In the freewheeling crypto economy, traders are often highly levered — after all, what drew most retail investors to Bitcoin if not jaw-dropping stories of overnight riches? 

The flip-side to a levered bet is, of course, the threat of liquidation. Last October, when crypto prices were hovering near record highs, a dip in global markets set off a liquidation cascade that hit $19 billion in less than 24 hours. 

The S&P 500 has since hit a new all-time high. Major cryptocurrencies, meanwhile, have yet to recover.

You might think a protocol that promises to protect levered traders from liquidation is selling snake oil. Not necessarily. In this week’s newsletter, we’ll dig into the first version of Hylo, a Solana protocol that has made — and kept — that very promise. But it came at a cost. Read on to learn more. 

The premise is straightforward. A pool of staked Solana backs both tokens. But hyUSD holders come first: their tokens are always worth $1. The xSOL holders, in turn, absorb all the volatility of the underlying asset. In other words, the leverage doesn’t come from borrowing, and xSOL holders don't owe anyone anything; they just hold a claim that can shrink toward zero in a crash or grow substantially during a rally.

Because xSOL represents a levered bet, it resembles a perpetual futures contract. Unlike “perps,” however, the first version of xSOL came with two big promises: first, that it presented no liquidation risk, and second, that it did not require a continuous funding fee. In Hylo’s words — or the words of its developers — xSOL delivered “3x SOL without hidden costs.”

Strictly speaking, those commitments were real. Nobody holding xSOL was ever forced out of a position. Thanks to the revenue generated by the staked Solana in the pool, nobody paid an ongoing fee to hold leveraged SOL exposure the way they would on a perpetual exchange.

There’s no such thing as a free lunch, of course. Trade-offs included dynamic fees to mint and redeem hyUSD, volatility decay, and variable leverage, to name a few. 

Hylo was immediately put to the test, launching in the fall of 2025. Since the October liquidation crisis, Solana has lost about two-thirds of its value. The drop has been so steep that it should have broken Hylo, wiping out xSOL holders and breaking hyUSD’s peg to the US dollar. 

Incredibly, hyUSD held its peg. So if nobody was liquidated despite the leverage, where did the extra loss go, and who ended up paying for it? 

The short answer is dilution.

Between October 2025 and June 2026, the supply of xSOL grew from a modest 8.4 million to 235 million, a 28-fold increase. 

This dilution took two forms. 

In one, users deposited fresh Solana into the pool and received xSOL at par. In the other, Hylo’s “stability pool” automatically burned a reserve of hyUSD tokens and issued new xSOL in its place. 

Both gave hyUSD some breathing room, but they came at a cost to people who already held xSOL. The dramatic increase in token supply, particularly via the stability pool, meant there were far more claimants for a pot of crypto that hadn’t grown nearly as fast in nominal terms. 

That means Solana would have to hit at least $630, and as much as $719, in order for xSOL to hit $1 again, a price it hasn’t seen since November. That is, of course, a tall order — Solana’s all-time high was just $293. 

But Halo’s developers didn’t rest. A newer version of the protocol features multi-asset collateral, the ability to mint and redeem hyUSD against a new USDC pool, and a redesigned stability pool. 

To learn more about what this means for xSOL holders and Hylo’s promise of liquidation-free leverage, read our full report here.

Tranching splits a pool into layers of risk. Junior depositors absorb losses first and get paid more for it. Senior depositors sit above them, protected until the junior layer runs out. Simple, old-school playbook, now onchain.

The math behind that protection looks solid within offerings: buffers can range from 9% to 39%, sized to absorb a normal bad day.

But DeFi exploits rarely result in a normal bad day. A median hack erases 34% of TVL, and a quarter of them erase 80% or more.

Running six major tranching protocols against 151 real exploits, and five of them get breached by the median case alone.

Find out which ones hold up.

Introducing V2 of the DefiLlama RWA dashboard, the deepest resource on tokenized assets and how they’re used in DeFi.

RWA perps, inflows & outflows, DeFi Active TVL, extensive historical data, and more.

Check it out here.