The Quiet Cost of "Premium Tier" 13U: It Buys Structure, Not a Bigger Tax Saving

Originally published: 2024-12 | Last verified: 2026-09-13 Scheme conditions are checked against MAS Circular FDD Cir 05/2026 (31 July 2026) and the MAS family office scheme page, linked inline. Earlier versions of this piece said 13U covered a broader set of investments than 13O, that the non-family IP requirement applied only to 13U, and used the pre-August-2026 local spending tiers. All three were wrong, and the breakeven analysis built on them has been replaced. Cost figures are my own practitioner estimates and are labelled as such. Tax incentive specifics are subject to MAS revision; please confirm against MAS notices and your licensed advisor before any structuring decision.

The way Section 13U is usually presented to a wealth holder considering a Singapore family office structure is something like: “This is the premium-tier scheme. Higher AUM threshold, more substance, broader coverage on the tax-exempt income side, and it’s the right answer once you’re past S$50M.” The framing is consistent across the family-office-setup advisory ecosystem — fund administrators, law firms, the tier of consultancies that build their fee model around 13U applications. It is also, in my read, wrong in one respect and incomplete in another.

The wrong part is the tax coverage. MAS publishes one annex of Designated Investments and Specified Income that applies across 13D, 13O, 13OA and 13U. The same portfolio earns the same exemption under either scheme; there is no incremental tax saving from moving up to 13U. The incomplete part is what 13U genuinely does offer — vehicle and structure flexibility — and what it costs at the mid-tier. This is the cost detail underneath the broader 13O vs 13U decision framework, and it’s a large part of why the mid-tier Singapore family office boom skews so heavily toward 13O. This piece walks through what the premium actually buys, the cost layers, and when paying them makes sense.

The headline pitch and what it leaves out

The standard 13U pitch points to three differentiators over 13O: (a) broader coverage of designated investment categories on the tax-exempt income side, (b) more permissive treatment of certain income streams that 13O excludes, and (c) signaling value with counterparties, banks, and home-country authorities that “we run a serious operation.”

The first two are not in the rules; both schemes draw on the same Designated Investments and Specified Income lists. The third is a matter of perception. What the pitch leaves out is the differentiator that is real, and the costs of using it:

1. What 13U genuinely adds: vehicle and structure flexibility.
2. The third IP seat, and who can fill it.
3. The tiered Local Business Spending floor — which, at the mid-tier, is the same under both schemes.
4. The compliance and infrastructure cost that scales with the number of entities.

I’ll take each in turn.

What 13U actually adds: structure, not coverage

13O requires the fund to be a Singapore-incorporated company that is tax-resident in Singapore — or, under 13OA, a Singapore-registered limited partnership — using a Singapore-based fund administrator. 13U accepts a standalone fund vehicle constituted in any form, or a master-feeder, master-feeder-SPV or master fund-SPV structure under one consolidated application, with the AUM and local spending conditions tested at structure level as though it were a single fund. The circular also sets out the “30/50” rule — a financial penalty on a Singapore-resident non-individual investor that owns more than 30% of a fund with fewer than ten investors (50% with ten or more), unless it is a qualifying investor — for 13O and 13OA, not for 13U.

That flexibility has real value for a family that already holds assets through an offshore fund it doesn’t want to redomicile, that wants holdings ring-fenced in separate SPVs, or that has a Singapore-resident corporate investor the 30/50 rule would catch. For a family whose assets can sit in one Singapore company, it has little value at all. The premium is worth paying exactly to the extent you would use the architecture.

Cost layer 1: The third IP seat

Section 13O requires two Investment Professionals (IPs), each Singapore tax-resident, each earning more than S$3,500 a month, each performing genuine investment functions — and at least one of the two must not be a family member. Section 13U requires three IPs, again with at least one non-family member. The headline framing is that you go from a 2-IP structure to a 3-IP structure, an apparent +50% headcount lift.

