San Francisco Stablecoin Week 2026: The Emergence of a 3D Ecosystem

Author: Charlie, Venture Partner @ Generative Ventures

Last week was Stablecoin Week in San Francisco, bringing together stablecoin industry leaders from around the world.

After going around, I have an increasingly strong feeling: everyone talks about "stablecoins," but they don't mean the same thing anymore.

In some circles, people are talking about Circle's stock price, financial reports, and valuation reassessment.

In some circles, people are talking about agents, wallets, authorization, payment protocols, and whether AI really needs a card.

In other scenarios, the discussions are less glamorous, focusing more on Brazil, Europe, corporate treasuries, local fund inflows and outflows, non-U.S. dollar liquidity, and the real-world challenges in cross-border fund transfers: each jurisdiction has its own compliance logic, banking restrictions, and clearing bottlenecks—money is never as simple as “just sending it over.”

On the surface, everyone is talking about stablecoins, but in reality, they are discussing three different businesses.

This was also my biggest takeaway from this year’s San Francisco Stablecoin Week: by 2026, stablecoins will no longer be a single narrative, but will gradually evolve into a three-dimensional coordinate system.

X-axis: Agentic Commerce.
The Y-axis represents RWA and Credit.
Z-axis is On-Chain FX.

Stablecoins are, of course, the underlying form of currency, the common monetary substrate.

But the businesses that have actually been built have clearly diverged.

Circle: This rally has made this change clearer.

On a financial reporting level, its data is certainly robust: as of the end of 2025, USDC's circulating supply reached $75.3 billion, a 72% year-over-year increase; in Q4 2025, on-chain transaction volume for USDC reached $11.9 trillion, up 247% year-over-year.

But what’s more worth noting is that it is consciously repositioning itself from a “stablecoin issuer” to a broader internet financial infrastructure: regulated stablecoins, tokenized money market funds, developer tools, Arc, and the Circle Payments Network—taken together, these form not just a single coin, but an entire stack.

Old perspectives are no longer sufficient. Stablecoins are still the same asset, but the businesses built around them are now entirely different.

I started writing about agentic commerce about a year ago. At that time, the term wasn’t this popular yet, and when people heard the concept, their first reaction was still “AI buys things for you” or “AI helps you browse Taobao.”

But I’ve always felt that what’s truly worth watching isn’t this.

The most important change in agentic commerce is not that the shopping experience becomes smarter, but that a new type of actor has emerged in commercial systems: software carrying delegated intentions.

This statement may sound a bit abstract, but its consequences are very concrete.

In the past, e-commerce centered on checkout. Whoever created a shorter funnel and a smoother payment experience had a better chance of winning.

But as agentic commerce moves forward, the question shifts from "how to pay" to "who can pay".

Who authorizes the agent? What are the boundaries of this authorization? How much can it spend, and in which scenarios?

How do I bind my identity? How is risk control handled? How are disputes resolved? How is auditing documented?

If these issues are not resolved, the payment channel itself is not that important.

So for the past few months, I’ve been telling a lot of friends,Circle’s deepest moat isn’t just reserve yields, not just distribution, not just regulatory advantages—it’s whether it can, in a more fundamental sense, become x402-native.

This is not about "owning x402," since x402 was not invented by Circle.

I’m referring to something else: if the future internet truly develops a machine-native payment layer, could Circle become the default dollar, the default wallet, and the default settlement asset within that payment layer?

This is a very important distinction.

Because if you look at agentic commerce from today’s perspective, despite OpenClaw breaking through, it is still largely"AI placing orders on behalf of humans."

But what will actually erupt first may not be the most visible front-end scenario.

In many mature markets, agents still prefer to use existing card networks, bank transfers, or merchant credentials, as these channels are cheaper, more familiar, and easier to manage for disputes.

What stablecoins are likely to reach first is a deeper layer:

Machine-to-machine settlements, pay-per-use APIs, pay-per-call for content and data, low-value high-frequency payments, autonomous treasury actions, and software-native global fund flows.

In other words, the stablecoin that wins first in agentic commerce may not be the one that "lets AI buy you a coffee," but the one that enables software to spend money like software should.

This is also why I keep emphasizing this one phrase:

If the internet grows a machine economy layer, the strategic high ground has never been "launching a dollar-pegged coin," but rather becoming the wallet and settlement system that best aligns with machine behavior.

Not an extension of human checkout, but the beginning of software-native money.

Today, the focus of RWA remains the migration of familiar financial assets onto the blockchain.

The most typical, of course, is U.S. Treasuries. In addition, there are credit, commodities, funds, and an increasing number of attempts around tokenized equities.

Ultimately, most RWA projects today are reimagining existing financial products: making them more programmable, easier to distribute globally, and more efficient to settle.

So I don't believe Messari's InfraFi proposed this year is the main theme of current RWA.

But precisely because of that, I pay even more attention to it.

In the entire RWA landscape, InfraFi is one of the few that truly feels like something that grew organically from the on-chain world, rather than a traditional financial product repackaged and redistributed.

A tokenized Treasury is still, at its core, a Treasury.

A tokenized stock is still, at its core, a stock.

