Asset Custody vs Payment Velocity: The $26B RWA and $650B Stablecoin Puzzle

General
1

I’ve been working on a DeFi protocol for the past year, and something’s been nagging at me about the RWA (real-world asset) tokenization narrative. The numbers tell two very different stories, and I’m trying to figure out what it means for where we’re all building.

The Tale of Two Blockchains

Ethereum’s Story: Tokenized RWAs hit $26.4 billion in March 2026—a fourfold jump from last year. Ethereum hosts about 60% of this value. BlackRock’s BUIDL fund, Franklin Templeton’s OnChain U.S. Government Money Fund, JPMorgan’s tokenized repos—they’re all on Ethereum. Even MetaMask integrated Ondo Finance to give users access to 200+ tokenized securities. This is the “institutional adoption” story we’ve been waiting for.

Solana’s Story: February 2026, Solana processed $650 billion in stablecoin transactions. That’s not total value locked—that’s actual transaction volume in a single month. Visa’s running pilot programs for cross-border settlement. PayPal, Stripe, Western Union, Fiserv—all live on Solana’s payments infrastructure. Sub-cent fees, sub-second finality.

The Question That Keeps Me Up at Night

Here’s what’s bothering me: Are we building a system where assets live on one chain but transact on another?

When I look at our protocol’s data, most tokenized treasury tokens just… sit there. Someone buys a tokenized 3-month T-bill on Ethereum, and it doesn’t move until maturity. The custody is on Ethereum, but the actual economic activity—the buying, selling, the real-time price discovery—that increasingly happens via stablecoins on faster, cheaper chains.

It’s like Ethereum became Fort Knox (we store the gold bars there) while Solana became the Visa network (we move the money there). Both are valuable, but which one actually captures more value?

The Math That Doesn’t Add Up

Ethereum: $26B in RWAs, but most sit idle after tokenization. Occasional trades, minimal transaction fees flowing to validators.

Solana: $650B in monthly stablecoin volume, but apps earn 3.5x what the network captures. According to recent data, Visa and PayPal keep most of the payment processing fees—SOL stakers get inflation and MEV, which is marginal compared to traditional payment processor margins.

So if you’re an institutional investor tokenizing assets, you choose Ethereum for the security and composability. But if you’re actually moving money around—paying for those assets, collecting dividends, rebalancing portfolios—you’re probably using stablecoins on Solana because gas fees matter at scale.

What Does “Winning” Even Mean?

I started thinking about this when we were evaluating which chain to deploy our next contract on. The conversation split our team:

  • Some argued: “Deploy where the assets are” (custody = king)
  • Others said: “Deploy where the transactions are” (velocity = king)
  • One person (our yield strategist) pointed out: “Neither matters if Visa and BlackRock extract all the value while we’re providing commodity infrastructure”

That last point hit hard. Are we building open, permissionless rails that ultimately just serve TradFi rent-seekers? Or is this what success looks like—institutions finally using crypto infrastructure, even if they capture most of the economic value?

I’m Genuinely Asking

I’m still learning about tokenomics and value capture, so I’m curious what this community thinks:

  1. Is asset custody or payment velocity the better value capture mechanism? Does it matter that Ethereum hosts $26B in RWAs if those assets barely transact on-chain?

  2. Should we care that apps and institutions extract most of the value? Solana’s apps earn 3.5:1 vs the network itself. Is that a feature (shows utility) or a bug (poor tokenomics)?

  3. Are we solving the wrong problem? Maybe asking “which chain wins” is the wrong question entirely. Maybe specialized chains for specialized use cases is exactly how this should work?

I’ve been building in this space for a couple years now, but I still feel like I’m missing something fundamental about how value flows in this system. Would love to hear perspectives from folks who’ve thought about this more deeply.

2

This is exactly the kind of conversation we need to be having. As someone who’s spent the last 6 years working on L2 scaling, I’ve watched this divergence play out in real-time, and I think it’s less of a bug and more of an inevitable outcome of blockchain architecture trade-offs.

Settlement vs Execution: Different Chains for Different Jobs

Here’s my take from the L2 trenches: Ethereum mainnet increasingly functions as a settlement layer, not a transaction layer. And that’s actually by design.

When we look at the numbers:

  • L2 TVL: $47B (up from $4B in 2023)
  • Daily L2 transactions: 1.9M (eclipsing mainnet’s 15 TPS)
  • L2s collectively targeting 100K+ TPS by end of 2026

But here’s the interesting part: RWAs stay on mainnet. BlackRock’s BUIDL, Franklin Templeton’s funds—they’re not moving to Arbitrum or Optimism. Why? Because custody demands the highest possible security guarantees, and despite all the innovation in optimistic and ZK rollups, mainnet Ethereum still has the most battle-tested security model and the deepest liquidity for high-value assets.

The Fort Knox Analogy is Actually Perfect

Your comparison of Ethereum as Fort Knox and Solana as Visa is spot-on. But consider this: Fort Knox doesn’t generate revenue from frequent transactions—it generates value from being the most trustworthy place to store valuable things. The security premium matters.

However (and this is the uncomfortable part), if most RWA transactions happen on L2s or Solana, mainnet captures minimal fees. The value accrues to:

  1. L2 sequencers (centralized, often private companies)
  2. Payment processors on Solana (Visa, PayPal—also centralized)
  3. Not necessarily to ETH or SOL token holders

Cross-L2 Fragmentation: The Real Problem

The bigger concern from my perspective isn’t Ethereum vs Solana—it’s L2 fragmentation. We now have 50+ L2s with separate:

  • Liquidity pools
  • Security assumptions
  • Governance models
  • Bridge risks

If a tokenized treasury lives on Ethereum mainnet but gets traded across Base, Arbitrum, Optimism, and Polygon, where does the economic activity actually happen? We’re solving scaling but creating interoperability hell.

Does Mainnet Become Just a Notary?

Your question about whether mainnet becomes irrelevant if L2s handle transactions—yeah, that keeps me up at night too. If Ethereum mainnet’s role is reduced to periodically checkpointing L2 state, does it matter that it’s “decentralized” when 90% of user activity happens on centralized sequencers?

But here’s the counter-argument: Being the notary is actually incredibly valuable. Legal systems have notaries precisely because someone needs to be the final source of truth. If Ethereum mainnet is the canonical record for $26B in institutional assets, that’s not nothing.

Pragmatic View: Specialization > Competition

Maybe asking “which chain wins” is the wrong frame. Different chains optimize for different things:

  • Ethereum mainnet: Maximum security, composability, institutional trust
  • Ethereum L2s: Transaction throughput while inheriting mainnet security
  • Solana: Payment rails, high velocity, low latency

This isn’t failure—it’s specialization. The internet didn’t fail because we have different protocols for different tasks (HTTP for web, SMTP for email, BGP for routing). Maybe blockchains specialize too.

The real question is: Can we build interoperability that’s as seamless as the internet’s protocol stack? Because right now, moving a tokenized asset from Ethereum to an L2 to Solana and back feels like three separate trips to three different countries, each with their own customs and currency exchange.