Why Financial Market Infrastructure Is Built the Way It Is

On September 15, 2008, Lehman Brothers filed for bankruptcy.

What followed in the next 72 hours was not just a financial crisis. It was a stress test of every assumption embedded in the infrastructure of global capital markets — and most of those assumptions failed.

I was not working in financial markets at the time. I was running a business. And what I remember most is not the operational paralysis inside the banks — it is the silence that followed outside them. Credit lines suspended without notice. Counterparties who stopped answering calls. The sensation that the rules of the game had changed overnight, and nobody could tell you what the new rules were.

That experience — watching a financial system seize from the outside, without the vocabulary to explain exactly why — stayed with me. It is part of what brought me, eight years later, to blockchain technology. Not as a financial operation. As an attempt to understand what an alternative infrastructure might look like: a system where the rules of settlement, transparency, and risk management were embedded in the protocol itself, not delegated to bilateral trust between counterparties. And whether the problems I had observed in 2008 were features of finance in general or of a specific way of building it.

The answer, it turns out, is the latter. The problems were infrastructural. And infrastructure can be rebuilt.

The problem nobody talked about

The public narrative of 2008 focused on subprime mortgages, leverage, and regulatory failure. All true. But underneath that narrative was a structural problem that only became visible when the system broke: market infrastructure had been built on bilateral trust between counterparties, and bilateral trust evaporates instantly when one counterparty defaults.

Lehman had over 906,000 derivative transactions outstanding at the time of its bankruptcy, with an estimated notional value of $35 trillion. The vast majority — 96% — were bilateral OTC agreements. No central counterparty. No standardised margining. No clear settlement procedure. Just contracts between two parties, one of which no longer existed as a functioning entity.

The resolution took years. Some contracts were still being disputed in courts in 2012 — four years after the default. Meanwhile, the markets that had operated through central counterparties told a completely different story: LCH’s SwapClear managed Lehman’s $9 trillion interest rate swap portfolio through five competitive auctions in a matter of weeks, using pre-posted margin. No systemic contagion. Orderly resolution.

Same defaulting entity. Radically different outcomes. The only variable was infrastructure.

Why rules look the way they do

Every regulation in financial market infrastructure exists because something broke.

T+2 settlement exists because longer settlement cycles create counterparty exposure that cannot be managed when markets move fast. Mandatory central clearing of standardised OTC derivatives — introduced through Dodd-Frank in the US and EMIR in Europe — exists because 2008 demonstrated what happens when $35 trillion in bilateral exposure unwinds simultaneously. Segregation of client assets exists because the alternative is insolvency administrators deciding whether your assets belong to you or to the estate of a failed institution.

These are not bureaucratic choices. They are engineering decisions made after observing failure at scale.

This matters for anyone building in financial markets today, because the same logic applies. The question is never “how do we avoid regulation.” The question is “which failures is this regulation designed to prevent, and are those failures still relevant to what we are building.”

Most of the time, they are.

What changed after 2008 — and what did not

The G20 Pittsburgh Summit in 2009 produced a clear mandate: standardised OTC derivatives should be centrally cleared, reported to trade repositories, and subject to higher capital requirements if cleared bilaterally. This was implemented across jurisdictions over the following five years.

The result was a structural shift in how risk is managed in derivatives markets. Central counterparties became the backbone of market stability rather than an optional service for exchange-traded products. Margining became standardised. Transparency improved significantly.

But not everything changed. Settlement cycles remained long. Cross-border reconciliation remained fragmented. The fundamental architecture of how assets move between counterparties — built on messaging protocols and correspondent relationships that predate the internet — remained largely intact.

This is the gap that became visible again. Not because another crisis exposed it, but because technology advanced to the point where the gap became measurable. When you can settle a token transfer in seconds on a distributed ledger, a two-day settlement cycle stops being a technical constraint and starts being a choice.

The continuity that most people miss

There is a tendency in financial technology to frame innovation as a break from the past. New infrastructure replacing old infrastructure. Distributed systems replacing centralised systems. Code replacing counterparties.

This framing misses something important.

The principles that make financial market infrastructure reliable — margining, segregation, transparency, orderly default management — are not artefacts of legacy technology. They are responses to real failure modes that exist regardless of the technology layer. A distributed ledger does not eliminate counterparty risk. It changes where that risk is managed and how. A smart contract does not eliminate the need for clear rules about what happens when conditions are not met. It encodes those rules differently.

The infrastructure being built today for digital asset markets is not starting from zero. It is inheriting the accumulated knowledge of every failure the traditional system has experienced — and, in some cases, repeating the same mistakes before learning the same lessons.

The regulators working on MiCAR, on the DLT Pilot Regime, on MiFID II’s application to tokenised financial instruments — they are not inventing new problems. They are applying old solutions to a new technical context. Understanding why those solutions exist is not background knowledge. It is operational intelligence.

Why this matters now

The transition period currently underway in European digital asset markets — with MiCAR fully applicable since December 2024 and the DLT Pilot Regime creating space for experimentation in market infrastructure — is not primarily a compliance event. It is an infrastructure event.

The questions being answered right now — how tokenised assets are settled, how client funds are segregated on-chain, how default management works when positions are held on a distributed ledger — are the same questions that were answered for traditional markets after 2008. The answers will shape the architecture of digital capital markets for the next decade, in the same way that post-crisis reforms shaped the architecture of traditional markets for the past fifteen years.

Knowing where we came from is not optional context. It is the foundation for building something that actually works.