The "13O vs 13U" Decision Framework: Why Mid-Tier Family Offices Are Defaulting to 13O

Originally published: 2024-04 | Last verified: 2026-09-13 Scheme conditions are checked against MAS Circular FDD Cir 05/2026 (31 July 2026) — the IP, AUM and local spending conditions in Annexes 6B and 7B, the vehicle rules in Annexes 6 and 7, and the common designated-investment list in Annex 3 — linked inline. Earlier versions of this piece said the non-family IP requirement applied only to 13U, that 13U covered a broader set of investments, and used the pre-August-2026 local business spending tiers. All three were wrong; they are corrected here and the cost model has been rebuilt on the corrected rules. Cost figures are my own practitioner estimates and are labelled as such. Tax incentive frameworks evolve; please confirm against MAS notices and your licensed advisor for current parameters.

For most of the period from 2019 through early 2023, the 13O vs 13U decision was treated as a wealth-band sorting problem with a fairly clean table of cutoffs. If your family office was sub-S$50M AUM, you took 13O. If you were S$50M+ and could absorb the additional substance commitments, the conventional advice was 13U — usually pitched as the scheme with the broader tax coverage. Most of the practitioner work was confirming the wealth-band cutoff and structuring around it. The broader-coverage pitch does not survive a reading of the current rules: MAS’s July 2026 circular applies one annex of Designated Investments and Specified Income across 13D, 13O, 13OA and 13U.

The MAS condition changes since 2023 broke the clean version of that table: for applications from 5 July 2023, a non-family IP requirement and the removal of the grace periods for hiring the second IP and building up to the minimum AUM; AUM measured against Designated Investments from 1 January 2025; and recalibrated single family office (SFO) conditions for awards approved from 1 August 2026. The decision is no longer a simple wealth-band sort. It is a four-factor decision: wealth size, read together with vehicle architecture, is the first input, and three other variables — who fills the third IP seat, the total cost of compliance, and political and operational optics — carry meaningful weight.

This piece lays out the decision framework I’d use today for a mid-tier family office evaluating the 13O vs 13U choice. The mid-tier here is the USD 30–80M wealth band — the cohort that the MAS substance updates have selected for, and the cohort where the framework actually matters because, for most of the band, both options are technically available — only the bottom end, below roughly USD 37M (S$50M) of Designated Investments, sits under the 13U floor. It’s the same cohort I described as the marginal applicant driving Singapore’s family office boom; this piece is the structuring decision that cohort actually faces.

The two regimes, as of the July 2026 circular

Briefly, for the practitioner-level reader:

Section 13O (and its limited-partnership twin, 13OA). Minimum S$20M of AUM in Designated Investments at the point of application and at the end of each basis period; for awards approved from 1 August 2026, the fund’s investments in the family’s own operating businesses don’t count towards that minimum. At least two Investment Professionals (IPs), of whom at least one must not be a family member, each Singapore tax-resident, each earning more than S$3,500 a month and engaging more than half their time in the qualifying activity. For awards approved from 1 August 2026, a 13O/13OA SFO fund may apply with a single qualifying IP, who may be a family member, provided the second IP — and with it the non-family requirement — is in place by the end of the basis period of the first year of assessment, failing which the award is revoked from its commencement date. Tiered Local Business Spending (LBS) — and this is the part most published summaries still have wrong, because MAS rewrote the tiers on 31 July 2026: S$200K a year below S$250M AUM, S$500K from S$250M to S$2B, S$1M at S$2B and above. They apply to SFO awards approved from 1 August 2026 and, from the year of assessment whose basis period ends on or after 1 August 2026, to existing awards on applications that reached MAS on or after 18 April 2022; awards on applications made before that date keep a flat S$200K spending condition. The old thresholds were S$50M and S$100M, which is a very different regime for a mid-tier holder — and at the time of writing, MAS’s own family office scheme page still displays them. The fund must be a Singapore-incorporated, Singapore-tax-resident company (13O) or a Singapore-registered limited partnership (13OA), and must use a Singapore-based fund administrator. The exemption applies to specified income from designated investments.

Section 13U. Minimum S$50M of AUM in Designated Investments, again with the family’s operating-business investments excluded from the minimum for new awards. At least three IPs, of which at least one must be a non-family member — the same non-family condition as 13O, applied to a larger team — and for awards approved from 1 August 2026 you may apply with two qualifying IPs, family members or otherwise, and add the third by the end of the first year-of-assessment basis period. Same tax-residency and salary conditions for the IPs. Same LBS tiers. The designated-investment list is not broader: MAS applies one common Designated Investments and Specified Income annex across 13D, 13O, 13OA and 13U. The real structural differences are the vehicle and the investor rules. 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. There is no requirement that a 13U fund be Singapore-incorporated or Singapore-resident; a Singapore-based fund administrator is required only where the fund entity is a Singapore-incorporated, Singapore-resident company. And 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 — is set out in the circular for 13O and 13OA, not for 13U.