They are certainly important and will only grow larger, but their economic identity has not changed.

InfraFi is different.

It refers to a class of assets or cash flows that were previously difficult to standardize, continuously verify, or effectively financialize, but now, thanks to on-chain verification, programmable ownership, and continuous data streams, are beginning to become truly investable.

That's also why I place greater importance on it.

It’s not because it’s the largest today, but precisely because it’s still not large today.

But it may represent a new way of generating assets.

In this direction, EnergyFi, represented by Arkreen, is one of the cases I’m most interested in right now.

My interest in it lies not in the surface-level narrative of "energy + RWA," but in how it might demonstrate what a truly on-chain native asset class could look like.

In the past, many infrastructure-related cash flows were not without value, but rather too fragmented, too dispersed, too reliant on offline verification, and too dependent on post-hoc aggregation, making it difficult to turn them into assets that could be frequently monitored, continuously priced, and effectively financed. Often, it’s not the assets themselves that are flawed, but rather the weakness of their underlying factual foundation.

EnergyFi is precisely designed to address this issue.

If the underlying energy production, usage, and settlement processes can consistently generate trustworthy data;

If this data becomes a verifiable, callable, and auditable stream of facts, rather than being secondhand through monthly reports, audit summaries, or third-party metrics;

Then, what is financialized is not just a packaged right to income, but a cash flow system that can be continuously verified.

Why is this important?

Because what it addresses is not just energy, not just DePIN, but the core pain point of private credit and even private investing more broadly.

Recent events involving Blue Owl Capitalhave exposed many issues with private credit. On the surface, these appear to be structural, liquidity, and valuation problems, but at their core lies a very fundamental question: Is the underlying information you receive accurate, timely, complete, and continuously verifiable?

If not, it’s still garbage in, garbage out.

The underlying facts are vague, delayed, and filtered; no matter how sophisticated the upper structure is, it merely repackages opacity.

What makes things like EnergyFi worth serious attention isn't that they introduce a new concept, but that they may offer an alternative:not making the "report" look better, but turning the underlying operational facts themselves into continuously verifiable data objects.

This will directly alter underwriting, as well as monitoring, and even the due diligence logic for primary and secondary markets.

If this path succeeds, the significance of on-chain finance will be more than just “an additional distribution channel”—it will add a new layer of truth.

In this sense,I view EnergyFi as one of the potential foundations for future private investing, rather than a niche narrative.

Moreover, this line has a very practical overarching context: the energy bottleneck of AI.

Over the past few years, energy has often served as a backdrop to technological narratives.

But as AI enters the infrastructure race, energy is no longer just a supporting factor—it is re-emerging as one of the toughest constraints.

If AI remains one of the most important industry themes over the next decade, systems surrounding energy production, financing, verification, and cash flow securitization will eventually move from the margins to the center.

From this perspective, EnergyFi isn’t just adding a “green narrative” to RWA—it’s more like demonstrating in advance whether on-chain finance can truly penetrate the core infrastructure of next-generation financing.

During this trip to San Francisco, I think the most underestimated type of conversation actually comes from on-chain FX.

Perhaps precisely because this group is closer to real liquidity, real corridors, and real balance sheets, their statements tend to be more restrained and less likely to reduce the issue to a simple narrative of “faster payments” or “cheaper transfers.”

My biggest takeaway is that many people still view on-chain FX as an extension of cross-border payments.

But what’s truly difficult about this has never been “sending money faster”—it’s the market structure.

Forex has never been just an information transmission issue, nor just a settlement issue.

It is first and foremost a balance sheet, funding, and liquidity issue.

Blockchain excels at atomic settlement; the mature FX market excels at netting.

The former is clean, direct, and minimizes trust, but requires substantial capital; the latter has complex governance but is extremely balance-sheet-efficient.

If every transaction must be gross settled, capital will be tied up.

The market maker's efficiency will decline, the spread will widen, and depth will fail to build up.

So over the past week, in multiple on-chain FX discussions, I kept hearing one clear assessment: what’s truly holding back this market isn’t that contracts can’t be written, wallets aren’t good enough, or even just compliance issues,it’s capital efficiency.

This judgment is important because it will directly change how you view this sector.

Simply understanding on-chain FX as "on-chain currency exchange" can lead to an overly simplistic conclusion: if we just build more infrastructure, issue more native currency stablecoins, and list more trading pairs, the market will naturally follow.

But the reality is completely different.

A better way to understand it is to break it down into three levels.

The first layer is fiat-to-fiat: This is the largest market, but also the hardest to disrupt immediately, as traditional institutions are deeply entrenched here.

The second layer is fiat-to-stable and stable-to-fiat: This is actually the fastest-growing segment today, especially in emerging market scenarios such as treasury management, remittances, and corporate settlements.

The third layer is stable-to-stable: It’s not large today, but it’s likely the real endgame—because only at this layer does FX cease to be “using crypto to access the old system” and begins to resemble a native internet foreign exchange market.

This layered structure also makes the boundaries of opportunities clearer.

