In 2026, the slogan on many private-bank pitchbooks still treats open architecture versus in-house funds as a branding choice. For families with substantial, multi-entity wealth, it is a governance question. True independence means the adviser can select third-party funds, separately managed accounts, private-market vehicles, and cash without a captive product factory, and that remuneration does not quietly prefer in-house vehicles.
Families who already sit across several banks hear the same phrase in almost every beauty parade. What used to mean access to other people’s products now often means a guided list, an affiliated asset manager, and a fee stack that is hard to read unless someone rebuilds it line by line. This note is educational, not personalised advice: mandates still depend on facts, licensing, and counsel in each jurisdiction.
What open architecture used to mean
In the late 1990s and through the 2000s, private banks discovered that wealthy clients would no longer accept a shelf made only of house funds. Open architecture, in that first generation, meant the bank would custody and advise on third-party UCITS, hedge funds, and later private vehicles sourced from external managers. The relationship banker was supposed to search the market, not only the group’s own factory.
That promise was never as clean as the marketing suggested. Access already ran through preferred platforms, due-diligence committees, and distribution agreements. Still, families could hold a competitor’s flagship strategy next to a house money-market fund. Two later changes then altered the phrase. Large groups built or bought asset-management companies whose economics sit on the same consolidated profit and loss as the private bank. European Union rules forced a sharper split between independent advice and advice that may still receive third-party payments. Pitchbook vocabulary did not always keep up.
Guided lists and in-house factories in 2026
By 2026, many so-called open platforms are guided architectures. A guided list can be a legitimate way to concentrate research hours. The problem starts when the list is presented as the whole market, when replacement of a manager is slow unless the replacement is affiliated, or when the economics of the list cannot be audited. Typical features include a short preferred list for each sleeve, a heavier operational burden for any fund that is not on that list, retrocessions or other distribution payments where the firm is not acting as an independent adviser, and an affiliated management company whose products appear with reassuring frequency in model portfolios. Together they recreate a factory, only with better lighting.
In the European Union, MiFID II still frames the inducement question. The ESMA Interactive Single Rulebook for MiFID II collects the directive text, including the investor-protection articles on information to clients, conflicts, and the conditions under which a firm may describe its advice as independent. At a high level, a firm that holds itself out as independent cannot accept and retain third-party monetary benefits in respect of that advice. A firm that does not claim independence may still receive inducements, subject to quality-enhancement and disclosure tests that supervisors continue to police.
The European Securities and Markets Authority publishes questions and answers that national authorities apply. In France, the Autorité des marchés financiers is the supervisor families should expect to see cited in a CIF or investment-firm file. This article does not invent a 2026 article number. The operational test is simpler: if the person who selects the fund is paid, directly or through group profit, when the house product wins the ticket, the architecture is not open in the sense that independence requires. In the United Kingdom, families who still use a UK-regulated firm should read the same conflict through the Financial Conduct Authority Consumer Duty: a high-level duty to deliver good outcomes, including on product value and foreseeable harm from conflicts. Professional-client classification does not erase the economic conflict. It only changes which retail-protection rules apply on paper.
How in-house funds quietly concentrate risk and fees
In-house funds are convenient. The bank already knows the operations, the reporting pack is standardised, capacity can be reserved, and the relationship team does not have to negotiate a new ISDA, side letter, or transfer-agency setup. Convenience is a real service. It is not the same thing as executing the family’s investment policy.
Fee stacking is the first quiet cost. The family pays an advisory or discretionary fee at the private bank, a total expense ratio inside the fund, sometimes a performance fee, and, where the model is not fee-only, a distribution trail that may be described as a retrocession, a rebate, or a platform payment. Each line can look modest. The look-through total is what compounds. Families should insist on a single schedule that adds every layer, including money-market and cash vehicles often treated as “just liquidity.”
Concentration is the second quiet cost. A house equity fund, a house credit fund, and a house private-markets feeder can share the same credit committee, valuation calendar, key-person risk, and incentive to keep capital inside the group. The portfolio then looks diversified on a factsheet and correlated in a stress. Conflicts sharpen when manufacturer and adviser share a profit and loss. If replacing a mediocre house fund with a competitor reduces group revenue, the replacement file needs a written policy that the family’s interest prevails, and evidence that the policy has been used.
Independence as a process, not a brochure word
Open architecture is a process. It starts with an investment policy statement that names permitted vehicles, prohibited conflicts, liquidity buckets, and who may override a preferred list. It continues with a manager search that can include firms the incumbent bank does not distribute, and a replacement protocol with dates: when a strategy is on watch, who decides, and how long the family will wait before capital moves.