What matters about this framing is that 13U is not simply “bigger 13O” — but the difference is not where it is usually said to be. I had this wrong in an earlier version of this piece on two counts: the non-family IP requirement is not a 13U-only constraint, it binds at 13O too; and 13U does not exempt a wider set of investments. The genuine 13O-to-13U delta on headcount is one additional IP, who may be a family member. The substantive differences are the AUM floor, the vehicle and structure flexibility, the investor rules, and that one extra seat.

The four-factor framework

For a mid-tier holder evaluating the choice in 2026, the relevant decision inputs are:

1. Wealth band and vehicle architecture — Can you put S$50M of qualifying investments into the fund, and do your holdings actually need what only 13U allows: a non-Singapore vehicle, a master-feeder, or SPVs under one award?
2. IP availability and the third seat — Can you fill a third qualifying IP, and is there a family member who genuinely qualifies for it?
3. Total cost of compliance — What do the extra seat and any extra entities cost, given that there is no incremental tax exemption to set against them?
4. Political and operational optics — Does running a 13U structure rather than a 13O structure send signals — to home-country tax authorities, to international counterparties, to your own family governance — that you actually want sent?

I’ll walk each of these in turn.

Factor 1: Wealth band and vehicle architecture

The first factor starts with the floor. For a wealth holder at USD 30M, 13U is technically off the table — they fall below the S$50M minimum of AUM in Designated Investments. For a wealth holder at USD 60–80M, both options are available on paper, but 13U requires committing at least S$50M of qualifying investments into the fund, and for awards from August 2026 the fund’s investments in the family’s own operating businesses don’t count towards that minimum. A family whose wealth is concentrated in the operating company can be well above S$50M overall and still short of the 13U floor.

The architecture question is sharper than it sounds, and it is where the conventional framing goes wrong. It is not about which assets are covered: the Designated Investments list is the same under both schemes, so a portfolio of listed equities, bonds and fund interests earns the same exemption in a 13O vehicle as in a 13U one. What differs is the container. 13O needs a single Singapore-incorporated, Singapore-resident company (or a Singapore limited partnership under 13OA). 13U will take a vehicle constituted in any form, and a master-feeder or SPV structure under one award. That matters for families that already hold assets through an offshore fund they don’t want to redomicile, that want to ring-fence holdings in separate SPVs, or that would have a Singapore-resident corporate investor caught by the 13O 30/50 rule.

The practitioner question I’d ask: map how the assets are held today and how they will be held over the next three years, entity by entity. If one Singapore company can hold all of it, 13U’s flexibility is theoretical rather than economic. The mid-tier holders who genuinely use 13U’s architecture are those with existing offshore vehicles or multi-SPV holdings — and most mid-tier holders don’t have them.

Factor 2: IP availability and the third seat

This is the factor most often mis-stated, and it is where the earlier version of this piece went wrong.

The 13O regime requires two IPs, at least one of them a non-family professional. So the external hire starts at 13O — a family cannot staff the fund purely from the family.

The 13U regime requires three IPs, still with a minimum of one non-family. On the letter of the rules the third seat can be filled by a family member, provided they genuinely qualify: Singapore tax-resident, paid more than S$3,500 a month, and working as a portfolio manager, research analyst or trader. In practice most family principals prefer not to count themselves as an IP for governance reasons, so the third seat often gets filled externally anyway — but that is a governance preference, not a MAS requirement, and it is worth knowing which is which before pricing it. In the current SG family office IP labour market, that distinction is meaningful. Senior IP-class hires — typically, on my estimate, professionals with USD 150K-300K all-in compensation expectations and Asia-region buy-side experience — are in genuinely tight supply, and in my read the families that have closed those hires have usually had to pay up to do so.

Timing has changed as well. For awards approved from 1 August 2026, a 13U applicant can apply with two qualifying IPs and a 13O applicant with one, completing the roster — including the non-family IP — by the end of the basis period of the award’s first year of assessment. How long that is depends on the financial year-end: in MAS’s own example, an award starting 1 August 2026 with a 31 December year-end has until 31 December 2026, about five months, and missing it revokes the award from its start. So the search for the non-family IP should begin before the application goes in, and the award’s start date should be planned against the financial year-end.

The cost of the third seat is therefore a range, not a point. On my estimates it runs from a little over S$42K a year — the statutory floor of more than S$3,500 a month — where a qualifying family member takes the seat and isn’t already on the payroll at qualifying terms, to ~S$150K-250K for an external hire at research-analyst level, and more for a senior one. That is the IP component of the 13U add-on in Factor 3, in the same currency. The tightness of the external search is itself part of the broader family office hiring squeeze — which is why the talent constraint weighs on families with no qualifying family member for the third seat, and much less on families that have one.

Factor 3: Total cost of compliance

Run the numbers honestly — and start with a correction to how earlier versions of this piece ran them. 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. The LBS floor is therefore a minimum that Singapore IP salaries, fund-admin and legal fees count towards, not a separate cost to add on top of them. The earlier model added it on top, which double-counted.

