Maximum Reach vs Minimum Trust: LayerZero's DeFi Lead Thlither on Omnichain, Institutions and Agents

In the latest episode of the Bullish on Bitcoin Show, BOB co-founder Alexei Zamyatin hosted Thlither, DeFi lead at LayerZero, for an episode about omnichain - what it costs, who it's really for now, and where agents fit.

His route in was trading. Stocks and options in TradFi around 2016–17, then discovering crypto in 2017 and mistaking it for a parallel stock market - different charts, different tickers, no sense that there was a universe underneath them. He graduated into COVID, watched finance recruiting stop dead, and found himself with unlimited time exactly as DeFi summer started. Four years of full-time trading, investing and long-form DeFi research later, he met someone from LayerZero. He'd been a heavy Stargate user already. He's led their DeFi vertical for about two years.

Here's a rundown of the highlights.

Maximum reach vs minimum trust

Alexei set the frame from his own research background. His PhD was on crosschain bridging - not the implementations but the underlying question of what's cryptographically achievable - and the finding was not the one he wanted:

“We proved that fundamentally trustless crosschain communication is fundamentally impossible. My whole goal for my thesis was to build something fully trustless. And we realised, yeah, it's not possible.”

You always trust someone or something. The best available case is two chains that can fully verify each other, and you trust only that. Everything else is a trade-off between how far an asset can reach and how little you have to trust - and the practical version of that choice, as he described it from BOB's side, is whether you build something users will actually use or something maximalists will admire and nobody will touch.

Thlither's answer started with perspective. His burned-in memory of the old world is a token that didn't exist on Ethereum mainnet with about $750 million bridged across on a simple wrapped-token bridge controlled by one person's multisig. Whatever the arguments about today's designs, that's the baseline they're measured against.

Why reach isn't optional

The case for spreading an asset across ten or fifteen chains is, in his telling, unglamorous: you have to bring assets to where users, applications and use cases already are. The activity on Ethereum mainnet, Solana, Plasma and Arbitrum is different enough that no issuer can afford to bet on one environment.

That pressure increases with the asset classes arriving now - RWAs, tokenized equities, private credit. They need to exist where capital already sits.

“Whether we go in a direction where in five years there are five million chains or five chains, in either case it doesn't really matter. You'll always need to have reach across that set of chains.”

How you get security without giving up reach

LayerZero's approach is configurability: let issuers assemble their own security rather than inherit one fixed model. The verification layer is built from what they call decentralized verifier networks - DVNs - and larger issuers typically run four, five or six of them. USDT0 runs its own DVN with custom logic and its own security monitoring on top.

The example he gave for why this matters is PYUSD, where LayerZero works closely with PayPal and Paxos. It operates under New York's regulatory regime, which means specific boxes have to be tickable in how its crosschain security is set up - not a matter of taste.

The redundancy extends past the protocol. Verifiers run in different parts of the world, on different clients, with different cloud providers, so that a single provider outage doesn't take the whole path down.

“Crypto will always be an adversarial space where people are testing security models and trying to poke holes in them.”

Institutions and open DeFi won't merge

Alexei framed the live debate: onchain finance is being pulled toward institutions and TradFi, while a core of DeFi protocols still exists to build the parallel, decentralized alternative. Do those two camps make different crosschain decisions?

Thlither's view is that both will persist and neither will absorb the other. Hyperliquid is an open, permissionless exchange that found enough product-market fit that price discovery for commodities and equities now happens inside its order books. On the same day, the DTCC published an update on its onchain proofs of concept. He doesn't expect those two things to converge into one category - there's room for both, serving different people.

The most surprising thing he's heard from an institution

Asked for something from behind closed doors, he offered a piece of vocabulary. A non-crypto-native fund he spoke to referred to vaults as “onchain strategies”, and he loved it.

Inside DeFi, the conversation about vaults is about standards - is it 4626, is it 7540, is it single-chain or multi-chain. From the outside, a vault is simply a fund that happens to be open and permissionless: raise capital, deploy a strategy, return a yield to holders. That reframing, he argued, is emblematic of how banks and funds actually think about blockchains. Not a way to do conceptually new things - a way to do what they already do more efficiently, more widely and with fewer people excluded.

What he wants next is regulatory clarity, which he reads as the signal that it's safe to build these things in production.

