In my view, the answer depends on the asset and the stress scenario.
The infrastructure today is much stronger than it was two years ago when speaking about Bitcoin and Ethereum exposure through large-listed products. Moreover, the authorised participant (AP) and market-maker ecosystem is deeper, while in-kind creation and redemption are now permitted for many US crypto exchange-traded products (ETPs).
Plus, major products already rely on large traditional names across creation, redemption, trading, and custody workflows. Naturally, all of this brings crypto closer to the model investors already understand from commodity ETPs.
The key problem is that a bank brand can create expectations that go beyond what the legal form of the product actually supports. Investors may assume daily liquidity, tight spreads, high counterparty quality, and predictable redemption behaviour, which can be ensured in normal conditions for the largest assets.
Under stress, however, there might be venue fragmentation, reliance on a limited group of trading counterparties, fiat rail unavailability, blockchain congestion, and the inability of APs to keep arbitrage working when spreads widen quickly.
The APs and arbitrage mechanism are what keep ETP prices close to net asset value (NAV), while NAV itself often depends on pricing consolidated on multiple crypto venues, so when several of those venues halt at once, the whole model ends up under pressure simultaneously.
Fragmentation across exchanges, OTC desks, custodians, and jurisdictions is widely acknowledged as a problem. Is that primarily a technology gap, a regulatory gap, a commercial one, or something else? And which is hardest to close?
I do not see fragmentation as a technology gap. Technology can connect venues, route orders, automate settlement instructions, and consolidate reporting.
The more serious problem is that digital asset markets grew around vertically integrated exchanges, bilateral OTC desks, separate custodians, and local regulatory regimes. Each part of the chain has its own credit and fee model, asset list, and custody rules.
This makes fragmentation a commercial and regulatory gap at the same time. Commercially, venues have little incentive to give up control over liquidity, client data, and collateral, while regulators treat the same activity differently depending on the region.
The hardest to close is the commercial aspect around credit and liability, because a unified execution screen is not enough. Institutions need to know who carries settlement risk, who guarantees delivery, how collateral is transferred, and what happens if a venue fails.
Unless these questions are answered, fragmentation — six legal entities, six boards, six local teams all expected to operate under one standard — remains an operating risk.
When an institution wants to execute at real size, what does the execution stack look like today, and at what point does it break down?
For a serious institutional trade, the stack usually starts before execution.
The asset has to pass internal approvals, compliance checks, venue eligibility, custody rules, and pre-trade risk limits.
Then the order goes through an order management system (OMS) or execution management system (EMS), often into an execution layer that can access OTC desks, exchanges, prime brokers, and sometimes derivatives venues for hedging. From there, the trade may be split between request for quote (RFQ), algorithmic execution, and block liquidity, while settlement is handled through a custodian, prime broker, exchange account, or off-exchange settlement model.
At a small size, this is manageable, but at a substantial one, the system starts showing its seams. Market impact is the first thing that breaks, as when the trade is too large for visible order books, the institution has to depend on OTC liquidity and internalisation. The second is collateral location, because if funds are held with a custodian, but liquidity is scattered on multiple exchanges, the institution either pre-funds venues, uses credit lines, or accepts slower settlement. The third thing that breaks is post-trade reconciliation, because books, wallets, custodians, administrators, and risk systems do not always update at the same pace.
Risk and operations teams in traditional finance are built around market hours and batch settlement. What is the single biggest operational failure point you see when institutions try to run a digital asset book alongside a conventional portfolio?
From my experience, the biggest failure point is the mismatch between real-time markets and batch-based control systems. Basically, digital assets trade continuously, and collateral moves continuously. Risk, in turn, can materially change on a Sunday night, with many traditional systems still assuming end-of-day valuation, overnight reconciliation, weekday settlement windows, and human escalation during office hours.
This is dangerous. The trading book may be live while the risk, treasury, compliance, and finance functions are only partly live, which makes a margin call or, say, a withdrawal queue, or wallet approval happen when the conventional operating model is essentially asleep. This is the point where the largest losses occur.
