Author: Don Wilson, CEO and Founder of DRW
Friday’s crypto selloff raised plenty of alarms and finger pointing, from blaming exchanges to faulting liquidity providers. But that misses the real question: If real-time, 24/7 infrastructure is the future of finance, are we building the best thing possible?
On days like last Friday, the market tends to dole out expensive lessons to some market participants. I always say it’s important to get your money’s worth from that tuition. The objective should be to learn the lesson so we don’t have to repeat the class.
Here are the three things I have been thinking about since Friday (plus one POV learned from decades in global markets):
Friday’s selloff offered a live stress test for one of crypto’s core innovations: real-time margining. The concept – adjusting collateral continuously as markets move – promises better risk control in theory. But as we saw, it also exposes new fault lines when infrastructure isn’t built for continuous, global stress.
Traditional markets have always functioned on batch margining. If you lose money on your derivatives position, your FCM will kindly ask you to wire some more money. They might even give you a couple of days to send it. The FCM acts as a buffer for the clearinghouse, which in turn has a default fund. Of course, that can go horribly wrong, as we’ve seen more than a few times.
Real-time margining, with auto-liquidation when accounts fall below a threshold, holds out the promise of reducing systemic risk. However, it also changes the rhythm of markets, especially in markets that operate 24-7. If the farmer has hedged his crops on a futures exchange and the futures markets have a massive dislocation in the middle of the night, that farmer may get stopped out with real-time margining.
This highlights the benefit of having an FCM stand between the customer and the exchange. The FCM can choose to provide additional credit, allowing the farmer to sleep at night while still capturing the benefits of real-time risk management. That layer of credit and judgment smooths the system’s response to volatility.
Most crypto platforms don’t have this type of FCM-like buffer in the mix, which makes this approach far more challenging. Positions are marked and liquidated instantly and when liquidity dries up, there’s no intermediary capital to cushion the shock, as we saw last week.
The other problem on Friday was that some platforms allegedly suspended deposits — a breakdown in market plumbing that would be unthinkable in traditional finance. Traders had no way to meet margin calls, forcing further liquidations and triggering additional volatility. That’s the kind of operational fragility that must be fixed for tradfi to function on these new rails.
When those buffers don’t exist, the system turns to auto-deleveraging, or ADL. ADL is a mechanism that kicks in on many crypto exchanges when there isn’t enough liquidity in the order book to absorb auto-liquidation orders. The exchange forcibly offsets those positions against open interest on the other side of the market, choosing which traders’ positions to reduce or close.
In traditional finance, this is reserved for extreme cases following large FCM defaults. It didn’t even happen during the Lehman Brothers collapse, one of the most severe stress events in modern markets. Unfortunately, in crypto, ADL is a fairly common occurrence, including on Friday.
Even worse, rumors circulated that certain large participants were exempt from the ADL process, thereby making it more likely that other participants would be subjected to it. This mechanism is problematic enough – when deployed in ways that aren’t uniform or transparent, it undermines trust in the market.
Crypto markets may also want to explore velocity logic (very short halts) to help order books rebalance. They’re not a cure-all; pauses can limit risk management abilities. While we’re at it, up/down limits should be off the table; they distort markets and delay price discovery.
Similarly, if crypto markets aspire to institutional credibility, then exchanges need to be just that: neutral venues for trading.
Exchanges that provide liquidity on their own venues – whether during liquidations or in the ordinary course of business – are inherently conflicted. In traditional finance, that’s a bright line. In crypto, it’s often blurred, and that’s a problem. If an exchange generates material revenues during liquidation events, then events such as deposits being suspended during large drops will appropriately be viewed skeptically.
A lot of what we saw on Friday was risk management in real time. Some did it well, others paid expensive tuition.
But let’s be clear. Liquidity providers did what they should be expected to do, and what mature markets anticipate they’ll do: reduce their size to reflect higher realized volatility, and then increase their size when markets normalize. That’s a repeated, proven pattern of rational market behavior that was again demonstrated here.
It's also important to note that volatility isn’t new. It’s not unique to crypto, and it’s not evidence of systemic failure. It’s part of how markets function globally – stress tests reveal where systems are resilient and where they’re not. The right takeaway from Friday is to review what happened and identify places where together we can strengthen infrastructure and resilience.
We should think critically about what Friday demonstrated. If DeFi is the blueprint for the next generation of markets, let’s build it the right way.
The Clarity Act and renewed engagement from regulators gives us a great forum to advance this work. Friday wasn’t a failure; it was a preview. The real test is whether we build on that lesson to create markets that can withstand the next one.