From "Speculative Asset" to "Next-Generation Financial Infrastructure," Is Crypto Building a New TradFi World?

To be honest, over the past few years, most of the crypto industry's narratives can be boiled down to one almost obsessive question:

"What's the next asset that will go up?"

Starting with DeFi Summer, NFTs, public chains/L2s, restaking, memes, and even AI tokens have taken turns in the spotlight. Each narrative has its own technological logic and market backdrop, but ultimately, everything comes back to price performance. Even though stablecoins, wallets, and cross-chain bridges have clear utility as tools, market attention on them is often limited to how much trading and speculation activity they can support.

But entering 2026, a series of changes across different tracks have begun to emerge intensively within the same window of time:

• The total market cap of stablecoins has reached approximately $300 billion, entering a high plateau phase of penetration into global payment networks;
• DTCC has completed its first batch of asset tokenization conversions in a production environment, with plans to officially launch related services in October;
• Prediction markets are moving from crypto-native products into brokerages and regulated exchanges;
• AI Agents, meanwhile, are starting to autonomously purchase data, model calls, and digital services using stablecoins;

On the surface, these changes seem unrelated. But when placed together, a more complete thread emerges: The issuance, custody, trading, payment, and settlement capabilities that the crypto industry has accumulated over the past decade are gradually opening up—from serving crypto assets themselves to a broader range of financial activities and the machine economy.

In other words, Crypto hasn't shaken off speculation, but beneath the speculative market, it is growing an increasingly complete layer of infrastructure.

1. Why Are These Breakthroughs Happening Almost Simultaneously?

Objectively speaking, RWA, stablecoins, prediction markets, and AI Agents are not gaining attention at the same time because of some sudden hotspot narrative.

The real reason is that the different components required for a new financial infrastructure, after years of independent development, have finally begun to connect with each other.

1. Stablecoins Turn Money into an Invocable Interface

First are stablecoins. They are nothing new, but the role they play is changing.

As we all know, early stablecoins were primarily used for exchange pricing, on-chain risk hedging, and settlement of crypto asset trades, with the vast majority of funds still circulating within the Crypto ecosystem. Today, a growing number of issuers, banks, payment institutions, and fintech companies are starting to use stablecoins for merchant collection, global payroll, corporate payments, treasury management, and cross-border settlement.

According to data disclosed by Circle in the first quarter of 2026, the annualized transaction volume of its network, based on a nearly 30-day trading period at the time, reached approximately $8.3 billion. Partner Nium's payment network covers over 190 countries and regions. This also means that stablecoins here are no longer just "on-chain dollars," but a form of currency that can be directly invoked by software.

They can be transferred 24/7, embedded in programs, automatically released based on conditions, and can directly become settlement assets after a transaction is completed. For internet applications, sending a stablecoin is increasingly becoming like calling a payment API—no need to understand correspondent banks, clearing times, or cross-border account systems, only need to confirm the amount, address, and execution conditions.

This is also the key change of stablecoins shifting from a Crypto trading tool to payment infrastructure.

2. RWA Turns Assets into Programmable Objects

If stablecoins solve the problem of "what currency to settle with," RWA solves "what assets can be traded and settled."

Past RWA products were mostly concentrated in U.S. Treasuries, money market funds, and private credit, with the core value primarily being giving Crypto users access to off-chain asset yields. But starting last year, TradFi infrastructure has visibly begun to proactively bring securities registration, custody, trading, and settlement processes on-chain.

On July 15, DTCC completed a tokenized asset trading test in an actual production environment, with over 30 traditional financial institutions and digital asset companies participating, and plans to officially launch related services in October. Unlike simple asset mapping, DTCC plans for tokenized securities to retain the ownership, investor protections, and entitlement arrangements corresponding to traditional securities.

Even earlier, the U.S. SEC approved Nasdaq in March to allow eligible listed securities to be traded in tokenized form. Tokenized stocks use the same CUSIP as traditional stocks, have the same substantive rights, and continue to trade under the existing market system and securities rules.

This is fundamentally different from simply issuing a "stock mapping token." It means that on-chain assets are beginning to connect with real ownership, custody relationships, corporate actions, and legal rights, and are starting to carry parts of the lifecycle of traditional assets.

Therefore, as this layer of connection is gradually established, blockchain is not just creating new assets—it is also starting to carry parts of the operational process of traditional assets.

3. Prediction Markets Turn Future Information into Prices

What prediction markets fill in is the information and price discovery layer.

Stocks trade on a company's future cash flows, bonds trade on credit and interest rates, and prediction markets trade on the probability of an event occurring. Election results, interest rate decisions, sporting events, corporate events, and even product launch dates can all be compressed into a continuously changing market price (Further reading: World Cup Fever, Prediction Markets Take the Table: How Are Polymarket and Others Breaking into the Mainstream?).

