Why settlement infrastructure should be chain-agnostic

Every stablecoin infrastructure company gets asked the same question in its first ten meetings: which chain are you on? It's a reasonable question with an unreasonable premise: that a settlement layer should have a chain the way a company has a headquarters. We think the premise is wrong, and that acting on it is one of the most common strategic errors in the category. This is the case for chain-agnosticism as a design principle: not indifference to chains, but the refusal to make the customer's settlement depend on any one of them.

The premise, and why it's wrong

The premise assumes the settlement layer chooses the chain. It doesn't. Three other parties do, and none of them coordinate:

Issuers choose. USDC is native on more than twenty networks; USDT on a dozen or more; Zand AED launched multi-chain on public blockchains; First Abu Dhabi Bank token (DDSC) is live on the ADI blockchain; JPYC launched on Ethereum, Avalanche and Polygon. The regulated non-USD tokens that matter most to us are launching on the chains their issuers and regulators prefer - and those preferences are diverging, not converging.

Counterparties choose. The exchange, market maker, or payout partner on the other side of a settlement has its own chain support, its own wallet infrastructure, its own preferences. A settlement layer serves them where they are or doesn't serve them at all.

Customers choose. A fintech that already holds balances on one network, or whose users do, doesn't want to be told the settlement provider only speaks another. Infrastructure that requires the customer to migrate isn't infrastructure; it's a platform with a lock-in.

Once you accept that the chain is chosen by others, the settlement layer's job is obvious: be wherever the value is, and route between wherever it needs to go. That is what "chain-agnostic" means.

The risks a single-chain bet inherits

Bet the infrastructure on one chain and you inherit that chain's roadmap, economics, and failure modes — and you pass all of them to your customers.

• Roadmap risk. Fee models, throughput, upgrade cadence, and governance are someone else's decisions. A settlement layer's uptime shouldn't depend on a foundation's release calendar.

• Liquidity risk. Non-USD liquidity is thin and unevenly distributed across chains; the deepest AED pair or euro pair might live on a network you didn't pick.

• Regulatory risk. Regulated tokens are being launched on chains chosen for compliance properties (permissioning, identity primitives, issuer control) - often not the chains with the deepest DeFi liquidity. A single-chain layer may find itself on the wrong side of an issuer's regulatory choice.

• Counterparty concentration. If every settlement runs through one network's bridges, sequencers, or RPC infrastructure, an outage there is an outage everywhere. The bridge-exploit history is a reminder that concentration is where losses cluster.

• Strategic capture. Chains court infrastructure companies with grants and ecosystem support - welcome, and worth taking - but building the architecture around a sponsor's chain converts a grant into a dependency. Take the support; keep the routing layer neutral.

What chain-agnostic actually means (and doesn't)

It doesn't mean supporting every chain on day one, and it doesn't mean having no opinions. It means:

The abstraction sits above the chain. The customer's API speaks in currencies, amounts, counterparties, and endpoints - not chain IDs. Chain selection is a routing decision inside the layer, driven by where the token is native, where liquidity is deep, and what the counterparty supports.

Multi-chain custody as a first-class capability. Key management, approval policies, whitelists, and monitoring that work identically across networks - the same custody discipline everywhere, not a per-chain patchwork.

A cross-chain routing policy, not a bridge. Burn-and-mint where the issuer supports it, liquidity networks within depth, vetted bridges only as a capped last resort, per-chain finality thresholds - evaluated per token, per chain pair, per ticket.

Deliberate expansion. New chains are added on evidence - an issuer launching there, a counterparty requiring it, liquidity worth reaching - with the same security review each time. Agnostic about which; disciplined about when.

Opinions, held loosely. We have views on chains - throughput, finality guarantees, tooling maturity, compliance primitives. They inform routing weights and expansion order. They don't get to become the customer's constraint.

What it costs

Chain-agnosticism is more expensive to build than picking a favorite. Every supported network means integration, custody support, monitoring, finality policy, and ongoing maintenance; cross-chain routing is an engineering discipline in its own right; and testing surface grows with the matrix. That's the real reason most infrastructure companies pick a chain: it's cheaper and it makes for a cleaner slide.

We think the cost is the point. The N×N problem is what makes settlement infrastructure a category at all - a layer that absorbs the matrix so customers don't have to. A settlement provider that solves the matrix by choosing one cell of it hasn't solved anything; it has asked the customer to hope everyone else chose the same cell. They won't. Paying the cost of agnosticism is how the layer earns the right to sit underneath everyone.

The dirham case makes it concrete

Look at the UAE's own token map: licensed dirham tokens are launching on different networks by different issuers - public multi-chain from one, a purpose-built chain announced by another - while the Digital Dirham runs on the central bank's own platform and mBridge. A settlement layer for AED that had bet on one chain would already be on the wrong side of at least one of them. The market didn't converge; it multiplied. Infrastructure that assumes convergence breaks on contact with exactly the tokens it most needs to serve.

That's why we're chain-agnostic - by conviction, and by arithmetic.

Frequently asked questions

What does chain-agnostic mean for stablecoin settlement?
That the settlement layer abstracts the blockchain away from the customer - speaking in currencies, amounts and endpoints - and routes across whichever networks tokens, liquidity, and counterparties actually live on, rather than requiring everyone to use one chain.

Why not just pick the best blockchain?
Because issuers, counterparties, and customers choose chains independently and increasingly divergently - regulated tokens launch on the networks their issuers prefer. A single-chain layer inherits that chain's roadmap, liquidity, and outage risk and passes it to customers.

Is chain-agnostic infrastructure more expensive to build?
Yes - more integrations, multi-chain custody, cross-chain routing, and larger testing surface. That cost is the substance of what a settlement layer provides.

Which chains does a chain-agnostic layer support?
Those where supported tokens are native, liquidity is deep, or counterparties require presence - added deliberately, with the same security review each time, rather than all at once.

Does chain-agnostic mean using bridges?
Not primarily. It means a routing policy: issuer burn-and-mint where available, liquidity networks within depth limits, vetted bridges only as a capped last resort, with per-chain finality thresholds.

Sources

• Zand AED launch — multi-chain on public blockchains (Nov 2025) · IHC/ADQ/FAB stablecoin on the ADI blockchain (Apr 2025)

• JPYC launch on Ethereum, Avalanche, Polygon (Oct 2025)

• CBUAE, Digital Dirham primer — issuance platform and mBridge