The architecture of private wealth management
The first time I walked into a private bank I was sixteen years old. My father had set up an account for my siblings and me so we could start learning how to manage wealth. The building was old and looked like one of those English ‘establishment’ clubs; dark wood floors, heavy mouldings, the recognisable codes of money. The banker was welcoming and spent time showing us spreadsheets that were, I remember, entirely lacking in depth, colour or anything fun.
Since then I have passed through four different private banks. The display of wealth; the impressive addresses, the recognisable art on the walls, the weight of the stationery, always seemed slightly excessive to me. Who was paying for the structure, the decoration, the enormous salaries? I would meet my banker once a year and I never particularly valued the prestigious events I was occasionally invited to but kept wondering: where did all the money come from?
When I thought of a bank, I pictured somewhere safe to put money. A vault with good carpets and a reassuring human in a suit looking after things. The private bank, reserved for clients with significant assets, typically above one or two million euros, presents itself as something more refined: a trusted advisor, a partner in the long-term stewardship of family wealth. The language is specifically chosen. Words like security, preservation, growth, protection. As if we were talking about a house or a painting rather than a liquid asset.
Through all my visits and meetings I never thought to ask about what I was paying the bank and the banker for this “security” and “service”. So I have done it now and here are my findings.
How the Money Flows
European private banking manages approximately €28 trillion in assets (the equivalent of the whole of European GDP), the US private banking manages roughly 37 trillion. McKinsey’s annual survey of the industry puts the average revenue margin at 73 basis points of assets under management, meaning the bank earns roughly €73 for every €10,000 of client money managed, every year, regardless of whether that money grows or shrinks. On a €2 million portfolio, that is approximately €14,600 extracted annually before a single transaction is made or a piece of advice given.
That fee, the management fee, is the first and most important revenue stream. It runs in good markets and bad. It is the reason the primary strategic imperative of every private bank is the same: grow assets under management (AUM). Everything else, investment performance, client service, advice quality, exists in relation to that imperative. That is, the imperative to accumulate assets inside the bank.
The second stream is transaction-based: the spread between the price at which the bank buys an asset and the price at which it sells it to the client. The client sees the net price. The bank keeps the margin. Fees embedded in transaction costs are less psychologically visible than explicit charges, a phenomenon behavioural economists call salience bias, which makes them a particularly comfortable source of income for the institution.
The third is product placement. When a bank recommends a fund, the fund manager often pays the bank a distribution fee for that recommendation. The bank is, in effect, being paid by the product it is suggesting to you. MiFID II (Markets in Financial Instruments Directive), the European regulation that came into force in 2018, tightened disclosure requirements around these so-called retrocessions. Belgium’s financial regulator, the FSMA (Financial Services and Market Authority), fined Degroof Petercam €800,000 in 2020 for inadequate compliance with MiFID II; for giving advice to clients that had not been properly verified against their risk profile. It is a rare moment of institutional transparency in a landscape where the inner mechanics are rarely exposed to public light.
These three streams often operate simultaneously on the same investment. On a single structured product, a client may be paying a management fee for portfolio oversight, a transaction spread on the purchase, and an implicit placement fee built into the product’s structure. Assembled from individually small components, it is routinely underestimated.
Writing this article, I asked my bank for the full extent of what I pay them per year, in numbers, not percentages. They gave me a specific figure for 2025, but it covered only the management fee. Not what the bank earns on the products I bought or sold, nor the fees linked to the bank’s own funds I am invested in. The three streams, it turns out, are not easy to see even for the banker.