An earlier version of this piece argued the lift was usually larger, because it treated the non-family requirement as 13U-only and concluded that 13U families needed two or three external hires. That was wrong. The external hire is needed under both schemes, and the third 13U seat may be filled by a family member who genuinely qualifies. So the real 13O-to-13U staffing delta is one seat, and its cost depends on who takes it:

• A qualifying family member. If they aren’t already on the payroll at qualifying terms, the floor is more than S$3,500 a month — a little over S$42K a year.
• An external hire. The talent market for senior investment professionals in Singapore — particularly those with the buy-side, Asia-region, multi-asset experience that would qualify them as a credible IP for a family office — is genuinely tight. On my estimates, an external third IP at research-analyst level runs S$150K-250K all-in, and a senior multi-asset hire (USD 150K-300K depending on seniority and asset-class specialization, plus typically 3-9 months of search) comes in higher.

Most family principals prefer not to count themselves as IPs for governance reasons, so many families do end up hiring the third seat externally. That is a governance choice rather than a MAS requirement, which is worth knowing before accepting it as a cost of 13U.

Timing also changed for awards approved from 1 August 2026: a 13U SFO fund may apply with two qualifying IPs, family members or otherwise, and must have the third — with at least one non-family IP among the three — by the end of the basis period of its first year of assessment, or the award is revoked from its commencement date. That moves the third hire after approval, but not far after: in MAS’s own example, an award starting 1 August 2026 with a 31 December year-end has about five months.

We’ve had searches sit open for nine months because the family wanted a 13U-credible IP and the candidate pool that fits the bill is small enough that the right person is always already employed somewhere they don’t want to leave for less than premium compensation.— Singapore-based wealth recruiter, family office practice

Read “13U-credible” as a seniority bar, not a separate standard: MAS applies one IP definition — portfolio managers, research analysts and traders who are Singapore tax-resident and earn more than S$3,500 a month — across 13O, 13OA and 13U.

Cost layer 2: All-in compensation and substance footprint

The IP compensation cost is only one part of the substance footprint. The rest scales with headcount and, more importantly, with the number of entities you run:

• Desk space and support for the extra IP.
• Fund administration, audit and legal work for each additional entity if you use a master-feeder or SPV structure — which is the reason to choose 13U in the first place.
• Governance and reporting to family members, which grows with the complexity of the structure rather than with the scheme label.

For a single-vehicle 13U with one extra IP, the surrounding cost is modest. For a multi-entity structure, my estimate of the incremental administration, audit and legal cost is S$30K-60K a year at mid-tier AUM. An earlier version of this piece put the all-in 13U-over-13O increment at S$300K-600K a year; that figure assumed a larger external team and a higher local spending floor than the rules require, and I no longer stand behind it.

Cost layer 3: Tiered Local Business Spending

The Local Business Spending (LBS) floor for single family office funds is tiered, and MAS rewrote the tiers in its 31 July 2026 circular: S$200K a year below S$250M of AUM, S$500K from S$250M to S$2B, and S$1M at S$2B and above. The same table applies under 13O and 13U, and it is not limited to new awards. Awards approved from 1 August 2026 carry it from the start. Under the circular’s tables for existing awards, awards whose applications reached MAS on or after 18 April 2022 move onto the same tiers from the year of assessment whose basis period ends on or after 1 August 2026; RSM and Baker McKenzie read it the same way. Awards on applications made before 18 April 2022 keep their earlier flat S$200K spending condition. The previous tiers had thresholds at S$50M and S$100M — a S$500K floor from S$50M and S$1M from S$100M — and at the time of writing MAS’s family office scheme page still displays them.

That changes this layer completely. An earlier version of this piece treated a S$500K LBS floor as a 13U burden in the S$50-100M band. Under the current tiers the floor at that band is S$200K under both schemes, so it adds nothing to the 13U premium. It is also not additional spend for most structures: MAS defines local business spending as operating expenses “including but not limited to remuneration, fund management fees and other operating costs” paid to contracting parties in Singapore. A family office paying two or three Singapore-based IPs and a Singapore fund administrator clears S$200K from costs it already carries. The pressure to over-purchase Singapore services to hit the floor, which I described previously, largely disappears below S$250M.

