Bitcoin trades over $20 trillion in notional volume annually, dwarfing the rest of crypto combined. Yet because of its technical design, it hasn’t seen the proliferation of on-chain, decentralized exchanges that Ethereum and Solana have experienced.
Ethan Marcus, CEO of Flashnet, believes this represents one of the biggest opportunities in crypto today. His company is building infrastructure to enable decentralized bitcoin trading markets using Spark, a scaling solution from Lightspark.
Below is an edited transcript of Marcus’s conversation with Blockspace’s Charlie Spears on the Bitcoin Season 2 podcast, discussing everything from why Lightning isn’t optimal for trading to the crucial role traders and degens play by bringing liquidity to markets.
What is Flashnet?
Do you think of your company as a Bitcoin company or a crypto company?
Ethan: This is a Bitcoin company. There’s a fine line between what is Bitcoin and what is the rest of crypto. I’d like us to be just an infra play. I don’t want to build consumer-facing products. I’m very happy sitting in the background, building cool tech that frontends and products can use.
What exactly is Spark?
Ethan: Spark is a new way to think about Bitcoin scaling. It’s an iteration on state chains, which have been around for years. The Bitcoin Layers guys do a great job categorizing scaling groups, and Spark falls under native scaling alongside Lightning and Ark.
Dive deeper: Making Bitcoin cheap to spend with Ark
Spark is basically a way to delegate ownership of on-chain funds. It doesn’t require new consensus, new opcodes, or covenants that we’re going to end up waiting for. It’s purely delegation of on-chain funds with really great trust characteristics. It enables you to transfer sats and tokens on Bitcoin very fast and very cheap. It’s just a signing protocol.
What does the operator actually do when a user is using Spark?
Ethan: The operator’s job is to help coordinate transfers between users. To simplify – when you deposit into Spark, the operators work together with you to create a shared key. When you want to transfer to someone else, the operators do some fancy math to find the difference between your keys and tweak their key accordingly. The new key plus the recipient’s key can then spend the UTXO.
The operators’ role extends to deleting their last key so the previous owner can never spend those funds again. The funds need both the operators and the user to spend – neither can act alone.
How does Spark work for Bitcoin?
How is Spark different from vanilla state chains?
Ethan: Vanilla state chains have only one operator, so you’re fully trusting that operator to forget its key. Spark uses either a threshold set or a set of operators. We did a lot of work around Frost, the signing protocol used, and ended up with nested Frost where there’s always one required signer – the user.
If you don’t have the user signing, none of the operators can do anything malicious. The only malicious action possible is if someone sends you funds and a majority of operators collude with the sender to double-spend what they’ve already sent you. And there are mitigations in place for this.
Why didn’t vanilla state chains take off?
Ethan: The UX sucked. You couldn’t split UTXOs. If you deposited a UTXO into a state chain, you could only spend that exact UTXO. There were limited time locks, so you had to exit the system within a certain timeframe. Users don’t care about trust assumptions at the end of the day – they just want things fast and cheap that work.
With Spark, we enabled UTXO splitting, token usage, and relative time locks instead of static time bombs. All these things culminated in a much better user experience.
What does an infrastructure company look like beyond operating a Spark operator?
Ethan: Flashnet is a Bitcoin company. The biggest opportunity in crypto today is bringing Bitcoin trading on-chain. Bitcoin markets do upwards of $20-30 trillion a year in notional volume, trumping the rest of crypto. But it’s the only asset that hasn’t started trading on-chain.
If you look at Solana and Ethereum, there’s been a gradual rise of DEX volume compared to CEX volume. That hasn’t happened for Bitcoin. We’ve seen this oligopoly start standing up – in the US you can only trade on Coinbase, Gemini, Kraken. If you’re in Asia you probably trade on Binance.
If you’re serving an Ethereum or Solana market, just fork Uniswap and you’re there. That’s the gap we’re trying to fill. Bitcoin trading can be massively more efficient if done in a permissionless way. Lowering the floodgates for anyone to run a Bitcoin market is a huge opportunity.
Why didn’t building on Lightning work for this?
Ethan: Lightning is great but it’s built for a very specific use case. It needs to be open to everyone. I can’t expect Joe Schmo to want to trade and have to spin up a node, open channels, balance liquidity. Even though Lightning operations don’t necessarily involve you opening channels anymore, it’s very relevant when talking about trading at high frequency and high volume. These problems get super annoying to deal with.
We tried building it for six months and decided there has to be a better way. Almost serendipitously, Kevin [Hurley, CTO of LightSpark] had the idea for Spark and we were like, this is literally the solution we need.
Why trade bitcoin on-chain?
What are the benefits to users trading Bitcoin on-chain versus centralized exchanges?
Ethan: If you just want to buy Bitcoin with stablecoin or trade spot BTC/USD, it’s really a question of fees. Coinbase will scalp you 70 basis points for a trade. Trading on Binance is a little better, Kraken about the same. These exchanges don’t have any pushback because there is no optionality. It’s not like Bitcoiners can just pack up and go somewhere else – there is nowhere else.
