Bitcoin is quietly becoming a settlement network

Beyond being a store of value, Bitcoin is becoming a neutral settlement network. Its base layer is deliberately slow — roughly seven transactions per second, with ten-minute blocks — but a stack of Layer-2 protocols moves value in seconds and settles back to Bitcoin for final security. The defining property is that they clear payments without trusting any single company: settlement finality comes from Bitcoin’s proof-of-work, not a corporate promise. Stablecoins can sit on top as the familiar-currency interface, but Lightning is the rail underneath — and the businesses packaging those settlement properties into remittances, cross-border payments, and merchant tools are where the value is captured.

This is the layer we build on, and the trust model matters more than the throughput numbers.

Why is Bitcoin’s base layer so slow?

Bitcoin’s base layer (Layer 1) processes around seven transactions per second, with a new block roughly every ten minutes. That’s slow by design. The base layer optimises for security and global accessibility — anyone, anywhere, on cheap hardware, can verify every transaction — and throughput is what it deliberately gives up to keep those properties.

Layer-2 protocols solve the speed problem without touching the base layer’s rules. They process transactions off-chain, quickly and cheaply, then periodically settle back to Layer 1, inheriting Bitcoin’s security for final settlement. The base layer stays simple and hard to attack; the fast, high-volume activity happens a layer up. The monetary properties don’t change — a Layer 2 can’t inflate the supply or rewrite Bitcoin’s rules.

How Lightning moves money instantly

The Lightning Network is the most mature of these. Users lock Bitcoin into two-party payment channels and transact off-chain, settling only the opening and closing transactions on the base layer. A payment from London to Singapore completes in seconds rather than the ten to sixty minutes an on-chain confirmation would take.

This isn’t theoretical infrastructure any more. Lightspark — founded in 2022 with $175M from a16z and Paradigm — has driven serious enterprise adoption: Revolut integrated Lightning in 2025, SoFi launched remittances on it, Xapo became the “first bank on Lightning,” and Tether is integrating Lightspark’s Spark protocol into its wallet kit.¹ Most major exchanges now support Lightning for deposits and withdrawals — Bitfinex, OKX, Kraken, Binance, and Coinbase among them.² By late 2025 the network’s public capacity stood at roughly 4,000 BTC across thousands of nodes and tens of thousands of channels,³ and in well-configured deployments it’s reliable: River Research reported a 99.7% success rate across 308,000 payments.⁴

The settlement thesis

This is the part that shapes what we build: Lightning can move value faster than traditional correspondent banking — final settlement in seconds rather than days — without requiring trust in any single corporate intermediary.

Stablecoins may well play a role in the experience. Customers often want to denominate transactions in familiar currencies, and on- and off-ramps to local banking need fiat-denominated touchpoints. But the stablecoin is the interface; Bitcoin and Lightning are the infrastructure that actually clears the transaction.

That distinction is the whole argument. A payment processor built on XRP depends on Ripple’s continued operation, its token economics, and its regulatory standing — decisions made in a San Francisco headquarters. Lightning has no such dependency: the protocol is open, the network is permissionless, and settlement finality comes from Bitcoin’s proof-of-work rather than any company’s promise. The opportunity for Bitcoin-native businesses is to build the products that make those settlement properties usable — remittances, cross-border B2B payments, merchant services, treasury operations. The rails already exist. The interface layer is where the value is captured.

What about people who can’t self-custody?

Bitcoin’s self-custody model demands technical competence most people don’t have. Chaumian eCash, via a protocol called Fedimint, offers a different trade. Trusted community members collectively hold Bitcoin while users transact with simple tokens that move instantly and work with Lightning. The trade-off is explicit: you trust a small federation rather than managing keys yourself. For a neighbourhood, a workplace, or a local merchant network — places where trust already exists — that often makes more sense than expecting everyone to become their own bank.

It already works in practice. Bitcoin Ekasi in South Africa runs a Fedimint federation that pays staff salaries in Bitcoin and lets local shops accept payments without on-chain complexity; in Kibera, Kenya, AfriBit Africa has onboarded over 2,600 residents on the same model. The underlying asset stays Bitcoin — a federation can’t inflate the supply or mint unbacked tokens. The packaging changes; the monetary properties don’t.