I’d also highlight governance among the weakest points. Institutions need 24/7 risk ownership, automated limits, real-time collateral visibility, and a clear escalation model. Without them, a digital asset book becomes an always-on position inside a mostly five-day operating machine.
Is the 24/7 nature of digital asset markets a genuine advantage for institutions, or is it mostly a liability until the broader financial system can catch up?
I would call it a genuine advantage that currently behaves like a liability for most institutions.
Continuous markets mean capital does not sit idle waiting for a settlement window to open. In tokenised markets specifically, programmable payment flows eliminate the frictions that batch systems impose by design.
That said, institutions do not operate in isolation. They depend on a chain of counterparties — banks, administrators, auditors — none of which match the market speed. If those functions are not available in real time, 24/7 trading creates a lopsided model in which the asset trades all the time while fiat settlement and internal controls still depend on office hours.
There is also a liquidity point worth making, as an open market does not mean a deep one. Weekend liquidity is thinner, and when fewer participants are active, price discovery reflects that: spreads widen and price swings can be exaggerated.
So, 24/7 access becomes genuinely valuable only when treasury and compliance infrastructure can run at the same speed as the market itself. For most institutions today, though, that condition is not yet met.
With MiCA live, a US framework taking shape, and the UK regime in progress, are we heading toward meaningful convergence or toward several incompatible regimes that force providers to build parallel infrastructure?
I believe we are heading toward partial convergence in principles and divergence in implementation. Regulators broadly want the same thing: segregated assets, proper disclosure, and someone responsible for failures.
The divergence is in perimeter, timing, and product treatment. MiCA in the EU is already live, the US passed the GENIUS Act in 2025, while the UK legal regime does not come into full force until October 2027, which means providers cannot build once and deploy everywhere.
The rules governing the same activity look different depending on where you operate.
A single global control architecture is achievable, but a licence alone does not solve the problem. Getting authorisation is easy, yet embedding each legal entity into onboarding, reporting, product eligibility, and internal controls is much harder.
Firms that treat authorisation as the finish line tend to discover the real work starts after.
Where do you see the biggest regulatory arbitrage risk emerging over the next two years? And which jurisdiction is most likely to attract a disproportionately high level of activity?
The biggest arbitrage risk over the next two years sits in stablecoins, yield, offshore derivatives, and custody standards. Jurisdictions treat the core questions differently — who can issue, on what reserves, and whether intermediaries can offer yield at all — which creates room for activity to concentrate wherever legal certainty and commercial flexibility overlap.
The US is likely to attract a disproportionately high share of stablecoin and dollar-liquidity activity because, for the first time, there is a federal framework that regulated institutions can actually explain to their boards.
At the same time, I would watch the UAE and Hong Kong.
Dubai’s Virtual Asset Regulatory Authority has a dedicated virtual asset regime with a functioning public licensing framework.
Meanwhile, Hong Kong’s stablecoin issuer regime came into force on 1 August 2025, and appears to be one of the first jurisdictions globally to reach that point.
If you had to identify the one problem that, if solved, would unlock the next phase of genuine institutional adoption and deep integration, what would it be?
For me, the core problem is the absence of a trusted settlement and collateral layer connecting venues, custodians, and fiat rails.
As a matter of fact, institutions already have the access and execution infrastructure. But what they still lack is a common operating layer that lets them transfer collateral and reconcile positions with the same confidence they have in traditional assets.
Yes, more ETPs bring in new capital, more licenses foster confidence, while more exchanges deepen liquidity. Even so, real integration needs, first and foremost, a predictable post-trade infrastructure.
If an institution cannot see collateral in real time or net exposures without manual workarounds, crypto stays a satellite allocation rather than a fully integrated asset class.
Next interview →
SCRYPT
Norman Wooding
Image: B2PRIME Group 