What was truly first transformed, as I mentioned in Airwallex’s founder on where stablecoins went wrong,wasn’t the deepest G10 interdealer market, but the long tail that the old system has always underserved: SMEs, exporters, cross-border platforms, freelancers, migrant corridors, and various participants without prime broker relationships or large balance sheets, yet who genuinely have ongoing cross-border needs.

This time, I heard several examples that were very representative.

A team mentioned that, on a treasury corridor in Latin America, Starlink has reduced settlement times from days or even weeks to approximately 35 minutes.

Another team launched in mid-2025, reached $1 billion in trading volume within six months, and achieved another $1 billion three months later, primarily serving emerging-market FX and regulated institutional demand.

These cases illustrate one thing: the growth of on-chain FX is not emerging from the center, but from the edges.

Instead of starting with the deepest markets, first make things smooth in areas that the old system has never served well.

More interestingly, these discussions have further convinced me of another point:On-chain FX is first and foremost not a supply-side infrastructure issue, but a demand aggregation problem.

I heard a phrase at the meeting that left a deep impression on me: start with the chicken, not the egg.

It means don't get caught up in designing a perfect liquidity market from the start—first, secure the real flow.

If you can aggregate real demand from neobanks, PSPs, remittance platforms, treasury software, or products with built-in distribution capabilities, market makers will naturally come.

Conversely, without a stable flow, adding another trading pair or another venue will hardly generate meaningful depth.

This perspective is especially important.

Because the winners in on-chain FX are unlikely to come from traditional exchanges; instead, they are more likely to emerge from the orchestration layer: these entities may not absorb all liquidity themselves, but they coordinate compliant deposits and withdrawals, aggregate demand, execute intelligent routing, and layer on netting and credit when appropriate.

Ultimately, once you acknowledge that the bottleneck of on-chain FX is capital efficiency, you're already close to credit.

If I had to pick the most underestimated line this week, it would be non-USD stablecoins.

Because it always sounds great on a PowerPoint: every country, every market, seems like it should have its own stablecoin.

Logically, that makes sense, but in reality, it’s a much harder business than dollar-stablecoins.

Why are USD stablecoins strong? Because in many countries, the US dollar is naturally a stronger store of value.

When currency volatility, capital controls, and banking limitations overlap, the dollar automatically becomes the default option. It’s no surprise, then, that today’s stablecoin world is centered around the dollar.

The issue with non-dollar stablecoins is that many people think "issuance" is the product, but issuance is actually just the easiest step.

The real challenge lies in local banking relationships, offshore liquidity, market makers, same-name pay-in/pay-out, integrating with local payment networks like SEPA and PIX, navigating regulatory requirements corridor by corridor, and delivering a user experience competitive with Wise, Revolut, local PSPs, and bank transfers.

More importantly, you must answer a question far more difficult than “Can you launch a token?”: Why should users hold this token on-chain?

This question is actually very sharp.

In many markets,the demand for the local currency is primarily not for value storage, but for payout purposes.

Users may still want to use the US dollar as a store of value, but they must revert to their local currency for payroll, invoices, supplier payments, taxes, and domestic spending.

This means that non-US dollar stablecoins must not focus on getting the market to “recognize this token,” but rather position themselves as a bridge: one end connected to offshore liquidity, the other linked to local payment systems and real-world business flows.

Brazil is a classic example.

What truly matters is not the fact that BRL has been tokenized, but whether it can bridge the onshore and offshore worlds in a market where cross-border transactions are inherently painful, order books thin out at scale, and holding fiat currency across borders is heavily restricted.

From this perspective, the token itself is less important; what matters is the entire corridor architecture.

I also believe that the market generally underestimates the path dependency of non-USD stablecoins.
USDT and USDC did not grow because of logical consistency, but because they tapped into real liquidity and distribution flywheels.

Non-dollar stablecoins must also find their own catalyst to achieve a growth flywheel.

Without finding that catalyst, it will remain stuck in the phase of “theoretically sound but practically thin.”

So I completely agree that non-U.S. dollar stablecoins are important, but they are by no means a natural extension of U.S. dollar stablecoins.

It is a slower, more difficult business that tests operational capabilities.

If you put these three lines together, the coordinate system for stablecoins in 2026 is already very clear.

Agentic commerce is fundamentally a matter of authorization.

RWA and on-chain credit face challenges in verification.

On-chain FX is ultimately a matter of capital efficiency.

Stablecoins remain the common foundation, but the real competition going forward is no longer just about "who issues more."

Someone will win on intent and permission.

Someone will win on truth and underwriting.

Some will win on corridor liquidity, routing, and netting.

This is also why I’m increasingly feeling that the term “stablecoin company” itself is becoming imprecise.

In the past, the market valued issuance. But in the next phase, what truly matters is who controls the outer layer surrounding stablecoins: authorization, verification, credit, and liquidity.

In this sense, Circle’s recent valuation is certainly worth watching, but it shouldn’t be viewed solely as a stock story.

It’s more of a signal: capital markets are beginning to subtly recognize that stablecoins are no longer just one coin, but the monetary foundation for three distinct businesses.

Stablecoins are still assets.

But the real business lies in who decides how this money is authorized, verified, financed, and exchanged.