The Real Cost of Discretionary Wealth Management: Evidencing Fair Value Across the Whole Stack

Take a reasonably ordinary arrangement. A GBP 1.5 million client sits on a wrap platform at 0.28%, in a hybrid model portfolio service charging 0.30%, holding funds with a blended ongoing charges figure of 0.35%, paying an ongoing advice fee of 0.70%. That is 1.63% a year, or GBP 24,450, before transaction costs and before VAT on the discretionary fee.

Now ask the harder question. Which committee in that firm has ever seen the figure 1.63%?

Usually none of them. Operations signed off the platform. The investment committee signed off the DFM. The board set the advice fee two years ago. Each decision was documented, each was defensible in isolation, and the number the client actually pays has never appeared in a single governance paper.

The layer nobody owns

The price and value outcome in PRIN 2A.4 does not ask whether each component of a service is separately reasonable. It asks whether the price the client pays is reasonable relative to the benefits they receive. For an adviser-intermediated discretionary arrangement, the client receives one service delivered through four contracts.

Layer	Typical range	Who signs it off in most firms
Platform or custodian	0.10% to 0.35%	Operations or the ops committee
DFM management fee	0.15% to 0.75%	Investment committee
Underlying fund charges	0.10% to 0.80%	Nobody, inherited from the DFM
Ongoing advice fee	0.50% to 1.00%	Board or pricing group

The third row is the one that should worry you. Underlying fund charges are frequently the second largest layer in the stack, and in most firms they are not assessed by anyone. They arrive as a consequence of a DFM appointment rather than as a decision. When a DFM shifts allocation from passive building blocks into active satellites, the client’s total cost moves without a single approval anywhere in the chain.

What the aggregate actually looks like

The spread across propositions is wider than most rate card comparisons suggest, and the ranking changes depending on which layer you look at. The compositions below are illustrative, built from the typical ranges above to show how the layers stack rather than taken from a market survey. Run the same exercise on your own propositions and the totals will differ; the shape rarely does.

Two observations from that shape. The institutional structure carries the second highest discretionary fee in the set and the second lowest total, because the custody and fund layers are compressed. And the advice fee is the largest single layer in three of the four propositions, which means a fair value assessment that scrutinises the DFM and waves through the firm’s own charge is examining the wrong end of the stack.

That is not an argument for cutting advice fees. It is an argument for evidencing them at the same standard the firm applies to third parties.

Why layer-by-layer assessment fails

Three failure modes show up repeatedly in fair value papers.

Benefit double counting. The platform assessment credits consolidated reporting as a benefit. The DFM assessment credits consolidated reporting as a benefit. The advice assessment credits the review meeting where that reporting is discussed. One deliverable has justified three charges.

Comparator mismatch. Firms benchmark DFM A against DFM B on the management fee, and reach a conclusion about value. But the client is not choosing between two management fees. They are choosing between two stacks, and the cheaper fee frequently sits on the more expensive platform or brings the more expensive fund list with it. Our guide to DFM charges and how the layers work sets out the components; the comparison has to be run on the sum.

Segment blindness. Aggregate cost almost never distributes evenly. Percentage-based fees across four layers compound against small portfolios and in favour of large ones, so the client paying the highest total cost of ownership is often the one receiving the least intensive service. A firm-wide average of 1.5% can conceal a cohort at 2.3%.

Building the evidence

The FCA’s finalised guidance on the Consumer Duty is explicit that firms in a distribution chain have to consider the effect of the whole chain on value, including their own remuneration. A workable assessment at stack level needs five things.

1. A worked model per segment. Not a rate card. An actual pounds and pence figure for a representative portfolio in each client segment, with every layer named, VAT applied where it applies, and transaction costs estimated rather than omitted.
2. A benefit set written at the same altitude. What does this client receive for 1.63%? List it, and be honest about which items are genuinely differentiating and which are table stakes.
3. Cohort testing. Run the aggregate across the whole book and look at the distribution, not the mean. Set a threshold above which a client is flagged for individual review.
4. An outlier register with actions. Flagging is not evidence. What changed for the clients above the threshold? A tier renegotiated, a fund list switched, a service level raised, a client moved to a different proposition, or a documented reason why the cost remains fair.
5. A named owner. One committee accountable for the aggregate, with the number on its standing agenda. An investment committee with a proper charter and MI pack is usually the right home, because it already sees the fund and DFM layers.

Where the cost is worth paying

None of this is an argument that cheaper is better. It is an argument that the firm should be able to say why the number is what it is.

The cases where a high total cost of ownership stands up to scrutiny share a feature: the benefit is specific to that client and could not be delivered by a cheaper structure. Genuine tax-aware trading around an individual’s capital gains position. Holdings that cannot sit in a model, such as concentrated legacy stock managed down over years. Multi-entity reporting across trusts, corporate holdings and personal portfolios. Our piece on when bespoke management earns its fee covers where that line sits.

The cases that fail share a feature too: the client is paying bespoke prices for model delivery. A stack at 1.9% built from a standard risk-rated model, an annual template review and a platform chosen in 2019 is a fair value problem regardless of how carefully each layer was assessed on its own.

Three questions for your next fair value paper

Bring these to the committee that will own the aggregate.

• What is the total cost of ownership for the highest-cost decile of our book, and what is that decile receiving?
• If the fund layer moved 20 basis points tomorrow, would we know, and who would have to approve it?
• Can we produce, for any client, a single page showing every layer they pay and what each one buys?

The third is the one most firms cannot do inside a week. It is also the one a supervisor is most likely to ask for. Firms running structured oversight already have most of the inputs; see our framework for ongoing DFM oversight for the reporting cadence that makes it possible.

Cost is not the enemy. An unexplained cost is. If you are reviewing your investment proposition and cannot produce the aggregate number by segment, that is the first piece of work, ahead of any provider comparison. Further background on the regulator’s expectations of managers themselves sits in our note on Consumer Duty for discretionary fund managers, and the FCA’s Consumer Duty hub holds the current rule material.