I also contacted friends at three Belgian private banks and asked for their tariffs, they sent me their public brochures. Puilaetco, a Quintet Private Bank, charges between 0.62% and 1.34% annually on discretionary mandates, with custody and transaction fees on top, and a minimum annual fee of €3,000 regardless of performance. Bank Degroof Petercam, now under IndoSuez Wealth Management, charges minimum annual fees of up to €12,705 on certain strategies, meaning a client at the lower end of the asset threshold pays an effective rate far above the headline percentage. Its deposit interest rate is set at ECB rate minus 1% for balances above €100,000, clients holding cash pay the bank for the privilege. Banque Transatlantique Belgium offers something rarer: a performance fee option, a reduced base rate plus 5% of any portfolio gains, which it markets explicitly as “an alignment of interests between you and the bank.” It is an exception. That it needs to be marketed as exceptional tells you something about the rule.
How the Banker Gets Paid
I have explained how the institution gets paid but we can also look at the banker level. The private banker, the person who calls when markets are volatile, who remembers your children's names, who sits across the table at the annual review, who hands you papers with small post it notes on the pages you need to sign, how are they paid?
The base salary is comfortable but not spectacular. The sleight of hand is in the bonus. And in private banking, the bonus is overwhelmingly tied to two things: net new money; how much new client assets the banker brought in and AUM growth; how much the total book grew. The actual performance of those assets, real returns for real clients, is typically a secondary input in the formula, if it appears at all.
The economists Jensen and Meckling identified this problem clearly in their foundational work on principal-agent theory: when the incentives of the agent are not aligned with the outcomes desired by the principal, the agent will rationally optimise for what they are measured on. A banker measured on AUM growth will grow AUM. A banker measured on client returns will optimise returns. These are not the same objective.
A respondent to my research told me about a conversation she had with her banker. She asked him to remove all petrol and defence investments from her portfolio. His answer: “But madam, how then are you going to make money?” She was well aware of the possible financial outcomes and was not looking to grow her wealth but that goal was simply not legible to the institution. It had no category for it.
The banker’s most valuable asset is not their product knowledge. It is their client book, the accumulated trust deposited over years of dinners, calls, crises navigated. Pierre Bourdieu would recognise it: social capital converted into economic capital. When a senior private banker moves from one institution to another, the clients, in many cases, follow. The negotiation that follows, retention packages, non-solicitation clauses, guaranteed bonuses on lateral hires, is the institution’s attempt to commodify what is fundamentally a human relationship. The warmth is as real as the commercial structure underneath it.
The question is whether that trust can hold when the client’s objectives move outside the bank’s incentive structure, toward redistribution, toward risk the bank does not support, toward simply keeping wealth in a safe place without growth.
A banker measured on AUM growth will grow AUM. A banker measured on client returns will optimise returns. These are not the same objective.
The Language That Followed
There has been a vocabulary shift across forty years. In the 1970s and 1980s, private banking spoke of service; what the bank did for you. Through the 1990s, advice; what the bank thought for you. By the 2000s, partnership; what the bank did with you. Each word tracked the fee model: service was transactional (commission), advice was professional (retainer), partnership is relational (AUM).
The shift to fee-on-AUM came with an elegant argument: the fee aligns interests. When the bank earns a percentage of what it manages, it benefits when the portfolio grows. The client and the bank are, supposedly, in the same boat. Again only if the goal is accumulation and concentration.
The argument is partially true. A commission-based banker who churns a portfolio; generating unnecessary transactions to extract fees is obviously worse for a client than a fee-based banker whose income is stable. But the argument contains a quiet omission. Aligning the bank’s revenue with assets under management is not the same as aligning it with investment performance; whichever the performance the client requests. The fee runs whether the portfolio rises or falls. In a flat or declining market, the bank continues to earn. They are not in the same boat but rather adjacent boats, travelling in the same general direction.
Nassim Nicholas Taleb’s framework of skin in the game names this precisely: the advisor who bears no downside is structurally different from the client who does. The banker experiencing a market downturn faces a professional challenge, the client however faces a major change. A true alignment of interests would require the advisor to suffer when the client suffers, which the AUM model, in its standard form, does not deliver. There is also no incentive for distribution or investment outside the bank. If the client removes money from the bank the banker then also loses AUM so there is no incentive to advise on a wider holding of assets.
