Leverage is crypto trading’s faultline

In all the Monday morning quarterbacking of last month’s crypto flash crash, one word turns up consistently: leverage. Leading up to the crash, open interest in crypto derivatives had reached an all-time high of roughly $115bn. Within a day, almost $20B in trades were liquidated — some not because the bets were wrong, but because they were mired in inadequate market plumbing. While this may not have been a credit event per se, leverage — seemingly everywhere and nowhere at the same time — plays an integral part in the crypto story. Collateralised and unsecured lending in crypto has been around almost as long as the asset itself. After all, what else were you (legally) going to do with it in the early years? It is cited as the weak link that sent the crypto market into a tailspin in mid-2022. Since then, other forms of leverage have taken on greater significance, including margin trading on centralised and decentralised exchanges, options and futures open interest, staking as collateral, perpetual futures (ie, futures with no or very long settlement dates), synthetic leverage, and rehypothecated tokens. Leverage in crypto can be built into smart (ie, programmable) contracts and can arise without explicit borrowing when specific parameters are met. While recent headlines have raised alarms about the borrowing undertaken by crypto treasury companies, at least you can see it on a balance sheet. But much of the leverage in crypto actually sits inside the code and the smart contracts. What might this look like in practice? Suppose a trader drops one BTC into a centralised exchange and ticks the box for cross-margining, meaning that single deposit backs everything they do. They put on a long perp at roughly 4x, throw on a small ETH/BTC short as a hedge, and — because people do this — lend part of their BTC back to the venue for a bit of yield. Depending on how you calculate it, their effective leverage now sits somewhere around 5.5x.  Meanwhile, the exchange doesn’t let the lent BTC sit idle; it rehypothecates it into liquidity pools or hands it off to market makers. Then a macro headline drops late Friday, right after traditional markets close. BTC slides 13 per cent, ETH drops 17 per cent. The trader’s margin (which is the BTC deposit) shrinks across every leg of her book. The perp goes red, the ETH/BTC short helps but not enough, and they blow through the minimum margin. The risk engine kicks in and liquidates the account. That BTC they lent to the exchange that was rehypothecated is no longer fully there — and what remains of it is in other exposures that are also being liquidated. Now imagine this is happening not to one person, but to thousands across multiple exchanges. Several aspects combine to make crypto leverage especially fragile, including a 24/7 market that lacks real-time infrastructure, a theoretically transparent underlying technology that lacks transparency, and volatility collateralising volatility. While black box risk and margining engines and protocols are continually updated, many accounting and reporting systems still operate in batch mode. All this occurs without co-ordination or standardisation of calculations and metrics across market participants. Humans — who occasionally need to sleep — are still essential, especially during times of market stress. You’re left with real-time price action and not nearly enough real-time risk management; a market that trades like FX with the infrastructure of an asset class that stops trading at the end of the day. While the on-chain ledger lets us see wallet balances and posted collateral, that visibility drops off quickly once assets are pledged, lent out, wrapped, or bridged to another chain — basically any time they leave the clean on-chain path. Transparency mostly ends at that “first hop”; it tells you very little about anyone’s actual leverage. And even if you can look at order books or pool depth, you still don’t know how many of those assets have been re-used, how intertwined the liquidity providers are, or how much of the liquidity is coming from transient market makers. Proof-of-reserves only shows the asset side of the picture, not the obligations sitting against it. In crypto, leverage often ends up being strangely self-referential. The same volatile tokens show up again and again as collateral — usually with haircuts that would raise eyebrows in traditional markets. When the underlying price drops, so does the collateral, and the whole thing picks up speed into liquidation. Instead of dampening risk, the collateral becomes the fuse that lights a very fast liquidity spiral. It’s not that leverage in crypto is inherently bad. But it needs to be manageable. It would benefit from many of the aspects of market plumbing that make leverage manageable in other markets. In fact, two paths are emerging, one resembling a more traditional market structure for participants who want those guardrails, and another, focused on leveraging crypto’s programmability. These need not be mutually exclusive. There are continued signs of improvement. Some are experimenting with proof-of-liabilities to complement proof-of-reserves. Startups are developing risk dashboards that provide participants with improved visibility into their collateral. Larger players are adopting more traditional, risk-weighted frameworks, though the 24/7 nature of that remains a challenge. Policymakers and regulators are also busy translating high-level objectives into actionable steps for market participants. For example, the CFTC recently requested comments on oversight of perpetual derivatives. These developments represent steps in the right direction — one in which leverage in crypto can be better managed through improved identification and visibility, better co-ordination, some consistency in standards and measurements, and our ability to see all this as it unfolds and not just in the rear-view mirror.

BY: Elizabeth Menke and Wendi Carver