The Physics of FX

The foreign exchange (FX) market with $9.6 trillion in daily global turnoverⁱ runs on T+2 infrastructure built for a world that no longer exists, even as equity markets have moved to T+1 in North America and will move by October 2027 in the U.K. and Europe. The physics are different.

Schrödinger’s cat is a thought experiment where a cat sealed in a box with a randomly triggered poison is considered both alive and dead until someone opens the box to see what happened. If you’re using FX to hedge other trades, you don’t know whether the FX leg has settled and can’t optimize until both settlement windows resolve. You can’t tell if that hedge is alive or dead.

Heisenberg’s uncertainty principle says you can measure a particle’s position or its momentum with precision, but not at the same time. If you’re managing FX as an asset class, your trading happens across a fragmented set of electronic communication networks (ECNs) where each allows for only partial price discovery. Optimizing for execution means sacrificing visibility across liquidity and counterparty costs and vice versa.

When FX is a hedge

The hedging thought experiment is a European fund taking a position in U.S. equities. When it does, it carries currency exposure from the moment the trade settles under T+1. But under T+2, the FX hedge against that currency exposure arrives a day later. There’s a delay in knowing whether the hedge did or didn’t provide an effective cushion. The portfolio manager is carrying live currency risk on an active position during the window the hedge is supposed to cover.

The infrastructure built to address this kind of asymmetry only partially resolves it. In 1974, Herstatt Bank collapsed mid-settlement day, leaving counterparties holding Deutsche Mark payments with no corresponding dollar receipts. The Continuous Linked Settlement (CLS) system was built to ensure that neither currency in a trade pays out until both parties are ready. Within defined daily settlement windows, CLS now covers 18 currencies. But equity positions now settle a full day before the FX hedges against them do, meaning that CLS protections don’t yet apply.

The timing gap degrades the hedge in two ways. Same-day FX closes the gap at a spread premium; pre-positioning the currency instead runs the hedge before the underlying exposure exists. That creates a window where the hedge is active, but the underlying equity exposure isn’t yet confirmed. Until the T+2 window closes, a portfolio manager optimizing hedge ratios is working against an exposure that isn’t yet resolved. When the UK and EU move to T+1 in October 2027, the same constraint extends to GBP and EUR pairs, adding two of the most widely hedged European currency pairs to the problem.

When FX is an asset class

Trading FX as an asset class creates a different kind of paradox because of the measurement uncertainty in a market built on fragmentation. FX liquidity splits across EBS (CME Group), Hotspot (Cboe), 360T (Deutsche Börse), and Reuters Matching (LSEG). Each of these allows for partial price discovery rather than providing a consolidated picture. A large order through any single venue is invisible to the rest of the market. But major dealer banks route across all of them simultaneously, so they see the full picture. Buy-side participants trade with counterparties who know the consolidated picture, and the spread they capture on every execution reflects that asymmetry.

That asymmetry is another function of generational market history. In the FX rigging scandals of the early 2010s, dealers used their knowledge of client end-of-day positions to trade ahead of the fix. Post-crisis reforms tightened surveillance, which led to ICE, CME, Euronext, and LCH acquiring ECN platforms.

But still, around 90% of FX turnover trades OTC, between institutions that built this infrastructure and have limited incentive to change it. Unlike equities, where DTCC provides centralized trade matching, FX runs on bilateral confirmation, which cascades through the investment manager, executing broker, FX prime broker (FXPB), and custodian. Each party has its own systems with different cut-off times, and a missed cut-off anywhere in that chain can leave a settled equity position without a confirmed FX leg.

Basel III/IV has made the bilateral model progressively more expensive to run. Under current capital rules, bilateral counterparty credit exposure consumes regulatory capital at higher rates than the original FXPB relationship structures anticipated. Major FXPBs have responded by raising minimums, pushing a growing tier of hedge funds and multi-strategy platforms to wider spreads, fewer counterparties, and more operational overhead. For these managers, the cost of accessing institutional FX is rising even while T+1 makes the operational demands harder.

Three forces compounding the pressure

No regulatory mandate to compress FX settlement exists yet, but T+0 matching platforms are putting pressure on T+2’s operating justifications. When mismatches surface on trade date and clear before settlement begins, the case for holding two days of runway weakens.

Crypto markets also provide a compelling continuous settlement model. With the use of stablecoins, FX could move to round-the-clock execution and atomic clearing with finality in seconds. JPMorgan launched a dollar deposit token with Mastercard and B2C2 testing 24-hour near-instant settlementⁱⁱ; UBS and Ant International ran tokenized-deposit cross-border paymentsⁱⁱⁱ; and Tradeweb completed the first real-time on-chain financing of U.S. Treasurys against USDC^iv. As Tradeweb’s 2025 annual letter put it: “There is no doubt markets are headed toward 24/7 trading.”^v

What this means for your operations

We think the continuous settlement model is the right destination. But the gap between where the market is going and where the infrastructure sits today is your operational problem to manage now.

Settlement-cycle visibility across both the equity and FX legs of every position is the starting point. A fund that can’t see its combined exposure on trade date is making hedging decisions with incomplete information. Relying on manual reconciliation or end-of-day reporting at this point is an unquantified risk, not a workflow preference.

October 2027 marks the next hard deadline. GBP and EUR hedges will carry the same structural mismatch as USD hedges, with additional time-zone constraints that make the affirmation window tighter in both transatlantic directions.

Counterparty concentration demands active monitoring. As FXPBs consolidate their books, it matters where single-counterparty FX exposure sits more than it did three years ago. A change in terms or a relationship exit can create significant disruption in a market, let alone the contraction in accessible prime brokerage relationships.

Finally, data coherence across fragmented ECNs now separates a fund making hedging decisions on the complete picture from one working venue by venue. Without a unified view of currency exposure, portfolio decisions contain structural blind spots that compound across positions.

The principle applies whether you're already trading FX or considering it. The goal is to tighten the way you see, fund, and confirm both legs of every position so you’re not betting on what’s happening in a space you can’t observe. First, prove you can see combined equity–FX exposure at your trade date, such that you can reconcile it across venues in real time. But be prepared. At present, this market still has an inherent structural uncertainty. Even though you can compensate for it, it's still there.