In 2022, European private banking profits hit a record €22 billion. Clients with balanced risk profiles earned zero or negative returns between 2018 and 2022. Profits rose because rising interest rates lifted the margin banks earned on client cash deposits. The bank’s record year and the client’s flat returns coexisted, without contradiction, in the same annual report. The partnership argument, examined under those conditions, begins to look rather one-sided.
The Gravity Well
The most interesting question is not whether the system is dishonest. In most cases it is not. The most interesting question is what the system produces, in the behaviour of clients who may not be aware they are inside it.
A gravity well, in physics, is a distortion in space caused by the presence of a massive object. Objects near the well do not feel themselves being pulled, they experience their own trajectory as natural. Only from outside does the curvature become visible. The incentive architecture of private banking creates something similar for clients.
Clients accumulate. Not necessarily because accumulation is the right strategy for their situation, but because the entire architecture of the relationship, the fee model, the banker’s incentives, the annual review format, the product range, is oriented toward accumulation as the default. The alternative, spending the money, giving it away, withdrawing it, is rarely the subject of an unsolicited call.
Clients optimise for tax. This advice is not improper, in many cases it is genuinely valuable. But every structure created, every holding company established, every offshore vehicle maintained represents assets that are harder to move and harder to transfer to a competitor. Tax optimisation and asset retention are not the same objective. They tend, however, to produce the same outcome. For example, a client who has restructured their patrimony through a Liechtenstein foundation; a cross-border legal structure offering tax and succession advantages, is less likely to move banks. And, moreover, if the bank acts as trustee of that foundation, or administrator of the holding company beneath it, it becomes a party to the legal architecture of the client's wealth. Replacing a party to a legal structure is a different order of problem from switching a bank account. The complexity becomes the lock.
Clients keep assets in the bank. The Lombard loan is a product offered by the bank that lets a client borrow money against their portfolio rather than sell assets. The client gets liquidity without triggering a tax event. The bank gets lending income. The assets must remain in the bank as collateral for the loan. It solves a real client need in a way that grows the bank's balance sheet rather than reducing it.
These pulls do not require bad intent. They are structural. The banker offering perfectly honest advice within the standard toolkit is still operating within a system that rewards accumulation, complexity and retention. The individual may be excellent. The field of force is what it is.
Most of the people I have spoken to service provider or wealth holders build their wealth strategy along the line of: structuring the portfolio to align with risk levels (talking only of financial risk and never of ecological or social risk), tax optimisation and talking early about how they want the wealth to move after their death. Again mostly for tax avoidance issues. After all that is set up, they then give liberty to the clients personal desires around how and where they want the wealth to go. I believe this process should be flipped.
What You Cannot See and What to Ask
There is one dimension of the system that is invisible almost by design. Clients of the same private bank do not see each other. They do not know what fees others are paying, what products others have been recommended, what returns other portfolios have generated. The relationship feels personal. Inside the bank, it largely is not, relationship managers work from the same product range, the same house view, the same fee grid. What presents itself as a personalised judgment is, in large part, a standardised output delivered through a personal relationship.
MiFID II moved toward individual transparency: the client can now see what they pay. It did nothing toward comparative transparency. The bank works with aggregate data and individual relationships. The client works with one relationship and no data.
Clients of the same private bank do not see each other. They do not know what fees others are paying, what products others have been recommended, what returns other portfolios have generated.
Personnally I hold a part of my wealth in a private bank as it offers something real, for me it still means security. A skilled, honest, client-oriented banker is worth the fee, worth the relationship. The question is simply: how do I know if mine is delivering it? Not from the annual review, prepared by the banker themself. Not from a comparison with my peers, which I cannot make. The structure; the fees, the incentives, the language, the complexity, is designed to make the question feel unnecessary. Asking it anyway is a reasonable place to start. My colleague Cassie has outlined some questions to start the conversation in another article in this issue.