Sometime in the coming months, the UK regulator, the FCA, will publish the findings of its multi-firm review into model portfolio services (MPS). The question it has been asking firms since late 2025 is deceptively simple: under the UK's Consumer Duty, are MPS providers delivering good outcomes, or just defensible ones?
For a lot of firms, the honest answer is we're not sure, because they've been measuring the wrong thing. For years the industry has reached for a single number to prove a portfolio represents good value: the Ongoing Charges Figure (OCF). It's clean, it's comparable, it sits neatly in a column on a factsheet. The trouble is that the OCF describes only one layer of cost, and it is rarely the layer that decides whether a client has actually been well served.
Cost has three layers; most firms measure one.
It helps to stop thinking of portfolio cost as a single number and start thinking of it as a stack. The first layer is fund cost, the OCF. This includes the management fee of the fund and
any other cost incurred on the individual fund level like trading cost within the fund. The second is implementation cost, the spreads you cross, the market impact you create, the platform dealing charges that land every time you trade. The third is the client cost stack, platform, custody, wrapper, DFM and advice fees sitting on top of everything else.
Most firms measure layer one and stop there. Which means most firms are making confident claims about "value" on the strength of perhaps a quarter of the actual picture.
What the OCF quietly leaves out
Here's why that gap matters, in two recent and very real stress tests, so let's start with the good news, because it's genuinely good. When markets seize up, ETFs have repeatedly proven to be the resilient part of the system rather than the fragile one. Through the Covid shock of March 2020 and again during the tariff turmoil of April 2025, ETFs kept trading, kept providing price discovery, and gave investors a way to access liquidity precisely when the underlying markets were at their most fearful. The structure did its job. Anyone who feared ETFs would break under pressure has now watched them pass two of the most severe tests markets can throw at an instrument.
But resilient is not the same as free. As State Street's own analysis of that April 2025 episode showed, spreads still widened materially even as ETFs held up - the average bid-ask spread on UCITS equity ETFs more than doubled, from around 0.37% to roughly 0.85%, in the space of days. Anecdotal evidence during the 2020 Covid crash tells the same story: spreads widened from January 2020 to March - as the cost of transacting in the underlying market spiked. The OCF on those funds didn't move a single basis point through either event. The structure stayed sound; the cost of trading it did not.
Tariff Tantrum, What Tantrum? How ETFs Performed During Liberation Day Turmoil by State Street Investment Management
And that is the point that matters for portfolio managers. A manager who rebalanced into or out of those positions during the worst of the volatility paid a real, additional cost, one that never appears on a factsheet and never touches the OCF. The instrument was resilient. The timing of the trade was expensive. Both things are true at once, and only one of them shows up in the number most firms rely on.
That is the heart of the issue. The OCF is a sticker price. Real cost is what the client pays at the moment of execution and that depends on a series of things the OCF was never designed to capture:
- Turnover and transaction costs. A cheaper ETF rebalanced too often can quietly hand back every basis point it was supposed to save.
- Spreads and execution timing. ETFs trade on exchange, and spreads widen exactly when markets are stressed and clients are most exposed. This is the worst possible moment to be trading without discipline.
- Tracking differences and securities lending. The realised net cost of an ETF can diverge from its headline figure, sometimes in the client's favour, sometimes not.
- The full client cost stack. The platform, wrapper and advice layers are routinely larger than the OCF difference between two competing ETFs in the first place.
Optimise the one number cost you can see, and you can still lose on the three you can't.
Under Consumer Duty, the question has changed
This is precisely where the Consumer Duty rewrites the brief. Under the price and value rules (PRIN 2A.4), the regulator is not asking 'is this portfolio cheap?' It is asking something far harder to game: is the cost justified by the outcome delivered?
That distinction has teeth. A 0.10% portfolio carrying drift, unnecessary turnover and thin governance can fail a fair value assessment. A 0.25% portfolio built systematically, governed properly, and able to justify every component it holds will very likely pass. The headline number is not the test. The process behind it is.
And the evidence points the same way. When the FCA reviewed firms' fair value frameworks, the ones that struggled weren't the expensive firms, they were the firms that couldn't evidence their reasoning. Cheapness is not a defence, demonstrability is.
The construction decisions that actually drive cost
So, what does demonstrating value require in practice? Three things, none of which appear in an OCF.