Spark and Ark: better UX, still your own exit

The thread connecting all of these is one property: unilateral exit — the ability to withdraw back to the Bitcoin mainnet without anyone’s permission. Axiom Capital, an investor in the space, calls it the defining characteristic of a genuine Layer 2; without it, they argue, a “Layer 2” is really just a multisig with no credible path to decentralisation.⁵

Spark, built by Lightspark, is a statechain-based Layer 2. It lets users transfer Bitcoin ownership without recording each transfer on the main chain, using a two-of-two signing arrangement between the user and a Spark operator. When ownership changes, the new owner receives signing authority plus a pre-signed, timelocked transaction that lets them exit to mainnet unilaterally if the operator stops cooperating. The trade is that the operator must cooperate for normal transfers — but the exit path means users aren’t handing it custody. We’re bullish on products built on Spark: it fits use cases where speed matters more than absolute trustlessness, and the mainnet exit preserves optionality that purely custodial wallets lack.

Ark takes a different route to the same goal. Launched in public beta on mainnet in late 2025,⁶ it uses Virtual Transaction Outputs (VTXOs) coordinated by Ark Service Providers, with no channels to open or close. That sidesteps Lightning’s biggest friction — channel liquidity management, where you need inbound capacity to receive and large payments can fail if intermediate nodes are short on balance. The trade-off is that Ark requires periodic interaction with a provider (currently about every four weeks) to keep VTXOs from expiring, whereas Lightning channels can sit dormant indefinitely. It’s earlier-stage than Lightning, but on a five-to-ten-year horizon the architecture solves real problems, and we’re bullish there too.

Why this is a durable business

A Bitcoin company building on these rails is defensible as a business, not just sound as technology. As software and production costs fall — and AI-assisted development is accelerating that — durable advantage shifts away from code, which can be replicated, and toward things that can’t: scarce assets, distribution, trust, and regulatory positioning.

Bitcoin-native companies build on the scarcest digital asset; that’s one moat. The other is regulatory friction — money-transmission licenses, banking partnerships, and compliance infrastructure take years and capital that any competitor must also spend. In a world where spinning up a product becomes trivially cheap, the barriers that remain are the non-technical ones. Bitcoin provides the asset-layer moat; the regulatory and institutional relationships provide the distribution moat. And every one of these protocols expands what Bitcoin can do without changing its base-layer monetary rules — greater capability, same hard money underneath.

Common questions

Can you actually pay with Bitcoin, or is it just for holding? Both. The base layer is built for security, not speed, but Layer-2 protocols like Lightning settle payments in seconds for a fraction of a cent — and they’re already integrated into banks, exchanges, and payment apps.

What is the Lightning Network, in simple terms? A network of payment channels built on top of Bitcoin. You move money off-chain instantly and cheaply, and only settle to the Bitcoin blockchain when you open or close a channel — so you get speed without giving up Bitcoin’s security.

Is Lightning safer than using a stablecoin like USDC? They do different jobs. A stablecoin is issued and controlled by a company; Lightning is a neutral protocol whose finality comes from Bitcoin’s proof-of-work. In practice the two combine well — the stablecoin as the currency interface, Lightning as the settlement rail.

What’s the difference between Lightning, Spark, and Ark? All three are Bitcoin Layer 2s that let you exit back to mainnet unilaterally. Lightning uses payment channels; Spark uses statechains (simpler transfers, an operator you can exit past); Ark uses virtual outputs that avoid channel-liquidity management. They solve overlapping problems with different trade-offs.

Lightspark — founded May 2022 by David Marcus (ex-PayPal/Meta), first round co-led by a16z and Paradigm; integrations include Revolut (May 2025), SoFi (Aug 2025), and Tether’s Spark/Wallet Development Kit integration (Aug 2025). Lightspark newsroom, https://www.lightspark.com/about; a16z, https://a16z.com/announcement/investing-in-lightspark/ ↩
Exchange Lightning support — Bitfinex (2019, first major exchange), OKX (2021), Kraken (2022), Binance (July 2023), and Coinbase (2024, via Lightspark). The Block, https://www.theblock.co/post/291662/coinbase-lightning-network-lightspark ↩
Lightning Network public capacity and node/channel figures (approximate, late 2025) — mempool.space, https://mempool.space/lightning ↩
River Research — 99.7% payment success across 308,000 transactions (Aug 2023). https://river.com/learn/files/river-lightning-report-2023.pdf ↩
Axiom Capital, “The Bitcoin Stack” and Ark Labs investment thesis — “Without unilateral exit, these are not real Layer 2s.” https://www.axiombtc.capital/stack ↩
Ark Protocol — mainnet beta launched October 2025; $2.5M pre-seed (Aug 2024) led by Draper Associates with Fulgur Ventures and Axiom Capital. https://ark-protocol.org/ ↩