For a mid-tier holder at S$60M AUM, my estimate of the 13O annual run-cost — IP compensation (~S$300K-500K), fund admin (~S$80K-150K), legal and compliance (~S$80K-120K), and corporate services (~S$30K-60K) — lands somewhere in the S$0.5M-0.8M per year range, before any office space or auxiliary headcount. The LBS floor at this AUM is S$200K under the 31 July 2026 tiers (it was S$500K under the old ones), and a structure paying Singapore-based IPs and service providers at those levels clears it without extra spend.

Moving the same structure to 13U adds the third IP (from a little over S$42K with a qualifying family member to ~S$150K-250K all-in for an external hire) and — if you use the multi-entity architecture that is the reason to choose 13U — incremental fund admin, audit and legal work across the extra entities (~S$30K-60K). Total 13U add-on at S$60M AUM, on my estimates: roughly S$70K-100K a year with a family third IP, S$180K-310K with an external one. The LBS floor is the same under both schemes at this AUM, so it adds nothing to the difference.

Earlier versions of this piece set that add-on against an “incremental tax saving” from 13U’s supposedly broader coverage. There is no such saving: the same designated investments produce the same exempt specified income under either scheme. The 13U add-on buys architecture — the vehicle and structure flexibility in Factor 1 — not a bigger exemption.

The practitioner-level summary: 13O is the default wherever a single Singapore-resident company or partnership can hold what you own. 13U is justified by architecture, and whether that architecture costs you roughly S$70K or S$300K a year depends mostly on who fills the third seat.

Factor 4: Political and operational optics

The fourth factor is the one that doesn’t show up in the cost spreadsheet but matters more than most advisors price for. There are at least three optics layers worth thinking through.

Home-country tax authority signaling. For wealth holders coming out of jurisdictions with active offshore-structure scrutiny — China, India, increasingly Brazil — the visibility of the Singapore structure to home-country authorities is a real consideration. But the disclosure rules don’t differ between the schemes, the local spending tier is the same at mid-tier AUM, and the staffing difference is one IP, so the discretion argument for 13O over 13U is weaker than it is often made to sound. What can be more visible is the 13U architecture itself: a master-feeder or multi-SPV structure spread across several jurisdictions presents more entities to any cross-border information exchange than a single Singapore company. For holders sensitive to that, “13O is the simpler choice” is a rationally weighted consideration.

International counterparty optics. Some counterparties — private banks, fund admins, prime brokers — in my read differentiate between 13O and 13U holders in their internal client tier classification; beyond the AUM floor, the rules give them little objective basis to do so. Where that differentiation has tangible service-quality implications for a mid-tier holder, the 13U add-on can be worth paying even without an architectural need. The mid-tier holders who optimize for service quality over cost (typically the 60–80M end of the band, with a long-term wealth horizon) sometimes choose 13U for this reason alone.

Family governance optics. Internally, the choice signals something about how the family thinks about its wealth structure. A 13O setup with one family member and one external professional as its two IPs reads as a family-managed vehicle with light external infrastructure. A 13U setup with two or three external professional IPs reads as a professionally managed family office with family governance oversight. Some families want one signal; some want the other. Neither is wrong, but the choice shapes the internal culture of the office for years.

The framework, compressed

For a mid-tier (USD 30-80M) holder evaluating the choice today, my decision framework looks like this:

The “depends on architecture” zone in the middle band is where the real practitioner work happens. For holders in that zone, the right process is to map the actual holding structure entity by entity, price the third seat honestly — family or external — against current market compensation, and run the political-optics overlay against the home-country sensitivity profile. There is no shortcut.

What might shift this in the next 18 months

Two things to watch that could move the framework, and one development that has landed without moving it.

First, the IP labour market. If the IP supply in SG genuinely tightens — say, due to a regulatory tightening that limits which professionals qualify as IPs, or due to the existing supply being absorbed by the next wave of family office formations — the 13U add-on gets more expensive for families that have to fill the third seat externally, and where their architectural case for 13U is marginal, the 13O default would harden. Families with a qualifying family member for the third seat are largely insulated.

Second, the political environment around home-country wealth-structure disclosure. If the Common Reporting Standard tightens further on family office structures, or if specific home countries (notably India, China) push for more granular disclosure on Asia-booked wealth structures, every additional entity in a 13U structure becomes one more thing to explain. That would favour the single-vehicle 13O set-up at the mid-tier band even where the architecture would otherwise have favoured 13U.

Third, the class exemption framework — which has now landed, and which doesn’t move this comparison. MAS’s revised SFO framework took effect on 15 June 2026, replacing case-by-case exemption with a straight-through class exemption: qualifying SFOs notify MAS, maintain an account with a MAS-licensed bank, and file a straightforward annual return, with existing SFOs given until 15 June 2027 to comply. Those requirements apply identically whichever tax scheme you sit under — and to SFOs that sit under neither — so they don’t change the 13O-versus-13U difference. What they change is the compliance floor for the family office itself.

The 13O default is not permanent. It is the equilibrium of the current rules, labour market and political configuration. Practitioners who treat it as a permanent feature will be slow to read the signals if any of the above shift. Practitioners who treat it as a contingent equilibrium will be positioned to advise the next cycle of holders accurately.