First published on Tabb Forum

Drawing lessons from recent crypto market volatility, Ben Santos-Stephens, CEO & founder of ClearToken, the UK-based Digital Financial Market Infrastructure, examines what the 2008 Lehman Brothers collapse would have looked like in a 24/7 digital market. He argues that without true settlement finality, a legal certainty most current DLT systems lack, a major insolvency could allow settled trades to be unwound, paralysing the market and creating systemic risk.

After last week’s crypto market crash that saw a record $19bn forced liquidations, market watchers are wondering what went wrong, and what will happen when, not if, a similar collapse envelops the broader market. In September 2008, when Lehman Brothers plunged, markets froze, trust evaporated, and institutions scrambled to understand their exposures.

But what if that Lehman scenario had happened in a different kind of market – one that never closes, where trades settle instantly, but nothing is ever truly final? What if transactions could be reversed or disputed after the fact, or later be undone?

That world is arguably already with us, in crypto markets – today’s public blockchains are not final settlement systems.

Settlement finality means that once a financial transaction is completed, it’s done. Unconditional and irrevocable. It can’t be reversed or disputed. Think of it like handing over cash for a coffee: once the money is in the till and the coffee is in your hand, the deal is final.

In financial markets, this certainty is critical. It allows institutions to manage risk, allocate capital, and trust that what’s been settled won’t be undone. Without it, every transaction carries a lingering uncertainty, a shadow of doubt that undermines confidence because in a counterparty insolvency, transactions can be undone.

In 2008, liquidity dried up because institutions didn’t know who was exposed to Lehman. In a non-final market, this uncertainty would have been magnified. Previously, settled trades could have been reversed when later challenged in court, title over collateral disputed, and exposures unclear.

The panic would certainly have spread faster and deeper as court orders issued by the bankrupt Lehman entities sought to seize control of all assets traded or transferred by Lehman whilst insolvent and continuing to trade. The moment of insolvency, previously announced as 2 a.m. Eastern time on a Monday morning, would be fought over in court for years to come, but the freezing of assets would be instant.

The technical moment of insolvency is a critical moment in a 24/7 market, as it determines the point from which trades can be undone, and defines what trading was conducted while insolvent. This is something worth fighting over and therefore a focal point which could cause paralysis in the system if it wasn’t known with absolute and irrefutable certainty – which in bankruptcy is an almost impossible ask.

Without finality, the claims over margin and other collateral would have become uncertain. Clearing houses and banks, use final settlement systems operated by securities depositories, and rely on knowing what’s settled to calculate risk as do parties to bilateral transactions.

If trades can suddenly be undone, their models break down and so does trust in the system as the risk of contagion escalates. We saw this to a much more limited degree in 2008 as some trades which hadn’t settled with Lehman were cancelled. Knowing that many trades could be undone would make the market’s rational course of action to stop trading with every counterparty in case they were also insolvent but didn’t know it yet.

Regulators acted swiftly in 2008 to contain the damage. But in a 24/7 market without settlement finality, intervention becomes harder. What do you freeze? What do you unwind? How do you enforce rules when the supposed source of truth, the ledger in every system is itself fluid? It’s questionable whether regulators or central banks could have even stopped the contagion if in a 24/7 world. Bankruptcy rules tend to trump most other laws, and for good reason. We might surmise that in the US, many financial firms would have entered Chapter 11 bankruptcy protection whilst they worked out their positions, painful but probably not terminal. However, in the UK, where most of the world’s banks booked their trading activities in 2008, and still do, the situation would have been nothing short of catastrophic.

The UK has no Chapter 11 equivalent, and it is a criminal offence for company directors to allow their firms to trade while insolvent. Admittedly, insolvency is not bankruptcy, and with big enough lending programs from the Bank of England and the Federal Reserve, most banks could probably have been saved. But even central banks may have been hesitant to lend to banks that might be insolvent, risking the later seizure by courts of any assets pledged as collateral.

Some may say 2008 was nearly two decades ago, and markets have evolved and become more resilient. We now also have blockchains and DLT which can immutably record and settle transactions in real-time.

But in a Lehman scenario, unless the transactions recorded to these new systems are final, they can and will be unwound by the bankruptcy courts and the degree of uncertainty will be far worse than it was in 2008.

This is not hypothetical.

Imagine a situation where banks are lending to one another in the repo market (repurchase agreement) – which is used for the majority of short term secured financing – and one bank goes bankrupt. The asset legs in repo transactions could be unwound leaving lenders with unsecured positions and potentially facing insolvency themselves as a result causing the risk of mass contagion.

As it stands, nearly every blockchain and DLT implementation globally is not a final settlement system. That means they’re subject to transactions being unwound when there are bankruptcies.

We’re moving toward the infrastructure that would be desperately needed in a Lehman-like collapse within always on, tokenised markets running on DLT and blockchain infrastructure. And it’s possible to preserve stability should the unimaginable happen again. But we’re not there yet.