Robinhood disclosed that its prediction market business attracted over 1 million users in its first year, with cumulative trading of approximately 9 billion contracts. It has also acquired a CFTC-regulated exchange and clearing infrastructure. From an infrastructure perspective, prediction markets provide a capability that traditional financial markets struggle to cover at scale—aggregating dispersed information into a probability that can be read in real time.

4. AI Agents Are Becoming New Economic Actors

Stablecoins and RWA solve the problems of assets and capital. The new variable brought by AI Agents is who initiates economic activity.

Traditional software can only execute operations according to pre-set processes. Agents, on the other hand, can understand goals, find services, compare prices, and make decisions within certain permissions. When an Agent can purchase APIs on its own, it is no longer just an information tool—it becomes a new type of economic actor.

The problem is, a large number of Agent payments may only be a few cents or even less, which is difficult for traditional card networks with fixed fees to cover. Their fixed fees, settlement cycles, and identity verification processes are not naturally suited for high-frequency, micro-transaction, automated machine payments.

This is precisely the scenario where stablecoins and low-cost blockchains can play a role.

Coinbase has integrated x402 and stablecoin wallets into AWS Bedrock AgentCore, allowing enterprises to set budgets and governance rules for Agents. Google's Agent Payments Protocol uses cryptographically signed authorization credentials to record what users allow Agents to purchase, the spending limits, and who specifically initiated the operation (Further reading: A Panorama of Crypto AI Protocols: Starting from Ethereum’s Main Battlefield, How to Build a New Operating System for AI Agents?).

2. So What Does the Next-Generation Financial Infrastructure Already Have?

Overall, these threads are happening at the same time precisely because they are different components of the same system.

Stablecoins turn money into an API, RWA turns assets into programmable objects, prediction markets turn future information into prices, and AI Agents allow software to directly participate in asset exchange for the first time.

But it's worth noting that judging whether Crypto is becoming infrastructure doesn't mean judging whether speculation still exists in the market. Just as stocks, forex, and commodities markets also have massive speculation, the more important standard is whether external businesses and users are beginning to rely on a technology to complete work that was previously difficult, too costly, or inefficient.

By this standard, Crypto and Web3, as next-generation financial infrastructure, have already initially formed multi-layered capabilities.

The first layer is asset issuance and mapping.

What can now be brought on-chain is no longer just native tokens. Stablecoins, Treasuries, money market funds, private credit, gold, fund shares, and stocks have all seen different forms of on-chain products emerge. The significance of bringing various assets on-chain is not just putting a certificate in a wallet.

When assets can be recognized by smart contracts, they can directly enter collateralization, lending, trading, treasury management, and automated investment processes. Operations that were previously dispersed across registrars, custodians, broker-dealers, and clearing systems have the opportunity to be compressed into a more unified execution environment.

The second layer is 24/7 payment and settlement.

Traditional cross-border payments typically require multiple correspondent banks and are subject to business hours, account system, and regional network limitations. Stablecoins can complete near-instant, 24/7 value transfer under a unified asset standard.

J.P. Morgan stated that Kinexys has processed over $4 trillion cumulatively since its launch, with average daily transaction volumes exceeding $7 billion. It has also expanded blockchain deposit accounts to multiple currencies including USD, EUR, GBP, JPY, HKD, SGD, and RMB.

Simply put, on-chain settlement doesn't necessarily require all funds to be converted into publicly issued stablecoins. In the future, there may be a coexistence of bank deposit tokens, regulated stablecoins, central bank digital currencies, and on-chain commercial bank money. What they have in common is that funds can be read and scheduled by programs, and synchronized with asset delivery.

The third layer is continuous trading and price discovery.

Crypto has already proven that markets can operate 24/7 and that smart contracts can automatically match orders and manage liquidity.

This capability is being brought to more asset classes. Tokenized securities can shorten the time between trading and settlement, and prediction markets can provide probabilities for events that traditional finance finds difficult to price directly.

In the future, an enterprise can not only hold on-chain money market funds but also automatically adjust cash positions based on changes in interest rate prediction markets. AI Agents may also simultaneously read asset prices, event probabilities, and liquidity conditions before deciding whether to execute a trade.

At that point, the market no longer just provides quotes for people to view, but a set of real-time signals that can be directly invoked by software.

The fourth layer is identity, permissions, and authorization.

Financial activities aren't just about transferring assets; they must also answer a series of questions: Who initiated the transaction? Who has the authority? How long does this authorization last? What's the spending limit? Who is responsible if something goes wrong?

Early Crypto answered these questions primarily through private keys—holding the private key meant having full control. But as enterprises, institutions, and AI Agents enter the on-chain space, a single private key clearly cannot meet complex permission management needs.

Google AP2 uses verifiable authorization records to capture user intent; Visa is building Agent identity directories, credentials, and scoring mechanisms; Mastercard's Agent Pay for Machines is attempting to provide identity authentication, permission settings, transaction, and settlement capabilities for machines.