Component-level cost transparency. You should be able to say what each ETF contributes to the blended cost and why it earns its place. It was the cheapest is not a reason. It delivers this exposure, at this cost, with this tracking quality - and here is the alternative we considered and rejected.
Rebalancing discipline. Threshold-based rebalancing, trading only when the portfolio has genuinely drifted far enough to warrant the cost, consistently beats rigid calendar rebalancing that trades on the date regardless of whether it's worth it. Every unnecessary trade is a spread you didn't need to cross.
Continuous monitoring. OCFs change. Fund structures evolve. A blend that was cost-efficient at launch drifts out of efficiency quietly and without announcement. An annual review isn't monitoring it's a snapshot of a moving target.
What this looks like in practice
Running model portfolios on adviser platforms is rarely as clean or as measurable as the theoretical framework suggests. While none of this is abstract for us, it's the discipline we aim to apply every day, the framework exists so that value isn't just something we try to deliver, but something we can evidence.
The challenge is that real‑world implementation data is messy. Obtaining a full attribution of trading costs is difficult, and access to the actual spreads at which ETFs were executed is almost never available at the model level. That limitation is universal across platforms.
Despite that reality, we operate with a clear playbook for managing our ETF Model Portfolios throughout their lifecycle. It's designed to capture the costs we can measure, control the costs we can influence, and govern the areas where transparency is structurally limited.
The building blocks are chosen, not assumed.
ETFs are the bricks our portfolios are built from, so the selection process is deliberately demanding. Every fund has to clear a series of filters before it earns a place - and lowest OCF is only one of them. Tracking quality, on-screen liquidity and spreads, fund scale, replication structure and the realised drag a fund delivers all matter, because the cheapest headline figure is not the same as the best-value building block. The result is a roster we can justify at the component level, exactly the transparency a fair value assessment demands.
The weights are optimised, not guessed.
We run a third-generation Black-Litterman optimiser monthly, targeting the best achievable performance per unit of risk for each mandate. Black-Litterman matters here because it produces stable, well-diversified weights rather than the concentrated, corner-solution allocations that naïve optimisation tends to throw out, which means less unintended factor risk and less of the drift the regulator treats as outcome drag.
Re-optimising monthly keeps each portfolio aligned to its risk target, but re-optimising is not the same as trading. We act on a recommended change only when the benefit justifies the cost of making it; small, expensive rebalances simply don't happen. That single piece of discipline is where the savings from choosing best-value ETFs are protected rather than handed straight back across the spread. The output isn't a discretionary call that's hard to explain after the fact, it's a defined, repeatable process that maps directly onto the portfolio's stated objective.
The inputs are forecast with data, not opinion.
What some firms describe loosely as house views, we treat as a data-science problem: forecasting expected returns by combining forward-looking macroeconomic indicators with market data. That distinction matters for fair value, because it turns the most subjective part of any optimiser into something systematic, evidenced and reviewable rather than a matter of conviction.
The risk is measured both ways.
We monitor risk ex-ante and ex-post, the risk we intend to take and the risk we actually took, to keep every portfolio true to its mandate. For target-risk portfolios this is the whole game. The promise to the client is a risk level, and we can demonstrate, continuously rather than once a year, that the portfolio delivered it.
Put together, that is what lets us answer the regulator's real question. We are not asserting that our portfolios are cheap. We are evidencing that the cost each client bears is justified, by best-value components, systematic construction, data-driven inputs, and risk measured against the mandate rather than assumed to be on target.
The better question to ask
Which brings us to the reframe that matters. Most firms still ask: 'Are our portfolios low cost?' The better question, the one the Consumer Duty is effectively forcing, and the one this year's findings will reward is: 'Can we demonstrate that every cost our clients bear is justified by what it delivers?
That single shift moves the whole conversation away from headline OCFs and toward the things that determine value: construction methodology, implementation discipline, monitoring and governance, and the full client cost stack.
ETF-based portfolios can deliver genuinely exceptional value, we see it work every day. But low cost and fair value are not the same claim, and only one of them survives contact with a regulator. The firms that thrive over the next few years won't be the ones with the smallest number on the factsheet. They'll be the ones with the infrastructure to measure real cost across all three layers, and the evidence to prove the value sitting behind it.
The factsheet shows the sticker price. The question is whether you can show everything else.
Irene Bauer