Solving the $30 trillion Liquidity Silo Problem w/ Ethan Marcus

Ethan Marcus founded FlashNet to address what he calls the biggest problem in crypto: Bitcoin markets trade upwards of 30 trillion dollars a year but operate almost entirely through trusted custodial intermediaries, with liquidity fragmented into geographic silos across Coinbase in the US, Kraken in Germany, Binance in Asia and Africa, and various exchanges in Europe. FlashNet originally set out to build a high-frequency trading network for Bitcoin on Lightning but concluded that Lightning does not scale for non-custodial users in practice, that the channel-based model cannot support multi-asset use cases, and that minimizing trust introduced too many steps and too much latency.

The pivot came when Kevin Hurley sent a white paper for Spark, a signing protocol where ownership of UTXOs on L1 is signed off to successive recipients using state chains. Spark has zero fees to transact today, provides pure unilateral exit to L1, and uses a threshold operator network where as long as one operator is honest during a transfer the user has perfect forward security. A user can verify Spark funds have not been double spent by confirming the UTXO has not been spent on-chain, making the trust assumption independently verifiable. Spark uses an account model rather than a channel-based model, eliminating the need to pre-fund or open channels. Lightning on Spark is implemented as an atomic swap between Spark Bitcoin and Lightning Bitcoin through a service provider, and Wallet of Satoshi had previously been forced to stop operating in the US due to non-custodial Lightning limitations until Spark became available.

FlashNet sits on top of Spark as a quasi execution layer with validators holding shards of a key that controls pools functioning as AMMs. Users swapping against FlashNet pools execute atomic swaps and validators reach consensus before signing off on fund movement. LPs in FlashNet pools incur longer standing risk because validators could theoretically collude. FlashNet's architecture requires LPs to hold only BTC and USD inventory, unlike THORChain which requires holding both assets of each swap pair. FlashNet is currently the primary LP in its own network to bootstrap execution quality, and Marcus says the LP dynamic is expected to invert once external LPs have enough alpha to gain from trading against FlashNet's distribution.

FlashNet's core focus today is B2C USD swaps. It works with CoinFlow for merchant payments, with neobanks Bellow and Ugly Cash for remittances in Latin America, and is integrated into Liquid Wallet for US onramps. FlashNet recently began powering BTC swaps for Trust Wallet, with Marcus describing the resulting flow into Bitcoin as very large. The biggest route on FlashNet is USDT on Tron, which Marcus says he would not have expected to be dominant, and the vast majority of FlashNet volume is BTC to stablecoin. Consumer swap fees are approximately 10 to 15 basis points depending on the originating network, and FlashNet currently sees daily demand for five to ten million dollar swaps it lacks the capacity to fill.

FlashNet has a stablecoin called USDB, issued by a regulated US company called Braille and redeemable one-to-one with USDC on any network, which FlashNet uses as an intermediate routing step between external chains and BTC on L1, Lightning, or Spark. Marcus views the market in three tranches: retail non-custodial users first, then active traders doing hedging and arbitrage, then large centralized products like River, Cash App, and Coinbase. He acknowledges the retail non-custodial market is not a multi-billion dollar per day market today, and the longer-term thesis is that centralized exchanges would tap FlashNet liquidity because on-chain execution fees and speed would be better than what they can do in house.

Marcus predicts Bitcoin will scale for payments invisibly, with stablecoins on the front end because users do not want a floating balance, and that financial businesses have a strong incentive to own their own stablecoin ledger rather than renting Visa, Mastercard, or Fedwire rails. The host, who discloses being an investor in FlashNet, raised the counterpoint that USDC on Solana is already fast, cheap, and composable, making it hard to reconcile why the wheel needs to be reinvented for Bitcoin payments. Marcus calls Bitcoin lending the golden goose of Bitcoin, with core demand today coming from whales holding 50 to 100 million dollars in Bitcoin who need cash liquidity without selling, though he concedes it is a poor fit for someone holding only one Bitcoin and that daily price swings of 10 to 20 percent make collateral positioning difficult for new users.

Marcus argues that upcoming IPOs will mint many millionaires and billionaires who will likely put money into Bitcoin rather than traditional vehicles, and points to Cash App with approximately 60 million monthly active users as committed to building out Bitcoin payments with a meaningful and non-zero portion of revenue already Bitcoin-denominated. He adds that airdrop-centric strategies have underperformed this cycle and are likely to underperform further, and that CT noise does not equal success.