Four eyes are no match for 13,000 lines. That is the average number of draft allocations a DFM checks at every rebalance, according to our own data from a spectrum of 25 firms, comparing the platform’s draft line by line against the trade sheet it is supposed to match. Somewhere in those 13,000 lines, a human being is expected to spot the one row that may be wrong.
This is the step that sits underneath the manual re-keying, the lengthy rebalance windows, and the potentially high remediation costs that form part of DFM operations today.
Once a model has been re-keyed into a platform, someone has to check that what landed matches what was instructed, line by line, before the trades are instructed against on the client accounts. That check is the last line of defense.
What that looks like in practice is someone working through a spreadsheet row after row against a deadline that is already tight. Attention is not a constant resource, of course – it degrades over a long, repetitive task long before anyone notices it has, and a reviewer is not scrutinising line 11,000 with the same care they gave line 100.
Now, nobody sets out to miss a misallocated weighting. The process is asking a tired, time-pressured investment operations professional to sustain their concentration for hours at a stretch.
The second pair of eyes is supposed to be the safety net, and in principle it is a sound one. The four eyes principle, or dual control, sits at the heart of the FCA’s own governance expectations: Senior Management Arrangements, Systems and Controls (SYSC) 4 mandates implementation of “governance arrangements … including the segregation of duties,” while SYSC 5 says that “A firm should normally ensure that no single individual has unrestricted authority to … initiate a transaction” or “account for [that transaction]”.
In practice, though, any dual control process carries the same structural weakness once volume and time pressure build up: the second reviewer can end up rubber-stamping the first, approving on the strength of someone else’s clearance rather than independently re-checking the underlying data. The knock-on effects of error can be far reaching.
The FCA’s Consumer Investments Regulatory Priorities report for 2026 is explicit that there has been significant consolidation and rapid firm growth across platforms and Model Portfolio Service providers specifically, and that this comes with opportunities but also risks if controls do not keep pace.
Checking 13,000 lines by eye is not in most people’s job description; it is what the team drops everything else to do when a rebalance window opens, which means other business-as-usual work queues up behind it. Some DFMs describe the lengthy rebalance process as diminishing the ability to react quickly to market events.
As model ranges grow, so does the checking workload, and firms are left choosing between pulling in more staff for a task nobody wants to do or asking the same small team to hold concentration for longer each cycle. The regulator is already asking whether controls are keeping pace with that growth. For a growing number of firms, the honest answer is not yet.
And when something is missed, it is rarely just the process that gets the blame; it is whoever’s name was on the check. That creates its own quiet cost inside a team: defensive checking, second-guessing, and potentially an operations culture that starts treating every rebalance as a personal liability rather than a shared control.
Senior Manager Conduct Rule 1 requires reasonable steps to ensure the business is controlled effectively, and Rule 3 extends the same standard to how responsibilities are delegated.
But arguably, as DFMs’ operational burdens have multiplied, a process built on manual checks and double checks is only defensible as a “reasonable step” if audit trails can be clearly evidenced and automated controls are unavailable.
A missed discrepancy at the checking stage is exactly how a mistyped weighting survives long enough to become a £900k remediation bill.
Inefficient checking processes are also part of why a rebalance can take the best part of a month, because when an exception is caught - however it is caught - corrections will have to be made, and instructions resubmitted. The checking process is not a footnote to the operational risk in model portfolio rebalancing; it’s close to the centre of it.
Four eyes was never a bad idea, and it’s the best risk mitigation process many firms have had to avoid costly errors. However, it’s simply not designed for the volume of models that currently sit across the industry and platform spectrum and is not fair to the teams being asked to run them.
Without technology that identifies and flags changes centrally, four eyes is the modern-day equivalent of the needle in a haystack.
Tom Whittle is CEO of Tikker
This article was originally published in: Wealth DFM