Education is the largest expenditure most of our clients never put in a spreadsheet. It is treated as a monthly cost, absorbed from income, until the year two children are at university simultaneously and the portfolio is sold at the wrong moment to cover it.
The size of the number
Take a representative case: two children, currently aged 4 and 7. International school in the Gulf or Singapore from reception to eighteen, then a three-year UK undergraduate degree as an overseas student.
- International school fees: roughly $25,000-$40,000 per child per year, before capital levies, transport and trips.
- UK university as an overseas student: £25,000-£40,000 tuition plus £15,000-£20,000 living costs, per child per year.
- UK private day school, if the family relocates: now materially higher following the addition of VAT to fees in 2025.
Inflated at a realistic 5-6% education-cost inflation — consistently above general CPI — the total for two children lands between $900,000 and $1.4m in nominal terms. It is a mortgage, paid over fifteen years, with no lender involved.
Why it should be funded separately
The liability has two features that make it unlike retirement. The dates are known to the month, and they are not negotiable. You cannot defer a university start by two years because markets fell 30%.
That argues for a dedicated pot with a glidepath: growth assets while the horizon is long, de-risked progressively so that each academic year's fees sit in cash or short-duration bonds by the September eighteen months before they are due. The pot should be denominated in the currency of the fees — a portfolio in dollars against sterling tuition is an unhedged bet on a date you cannot move.
Wrappers that fit
- Grandparent contributions. Often the most efficient route. Gifts out of surplus income are immediately outside the estate where properly documented, and paying a school directly can be structured as a normal expenditure gift.
- Bare trusts and designated accounts. Simple, tax-efficient in many jurisdictions because income and gains are the child's, but the child takes control at majority. That is a feature for some families and a problem for others.
- Discretionary trusts. Control retained, flexibility across siblings, at the cost of complexity and trust tax rates.
- UK Junior ISAs where a child qualifies, subject to residence conditions.
- Offshore bonds, where the 5% annual withdrawal allowance and assignment to an adult child in a lower bracket can be genuinely useful — but only outside a commission-loaded contract.
The mistakes we see
Funding from income only. Works until it doesn't — a business slowdown coinciding with two sets of fees is the classic failure.
One pot for everything. Retirement and education compete, and education always wins in the moment. Separate accounts create separate discipline.
Ignoring the university currency. Families in the Gulf frequently plan in dollars for sterling liabilities. A 15% currency move is a year of tuition.
Leaving it too late to be tax-efficient. Most of the useful structures — regular gifts out of income, bare trusts, insurance — depend on time. At five years to first fees, the options narrow to saving harder.
What we do
We build a fee model by child, by school year, in the fee currency, then design a glidepath and a wrapper set to match. It is a fixed-fee piece of work, usually a fortnight, and it typically changes the family's monthly savings target — up or down — by more than any investment decision they will make that decade.
