframes the question institutions should be asking. It is no longer simply which vendor sells a particular dataset. It's how data should be sourced, governed, distributed, consumed, and combined across an organization that increasingly runs on both conventional and blockchain based rails.

For decades, the market data supply chain has followed a familiar shape. Financial institutions and trading firms generate valuable market information. Intermediaries package and redistribute it through closed systems. Downstream users consume the result through separate contracts, terminals, feeds, and integrations, each one its own negotiation. That model was built for an earlier generation of financial technology, and the rise of tokenized assets, automated treasury, AI driven analysis, and always on markets is creating real demand for something different: market data that can move directly into applications, risk systems, portfolio tools, and programmable financial products, without every user rebuilding the same integration from scratch.

This is where Pyth Pro fits into the picture.

Pyth's model sources market data directly from institutional contributors and distributes it through a single layer across asset classes and geographies, rather than routing it through the usual chain of intermediaries. In a recent solution brief, the analyst firm Celent described Pyth's architecture as a single source of truth model built around three connected products: Pyth Pro for subscription based price access, Pyth Indices for continuous benchmark style products, and the Pyth Data Marketplace for distributing proprietary institutional datasets. Data flows in from more than 120 exchanges, market makers, and financial institutions, including familiar names like Fidelity Investments, Revolut, and Jane Street, and flows out through a single integration rather than dozens of separate ones.

Celent's assessment lands on a straightforward strategic choice facing data owners. They can keep treating distribution as a closed, downstream process, the way it has worked for decades, or they can participate in infrastructure built for broader reach, machine consumption, and programmable financial workflows. That isn't a traditional finance versus digital finance question. It's a question of whether the data supply chain is actually designed for how financial systems operate today.

What that looks like in practice

The clearest evidence for this shift isn't theoretical. Celent's briefing on Pyth points to Hyperliquid as a working example: a venue that ran into the familiar limits of traditional market infrastructure, restricted trading hours, licensing constraints, fragmented data access, and used Pyth Pro alongside Pyth's HIP-3 infrastructure as the pricing layer for its real world asset markets. That gave Hyperliquid continuous, institutional grade pricing across equities, commodities, FX, metals, and digital assets through one framework, and Celent credits it with helping the venue evolve from a crypto native exchange into a genuinely 24/7 global markets platform.

It's a useful case study precisely because it isn't about crypto adopting crypto infrastructure. It's about a trading venue solving an ordinary market structure problem, continuous markets need continuous pricing, with a data layer designed for that reality from the start.

The next market data layer

Market data infrastructure built for tokenized money needs to connect institutional quality with software native delivery. Based on where the research points, that comes down to four things.

Direct sourcing. The closer data sits to the institutions and venues actually forming prices, the clearer its provenance. Pyth's publisher model, sourcing directly from exchanges, market makers, banks, and trading firms rather than acquiring data through downstream vendor chains, is built around exactly this.

Cross asset coverage. Treasury and investment workflows rarely stay inside one market. They combine currencies, rates, equities, commodities, fixed income, funds, and derivatives in the same decision. Pyth's own network update from earlier this year reported more than 138 participating institutions and a catalog above 3,500 feeds, including nearly 1,900 equity feeds, evidence of a network expanding well beyond its crypto native roots into broader market data territory.

Continuous availability. Financial software increasingly runs across time zones and outside traditional market hours, and its data infrastructure has to keep up. This is the specific gap Pyth Indices was built to close, blending on-chain and off-chain inputs into 24/7 composite and single asset benchmarks, with early traction concentrated in commodities and equities.

Programmability. Data has to be usable by APIs, automated workflows, risk engines, financial applications, and AI systems, without forcing every user to rebuild the underlying distribution architecture themselves. On the institutional side, the Pyth Data Marketplace extends that same principle to proprietary datasets, giving institutions a way to distribute specialized data such as reference data, fixed income pricing, and economic indicators through one technical and commercial connection instead of building a separate pipe for every counterparty.

Convergence, not replacement

None of this reads as a wholesale replacement story, and it shouldn't. Traditional institutions still depend on decades deep historical time series to power risk and valuation models, and on established providers to support reconciliation, regulatory reporting, and integration with market infrastructure that remains only partially blockchain native. That's precisely why analysts covering Pyth's model tend to frame it as additive rather than replacement led: complementary to incumbents like Bloomberg, LSEG, ICE, SIX, and S&P Global, strongest where it can plug into new products, new customer segments, and new distribution rails, rather than displacing an established data estate overnight.

That framing matters because it's honest about where the real opportunity sits. The market data infrastructure built for tokenized money doesn't need to win by tearing out what already works. It needs to become difficult to leave out of the architecture of any institution building always on, tokenized, or agent enabled financial products, while continuing to strengthen the depth, historical coverage, and enterprise controls that regulated institutions will keep expecting as adoption grows.

Tokenized money needs prices to operate intelligently. Institutional data needs modern distribution to reach the systems that actually consume it. The next market data supply chain will be built around both, and the research increasingly suggests that build is already underway.

Want the full picture? This piece draws on an independent solution brief that Celent, a division of GlobalData, published on Pyth's data distribution model, including a deeper look at Pyth Pro, Pyth Indices, and the Pyth Data Marketplace, along with Celent's assessment of where the architecture still has room to grow.

To explore the product directly, visit Pyth Pro.