Andrew Walker

Not every exception becomes a settlement failure, but under T+1 it has a much higher chance of becoming one.

Accelerated settlement changes one of the most important variables in exception management: the time available to do something about the problem. Under T+2, a break discovered late in the day could often join the following morning's workload. Someone would investigate it, speak to another team, correct the data or contact a counterparty.

Under T+1, that window becomes significantly smaller. Exceptions have not suddenly become more complicated. They have simply become much more urgent.

Tomorrow morning may already be too late

An exception queue is effectively a prioritisation mechanism. Some issues need attention immediately, some can wait and others will resolve through an existing process.

That works when there is enough time between identifying a problem and the point at which it becomes consequential. T+1 changes that calculation.

An incorrect instruction, missing field, mismatched identifier or late allocation can quickly affect the ability to complete the next stage of the process. Each delay consumes a much larger proportion of the remaining settlement window.

The FCA estimates firms will have around 80% less processing time under T+1 and is specifically encouraging organisations to uncover manual arrangements that delay settlement or contribute to failures.

For operations teams, that creates a fairly unforgiving reality. It is no longer enough to know that an exception will eventually be resolved. You need to know whether it will be resolved in time.

The best settlement exception is the one you never create

Exception management tends to focus naturally on resolution. Something has gone wrong, so the task is to find it, investigate it and fix it.

Faster settlement increases the value of prevention.

Many operational exceptions originate well before the final settlement stage. Missing or inconsistent data, incorrect instructions, unusual combinations of inputs or rules applied differently in disparate systems can all create problems later.

Catching those issues at the point of entry is much easier than untangling them downstream. That requires good data validation and predictable decision logic.

Rather than accepting information, allowing it to travel through several stages and then discovering an issue during reconciliation, firms should increasingly ask whether the relevant rule can identify the problem immediately. Is the data complete? Does it meet the expected format? Does the transaction satisfy the relevant business conditions? Is an exception genuinely unusual, or is somebody performing the same manual check hundreds of times a week?

There is a significant difference between automating the handling of an exception and preventing unnecessary exceptions from being created in the first place. Under T+1, both matter.

Manual processes introduce unpredictable time

There is another problem with manual exception management: nobody can tell you precisely how long it will take.

An email may receive an answer in two minutes or two hours. A spreadsheet may be picked up immediately or wait in a queue. An investigation may require one person or bounce through several teams before reaching whoever can actually resolve it.

That variability becomes dangerous when the available window is fixed.

North America's experience is instructive. PostTrade 360 has reported that manual processing continued to create pressure following the US move to T+1 and that some organisations increased staffing to absorb the additional workload. It also reported growing investment in automation as firms looked to remove manual activity from their operating models.

More people can increase capacity, but they cannot make a fundamentally unpredictable process predictable.

Clear ownership matters

Anyone who has spent time around a complex operating model will recognise another familiar problem: the exception that seems to belong to everybody until somebody actually needs to own it.

A trade break can move between investment operations, middle office, settlements, technology teams and external providers. Each hand-off takes time and each creates another opportunity for context to be lost.

Under T+1, firms need much greater clarity around what has happened, why it happened, how urgent the issue is, who owns the next action and which rule determines that action.

This is also where exception-first monitoring becomes much more valuable. A dashboard showing that the vast majority of transactions are progressing normally may be reassuring, but an operations team working against a compressed deadline needs something more useful: a clear view of the transactions that are not progressing normally, ordered by what needs attention now.

Speed does not mean rushing

Fast exception resolution should not mean operational teams simply working harder or with greater urgency. It should mean designing the process to remove unnecessary delay.

That means detecting problems as close as possible to the point where they originate, using clear and repeatable rules to classify them, automatically validating or enriching information where appropriate and routing exceptions directly to the right process or person.

It also means retaining a record of the decision logic so the organisation can understand what happened and improve the process later.

In other words, pace should be engineered into the operating model rather than supplied by human urgency.

Where rules technology can help

A rules engine such as Heywood Idiom can help with many parts of this problem by applying consistent business logic, validations and calculations automatically.

Rules that currently depend on manual interpretation, spreadsheets or bespoke legacy tools can be embedded into operational workflows and applied (or executed) consistently and at scale. This can help identify invalid or incomplete information earlier, automate predictable decisions and reduce some of the manual friction surrounding exception handling.

Crucially, Idiom can be introduced alongside existing systems. The objective is not to rip out functioning settlement technology, but to identify areas where rules or calculations create delay and introduce greater control around them.

Not every exception can or should be automated. There will always be genuinely unusual cases requiring experienced human judgement. The opportunity is to stop asking those people to spend their time dealing with problems that are repetitive, predictable and rules-based.

Under T+1, every hour spent unnecessarily investigating, rekeying, reconciling or chasing brings a transaction closer to failure. The firms best prepared for accelerated settlement will therefore not simply become quicker at fixing problems. They will become much better at preventing predictable ones from occurring in the first place.

Get ahead of the exceptions

If exception management is becoming a key part of your T+1 planning, speak to Heywood about where rules-driven automation could help reduce manual intervention and identify issues earlier in the process.

We can work with you to understand where exceptions originate, which decisions can be automated and how Idiom could complement your current systems without requiring a full re-platform.

Talk to Heywood about improving exception management for T+1.