Alexei agreed vaults are interesting and then argued the other side, which made for the sharpest exchange in the episode.

There are, he said, two kinds. The first are genuinely DeFi-native onchain strategies that deploy into Aave and Morpho - but liquidity there is thin, because onchain yield opportunities and incentives are smaller than they were. The second, where most of the money actually is, are basis trades, options strategies, tokenized money market funds and private credit funds wrapped in a vault.

What's fascinating to him is the direction of travel. Everyone assumes crypto has to go begging TradFi for distribution; in practice these funds are using onchain vaults as distribution, because a wallet deposit is easier than onboarding to a brokerage. His reservation is about where the capital ends up:

Deposit into Aave and you're adding to onchain liquidity. Deposit into many vaults and the capital is pulled offchain to be traded somewhere else. It gets better when those positions are tokenized back onchain - Ethena being the obvious example - because then the receipt becomes collateral again and composability returns. But he was candid that a lot of vaults are, at bottom, a smoother funnel for institutions to reach less sophisticated investors.

The spicy take: interop grew up

Asked for his most controversial view, Thlither said the nature of interop has changed completely in three years, and the evidence is in who moves the volume.

In 2022 and 2023 there were games to play: airdrops to farm, absurd yields to chase, a new Ethereum killer every quarter. Today the flow is exchanges rebalancing inventory, market makers, and asset issuers moving eight and nine figures between chains for the cost of gas. When Plasma launched, someone bridged $800 million of USDT0 into the chain in a single transaction - the kind of thing that would have been extraordinary three years ago and now isn't.

“My spicy take would be that it's very emblematic of the fact that the space is institutionalizing. It's less games for retail participants to play and more serious actors needing to move money from point A to point B in the most capital-efficient way.”

WBTC on 21 chains

On Bitcoin, Alexei raised WBTC's long-standing weakness: it was slow to go multi-chain, and newer wrappers took those chains while it waited. LayerZero's work with BitGo changed that, and WBTC is now on 21 chains.

The technically neat part is how. Because WBTC already existed on a set of chains before the integration, LayerZero's omnichain token standard let them add interop without deploying new token contracts - no liquidity migration, no rebalancing, no exiting one system to enter another. They plugged in and kept what they had.

Flows still centre on Ethereum mainnet, which he still considers the hub for size, moving in and out toward Monad, Base, Starknet, Hedera and other networks that want an asset tracking Bitcoin's price.

He was also frank that Bitcoin has more open questions right now than at any point he can remember - quantum breakthroughs and what they do to encryption, what happens to the coins in Satoshi's wallet, MicroStrategy as an external factor, and a broad market downturn. Through all of it, WBTC's TVL denominated in Bitcoin has kept growing, which he reads as the more meaningful signal.

Two kinds of Bitcoin holder - and what the data shows

His generalization: one group holds Bitcoin natively and never touches it, treating it as a hedge; the other believes in the DeFi mullet, where the point of holding an asset onchain is to use it. He sees the first group gradually joining the second, and almost nobody making the trip back.

His argument for why they stay is verifiability. Deposit WBTC into a lending market and you can open a block explorer and see it sitting in the contract. A centralized exchange cannot offer that, and for people who hold size and share the original crypto ethos, that guarantee is worth a lot.

Alexei then pulled BOB's own numbers, which complicate the picture in a useful way. Over the last six months the top pair for native Bitcoin swaps was BTC to USDT, followed by BTC to ETH, with BTC to WBTC third - and until about two months ago BTC–ETH was ahead, which he read as whales rotating between the two majors.

Underneath that are two distinct populations. One is cashing in and out: selling Bitcoin for stables, or buying Bitcoin and moving it straight to a hardware wallet. As he pointed out, if you're sophisticated enough to keep coins on a Ledger, there's no reason to hold WBTC you aren't using. The other is DeFi power users, and when you follow those transactions onchain they always end somewhere - Aave, Morpho, Yield Basis. They don't just sit.

Thlither's read was the same: the cash-in/cash-out crowd is probably a huge volume driver, and doing it without an exchange account or an OTC desk is a real service.

Watch the full episode

This recap covers the main threads, but the full conversation goes deeper on DVN configuration, the vault debate, and how cross-chain flows have shifted from retail to institutional. Watch the complete episode of the Bullish on Bitcoin Show with LayerZero here:

Watch on YouTube →