HNW family-office software is four products in one: wealth management, tax planning, estate planning, entity administration. The platforms in market — Addepar, Black Diamond, Eton Solutions, Mastery Wealth, the newer entrants — handle features that mass-market wealth-tech does not have to know exists. Their test corpora reflect it.
A platform serving family offices has to model multi-entity household structures, multi-generation trust arrangements, illiquid alternatives, concentrated equity, pre-IPO holdings, multi-jurisdiction tax, and the documentation chain that ties the four pieces back together. Each of those is a primary structural element of the corpus, not a tagged-on attribute of a household record.
What HNW actually means for the platform
The conventional HNW threshold is $5M of investable assets; ultra-HNW is $30M+. The numbers matter less than the structural differences:
| Mass affluent | HNW / UHNW | |
|---|---|---|
| Number of accounts | 3-7 (taxable, IRA, 401(k), HSA, 529) | 20-100 (across spouses, trusts, entities, jurisdictions) |
| Account types | Standardized brokerage products | Mix of brokerage, custodied private investments, direct holdings, partnership interests |
| Entity structure | Joint or individual | Multiple trusts, LLCs, S-corps, family partnerships, foundations |
| Beneficiary structures | Spouse + children | Multi-generation, with conditional / discretionary distributions |
| Tax exposure | Federal + state | Federal + multiple states + international + GST + estate |
| Illiquid positions | Rare | Frequent — private equity, real estate, art, collectibles, private credit |
| Privacy / security profile | Standard PII handling | Higher-touch — physical security, cybersecurity, public-figure considerations |
Each row is a different data shape. The platform has to support all of them, and the synthetic corpus has to exercise all of them.
The entity structure complication
A typical HNW household isn't a household — it's an entity diagram. The structures we model in our HNW corpus:
Common HNW entity structures
- Revocable trust (a.k.a. living trust) — owns most assets during life, becomes irrevocable at death
- Irrevocable life insurance trust (ILIT) — owns life insurance policies, removes death benefit from estate
- Spousal lifetime access trust (SLAT) — funded by one spouse, benefits the other
- Grantor-retained annuity trust (GRAT) — transfers appreciation while retaining annuity stream
- Intentionally defective grantor trust (IDGT) — used for sales of appreciating assets to remove from estate
- Charitable remainder trust (CRT) — provides income to grantor with remainder to charity
- Charitable lead trust (CLT) — pays income to charity for term, remainder to family
- Generation-skipping trust (dynasty trust) — multi-generation, GST-exempt
- Family limited partnership (FLP) — operating vehicle for family business and investment holdings
- Personal residence trust (QPRT) — transfers residence with retained occupancy
- Investment LLCs — pass-through investment vehicles for HNW investors
- S-corps, C-corps — owned via the family structure for active businesses
A real HNW household might have 5-15 of these layered together. A wealth-planning platform that doesn't model the entity diagram explicitly produces wrong tax projections, wrong beneficiary distributions, and wrong post-mortem analyses. The dynasty-trust slice in particular requires explicit GST exemption tracking; see generation-skipping transfer tax planning for the inclusion-ratio mechanics.
The illiquid-position complication
HNW portfolios contain assets that don't have clean daily marks: private equity stakes, real-estate holdings, fine art, collectibles, royalty streams, intellectual property, partnership interests in operating businesses. The platform has to:
- Capability 1Period valuation handlingQuarterly capital-account statements from PE managers; annual appraisals on real estate; ad hoc valuations on art. The platform has to ingest each on its own cadence and surface stale values clearly.
- Capability 2Capital-call and distribution schedulingPrivate partnerships have committed capital that's called over years and distributed irregularly. The platform has to track commitment / called / distributed / remaining for each fund.
- Capability 3K-1 ingestion and cascadeEach partnership produces a K-1. The cascade through to the household 1040 has to be modeled correctly (covered separately in our pass-through tax modeling article).
- Capability 4Concentrated-position managementHNW founders or executives often hold concentrated equity in a single company. The platform has to model concentration risk, hedging strategies (collars, prepaid forwards), and disposition planning.
- Capability 5Cost-basis reconstruction for legacy assetsReal estate purchased 30 years ago with multiple basis adjustments. Art with a chain of valuations. Engines that can't reconstruct basis defensibly produce wrong sale-tax projections.
The estate-planning complication
HNW estate planning is the headline service of family offices. The platform has to support modeling for:
- Lifetime exemption usage tracking (especially given the 2025 sunset)
- GST exemption tracking (separate from estate)
- Annual exclusion gift bookkeeping (per donee per year, with present-interest qualifications)
- 529-plan superfunding (5-year forward use of annual exclusion)
- ILIT funding via Crummey gifts
- GRAT performance vs §7520 hurdle rate
- IDGT promissory note tracking
- SLAT cross-trust mechanics
- Charitable-vehicle performance (CRT / CLT / DAF / foundation)
Each is a category of edge case. A platform that handles only outright gifts — the typical mass-market estate planner — covers maybe 20% of an HNW household's actual estate-planning activity.
What synthetic test data has to look like
An HNW family-office platform's test corpus, at minimum:
HNW corpus essentials
- 100+ households at varied net worth levels — $5M, $15M, $50M, $250M, $1B+
- Entity structures spanning the inventory above. Each household has 5-15 entities; the corpus across households has examples of every structure type.
- Multi-generation beneficiary structures with realistic relationships, ages, and conditional distributions.
- Illiquid positions — at least 30% of corpus value across the whole corpus is illiquid.
- Pre-IPO equity positions at various stages of the QSBS 5-year clock.
- Lifetime gift histories — mix of pre-sunset, at-sunset, and projected post-sunset planning.
- International exposure — foreign accounts, foreign-grantor trusts, FIRPTA-relevant real estate, FATCA reporting requirements.
- Multi-state and multi-jurisdiction tax considerations.
- Concentrated-position scenarios with hedging strategies in place.
A test corpus missing any of these is a corpus where 30-60% of the platform's code paths are untested — every branch that handles a trust, a foreign holding, a §6166 election, a concentrated position is in scope for an HNW household and out of scope for a mass-affluent one. Family offices know their own situation better than any vendor's data and surface a wrong output the same day they see it; the corpus has to model what the demo will land on, not the median household the platform was originally designed around.
Key takeaways
- HNW family-office platforms span a wider modeling surface than mass-market wealth-tech. The corpus has to be a different product, not a multiplier on the mass-affluent one.
- Entity diagrams routinely have 5-15 structures per household: trusts (revocable, ILIT, SLAT, GRAT, IDGT, CRT, CLT, dynasty, QPRT), LLCs, partnerships, S-corps. The platform has to model each.
- Illiquid positions require period valuations, capital-call scheduling, K-1 cascading, concentrated-position management, and cost-basis reconstruction for legacy assets.
- Estate planning is the headline use case. Platforms have to model lifetime exemption tracking, GST tracking, annual exclusion bookkeeping, and the sunset-window planning the 2025 sunset has triggered.
- Test corpus has to include 100+ households across the wealth spectrum, with entity structures, beneficiary structures, illiquid positions, pre-IPO holdings, lifetime gift histories, and international exposure.