Account abstraction, Passkeys, multi-sig wallets, session keys, and spending policies also allow users to delegate limited permissions to an application or Agent, rather than handing over full account control.

This means the role of wallets may also change. Future wallets won't just store assets and private keys; they will also need to manage user identity, institutional credentials, Agent permissions, spending budgets, and authorization records, becoming a control interface for users entering the on-chain economy. (Further reading: A Decade of Web3 Wallets: As the AI Inflection Point Accelerates, a New Map of Crypto Users).

The fifth layer is connection with real-world legal and regulatory systems.

Whether a financial system can become true infrastructure depends not only on whether the technology can run, but also on whether real-world law recognizes the results of transactions.

In January 2026, the U.S. SEC issued a statement on tokenized securities, clearly distinguishing between tokenized securities issued directly by issuers, tokenized interests formed by third-party custody of underlying assets, and on-chain products that only provide synthetic price exposure. This distinction is very important because while all these products may look like "on-chain stocks," the legal rights holders actually possess can be completely different.

The CLARITY Act attempts to further delineate the regulatory boundaries between the SEC and CFTC, establishing clearer rules around digital asset issuance, trading platforms, software developers, DeFi, and investor protection. The Act remains controversial and has not yet been enacted into law. But the regulatory focus has gradually shifted from "whether Crypto should be allowed to exist" to "who can issue, who is responsible for custody, and what rules apply to which assets."

This shift itself is an important signal of infrastructuralization, because only when participants can roughly assess their legal responsibilities will banks, brokerages, asset management institutions, and payment companies be willing to make long-term commitments, rather than just running isolated pilot projects.

3. The Inevitable Path from "Speculative Market" to "Infrastructure"

Is Crypto moving from a speculative market to infrastructure?

The answer is yes, and this process is irreversible—but it's not an either-or replacement process.

Crypto won't suddenly lose its speculative nature just because of stablecoin payments and RWA growth. More accurately, it is building an execution system beneath its existing trading market that can be used by real assets, traditional institutions, and intelligent software alike.

This change is first reflected in the expansion of funding sources within the industry. In the past, much protocol revenue came from leveraged trading, asset issuance, liquidations, and on-chain capital cycles. Now, a second type of cash flow from external economic activity is starting to appear on-chain: enterprises using stablecoins for cross-border settlement, funds distributing and managing assets through on-chain channels, software paying for APIs per use, and Agents automatically paying for data and model fees.

Second, the participants in the on-chain economy are also expanding. The typical user of the past was a human trader sitting in front of a screen clicking "Confirm" and "Sign." In the future, a large volume of on-chain interactions may be initiated by enterprise systems, payment programs, and AI Agents. Humans set the goals, boundaries, and permissions, while software handles the specific execution.

Regulatory discussions are also changing. The past debate was mainly about whether Crypto should be incorporated into the existing financial system. Now the question is gradually becoming how to delineate regulatory boundaries, protect investors, constrain intermediaries, while preserving space for self-custody and open software.

However, from "it can run" to "worth relying on long-term," crypto infrastructure still has a long way to go.

First, on-chain confirmation does not equal legal finality. Questions such as who custodies the assets behind the tokens, whether investors can recover assets if the issuer goes bankrupt, whether different jurisdictions recognize on-chain ownership transfers, and whether token holders have dividend rights, voting rights, or only price exposure—these cannot be solved by smart contracts alone.

Agentic payments also face liability boundaries. When an AI Agent executes a wrong transaction due to misinformation, prompt injection, or model hallucination, who is liable—the user, the model provider, the wallet, or the merchant? Mature handling mechanisms are still lacking. Wallets of the future need to solve not just how to let Agents pay, but also how to limit which assets they can use, who they can pay, spending limits, and how to pause and revoke permissions in case of anomalies.

At the same time, the more assets and networks there are, the more prominent the problem of liquidity fragmentation may become. The same stablecoin, fund, or security can exist across different public chains, bank ledgers, and permissioned networks, yet may not circulate freely. In the next phase, establishing unified asset standards, cross-network communication, and secure settlement mechanisms is more important than continuing to issue more assets.

Privacy is also an unavoidable part of institutional adoption. Public blockchains are good for verification and auditing, but enterprises won't want to expose all their customers, suppliers, payroll, and fund flows publicly. How to leverage zero-knowledge proofs, selective disclosure, and on-chain credentials to preserve necessary privacy while meeting compliance requirements will directly determine how far on-chain finance can go.

The more fundamental issue is that blockchain can improve transaction and settlement efficiency, but it cannot automatically create credit. Credit loans, insurance, accounts receivable, default handling, and liquidity support in a real financial system all require complex risk control, legal, and liability frameworks. Prediction markets won't automatically solve problems like insider information, insufficient liquidity, or outcome adjudication just because prices are public.

Therefore, today's Crypto is more like having set up the basic framework for assets, currency, trading, and settlement