If you lower the barrier to entry and any market can define its own fees, then it’s a question of where liquidity is parked. I can go on-chain and buy Bitcoin for 20 or 30 basis points depending on the market.
The second is Bitcoin native assets. The DeFi opportunity for Bitcoin is huge. There’s this latent value waiting to deploy. We’ve seen it with Ordinals, BRC-20s, Runes. We’ve even seen it with Unisat doing however many millions in volume since relaunching after getting hacked. People have an appetite for this, and we’re enabling exactly that with lower trust requirements. It’s about lowering the floodgates and allowing people to trade.
Do you see this evolving beyond just trading Bitcoin on-chain forever?
Ethan: What we’re building is a primitive to enable more financial infrastructure on Bitcoin. There’s a world where most stables stay on Bitcoin, and you can use that stablecoin as collateral to buy something else on top of Bitcoin, never needing to move out of the ecosystem.
First you need markets – you need to be able to swap in and out of any asset. Then we can start talking about what lending looks like on Bitcoin, what are the pain points of going to other networks, what does issuing debt look like. These are all things that are downstream.
What are your thoughts on Tether’s announcement about coming to Bitcoin?
Ethan: A press release isn’t real until it’s real. Unfortunately, we haven’t seen anything come to fruition yet. The harsh truth is I don’t think stablecoins on Lightning would ever work. Multi-asset Lightning is a disaster waiting to happen.
Lightning for just payments with sats took years to get to 95% payment efficiency. How long will it take with stablecoins? We’ll have to convince node operators to open stable channels, get liquidity in place. In the beginning it’s just going to be super centralized hub and spoke.
The other issue – and a lot of people will hate when I say this – is privacy. A stablecoin by nature has to be regulated and compliant. If you don’t know what’s happening with the stablecoin, you can’t be compliant. If you can’t freeze an asset after a terrorist holds it, you can’t be compliant. Lightning cannot offer this. For that reason, I don’t see Tether going to Lightning or RGB anytime soon.
Flashnet launch and the future of Bitcoin
What happened when Flashnet launched and got flooded?
Ethan: Huge validation on one hand. On the other, we miscalculated the appetite like crazy. We load tested for a couple hundred concurrent swaps because we stuck the beta flag everywhere thinking people would be nice. We underestimated the degens.
We attempted to serve thousands of concurrent swaps – $300,000 in nine minutes was trying to flow through the system, which decided this was too much and started having hiccups. We shut it down, took a step back, did two audits – mainly because Odin got hacked around the same time so we thought we might as well not have the same fate.
Our load testing philosophy is very different now. We’re loading until it breaks, then pushing it further every time. That will yield much smoother results.
How do you navigate building for serious capital markets when your first users are degenerates?
Ethan: It’s the same playbook every successful DEX ran. Every network has run it. Solana is now peaking in capital markets – it has probably some of the most sophisticated high-frequency trading and MEV of any network, and it all started out and continues to be powered by degens.
I’m happy to go talk to all these LPs and big market makers to get them to allocate liquidity. Or we can just ship something that anyone can build on and let them run wild.
What do you think about all these new chains being announced?
Ethan: I think we’re speaking on the perfect day because Stripe and Paradigm just announced Tempo, their new blockchain. I think it’s a mess.
There’s a pattern with crypto where if you stay long enough, you end up just back at Bitcoin. You’re an engineer working on Ethereum scaling – I’ll see you in Bitcoin in five years. There is this return to the beginning. I think it’s just because there is this uncompromising level of rigor and standard that Bitcoin Core does have that enabled this great thing that is Bitcoin. That is not the case on other networks.
Sadly I think it’s going to cause immense amounts of liquidity fragmentation. I think it compromises the entire reason that we’re in crypto in the first place. I think it’s fully just regulatory arbitrage. People are saying, hey, I don’t need licenses to do this anymore if I just run a sequencer and post this to this chain every five minutes or whatever it may be.
I think that’s going to end pretty quickly. Quickly is actually a longer timeline than a couple months – it’ll probably be like 5 to 10 years. But I don’t see a world where there’s a million chains and we have to use Layer Zero to go from this chain to that chain. At what point does the regulatory arbitrage get rubbed out and you just call the sequencer what it is – a server? I already send payments with Stripe, why do I need a blockchain?
What’s your opinion on soft fork proposals for Bitcoin?
Ethan: My unpopular opinion is if it ain’t broke, don’t fix it. Spark is a perfect example – we’ve been able to use the base layer rules of Bitcoin to create something great. Would Spark be better with CTV or OP_CAT? 100%. Would that make our lives easier? Yes. Is that something I’m going to hold my breath about? Probably not.
The Bitcoin community can sit around advocating for these things, but the horizon for these changes is years and years. Bitcoin is a snail when it comes to adopting new changes, which is a feature and a bug. I’m all in favor of them, I just wouldn’t hold my breath.
What else in Bitcoin do you find intellectually interesting?
Ethan: I’m honestly so locked in that I don’t see anything on my periphery. What we’re building is probably some of the most fascinating problems I’ve ever worked on, and probably some of the most fascinating problems to work on in Bitcoin in general. My brain is too exhausted to look outwards for now.