Cost layer 4: Compounding compliance and infrastructure

The fourth layer is the one that compounds over the holding period, and it tracks the architecture rather than the scheme label. A multi-entity 13U structure has more entities to administer, audit and update, and structural changes — capital top-ups, a new SPV, family-member changes to the IP roster — each carry an annuity cost that is small per year but compounds over the typical 10-15 year structure life. A standalone 13U vehicle doesn’t carry much more of this than a 13O vehicle does.

The infrastructure cost also tends to drift upward once a structure has several entities and a larger team: tech stack, performance reporting, regulatory engagement, family communications. On my rough estimate, not a measured figure, the all-in cost of such a structure five years in is 15-25% higher than the pro-forma cost projected at setup.

The breakeven math, honestly run

Pull the layers together for a S$60M AUM family considering 13O versus 13U for the same portfolio.

The add-on runs roughly S$70-100K a year if a qualifying family member takes the third seat, and S$180-310K if it is an external hire. Those totals assume a multi-entity structure; for a single-vehicle 13U the extra-entities row drops out and the add-on is the IP line alone — a little over S$42K to about S$250K.

On the other side of the ledger: nothing. Earlier versions of this piece estimated a S$35-100K incremental tax saving from 13U’s broader coverage and concluded that the premium paid back only 10-30% of its cost. The direction was right, but the arithmetic rested on a false premise: there is no broader coverage, so the incremental tax saving on the same portfolio is zero. The 13U add-on never pays for itself in tax. The one near-tax exception is the 13O 30/50 rule: a Singapore-resident non-individual investor holding more than the limit is penalised under 13O and 13OA, not under 13U. But that is a question of who invests in the fund, not of what the fund earns exempt, and it belongs under (b) below. Otherwise the add-on pays for itself only if you need the architecture.

Where the math does work

Because there is no tax premium to earn back, the 13U math clears on architecture alone. Three things need to be true together:

(a) the fund can hold at least S$50M of Designated Investments — and for awards from August 2026, the fund’s investments in the family’s own operating businesses don’t count towards that minimum; (b) the holdings genuinely need 13U’s flexibility — an offshore vehicle, a master-feeder, SPVs under one award, or investors the 13O 30/50 rule would catch; and (c) the third seat can be filled, ideally by a qualifying family member, at a cost the family is comfortable carrying.

When all three hold, the add-on can be modest. When (b) doesn’t hold, the 13U decision is being driven by signaling and counterparty optics rather than by anything in the rules — a legitimate reason to choose it, but not the reason the standard pitch implies.

What the pitch should say

A more honest 13U vs 13O conversation, for a wealth holder in the S$50-150M band, looks like:

1. “13O is the default if one Singapore company or partnership can hold your assets. The exemption is the same under either scheme.”
2. “13U is available, and the add-on is smaller than it is usually made to look — roughly S$70-100K a year on a multi-entity structure if a qualifying family member takes the third seat, S$180-310K if it’s an external hire, and less on a single vehicle — but nothing on the tax side pays for it.”
3. “If you choose 13U without an architectural need, the reason is signaling, counterparty access, or institutional optics rather than tax. That can be the right reason, but let’s be explicit about it.”
4. “If your holdings sit in, or need, an offshore vehicle, a master-feeder or a set of SPVs, 13U’s architecture is the reason to pay. Map the entities before deciding.”

The reason the pitch usually doesn’t go this way is that the advisory ecosystem has fee structures aligned with 13U’s setup complexity — application fees, ongoing compliance fees, fund admin fees. The economic incentive for the advisor pool is to qualify families into 13U rather than out of it.

For the family principal, the right question to ask any advisor is: “Walk me through what 13U lets my structure do that 13O doesn’t, who fills the third IP seat, and the line-by-line cost of the extra seat and entities at current market rates. If the case involves a bigger tax exemption, show me where it is in MAS’s designated-investment list.” If the answer comes back vague, you have your answer about whether the recommendation is being made on your math or theirs.