The OECD work on financial consumer protection is written for a broader public than UHNW families, yet the principles travel: disclosure that can be used, conflicts that are managed rather than narrated, and products whose costs are intelligible. A slide that says “open architecture” without a search log, a look-through fee file, and a right to hold a competitor is advertising.
Fee-only advice is the remuneration design that makes the process believable. When the family pays the adviser, the product does not have to. That does not make every fee-only firm competent, and it does not make every private bank dishonest. It does remove a systematic reason to prefer the house factory. Vellum is independent and fee-only: the family pays for advice, not for a product shelf. Custody, execution, and advice can also be unbundled. Reporting should show look-through exposures, not only share-class names. For how that operating model differs from a traditional private-bank package, see how an independent office compares with traditional wealth-management firms. The services map shows how advice, structuring, and reporting sit apart from a captive catalogue.
Five tests for open architecture versus in-house funds
Beauty parades reward vocabulary. A mandate review should reward documents. The five tests below can be run on an existing relationship or on a request for proposal. They require someone, inside the family office or as an independent second pair of eyes, to sit with the files.
Look-through costs, affiliated share, and competitor funds
First, rebuild TER plus any retrocession, rebate, or platform payment for every line, including cash. Ask for the same look-through on private-market vehicles, where placement fees and feeder expenses hide easily. If the bank cannot produce the file, the architecture is not open enough to supervise.
Second, measure the percentage of the book in affiliated funds, affiliated SMAs, and affiliated feeders. There is no universal “right” number. A concentrated house allocation after a documented search is different from a default of most liquid risk sitting in group products that nobody has put on watch. Ask for the figure at market value, by sleeve.
Third, test the ability to hold a competitor fund that is operationally ordinary: a UCITS the family already owns elsewhere, or a well-known SMA. If the answer is “yes, but the extra due diligence will take two committee cycles and a higher custody tariff,” the list is guided in practice even if the brochure says open. Record the tariff. Convenience pricing is sometimes the real gate.
Private markets, cash, voting, and side letters
Fourth, ask whether private-market access requires a captive feeder. Many banks add value by aggregating tickets, running KYC, and negotiating a house side letter. That can be worth paying for. It is not the same as being unable to underwrite a primary fund or a secondaries ticket on the family’s own paper. Independence includes the right to decline the feeder when the extra layer is not earning its keep.
Fifth, inspect cash and money-market parking. Liquidity is where house products often sit without debate. Sweep vehicles, internal money-market funds, and group term deposits can be sensible. They can also be a quiet yield share. The family should know the alternative rate, the credit exposure, and whether cash can sit at a third-party treasury fund without drama.
Two further checks sit beside those five. Voting and ESG overlays should be executable on external funds and SMAs, not only on house vehicles with a house voting policy. Side letters on private funds (most-favoured-nation clauses, co-invest rights, reporting, key-person, and excuse rights) should be negotiable for the family, not only for the bank’s feeder. If overlays and letters exist solely inside the factory, the architecture is closed at the points that matter for control.
Where a multi-family office sits versus a private bank
A private bank remains useful. Custody, lending against a securities portfolio, foreign-exchange, and a local desk are genuine utilities. The conflict appears when the same institution is also the manufacturer, the distributor, and the adviser, and when the family has no separate agent whose only client is the family.
An independent multi-family office sits on the family’s side of that table. It can still use a private bank as custodian or lender. It should not need the bank’s factory in order to complete a portfolio. Manager search, replacement, look-through reporting, and the investment policy statement belong to the office. Product manufacturing does not. That is the 2026 meaning of independence: not a refusal to hold a well-run house fund when it wins a search, but a refusal to let the search be won in advance by the factory. Families who run this comparison well treat it as a recurring control, like a valuation policy, not a one-off beauty-parade question.
Conclusion
Open architecture earned its reputation when private banks opened their shelves to third-party funds. In 2026 the phrase often describes a guided list sitting in front of an affiliated factory. Independence is the ability to select third-party funds, SMAs, private markets, and cash without that factory, and to pay for advice in a way that does not secretly prefer in-house vehicles. Families can test that claim with look-through fees, affiliated weights, competitor access, feeder optionality, and cash parking, then with voting overlays and side letters. The process lives in the investment policy statement, the search file, and the replacement clock, not on a pitchbook slide.
Discretion. Stability. Prosperity.
Team Vellum
A team of passionate professionals who combine their expertise to bring knowledge through Vellum Finance & Patrimoine blog articles. Each member writes about their own field of expertise, cross referencing with our colleagues own fields to ensure the highest quality of information possible in all our content.
