Could Payment Be Deferred Until an Asset Is Sold?

Norway lets an owner put off the wealth tax on business assets for up to three years. The price is interest at the central bank's policy rate plus five percentage points, and that margin is deliberate, set high enough that nobody mistakes the deferral for a discount.

That one detail answers most of this objection before it is made.

The objection, in its better form

The weak version says deferral is a loophole: let people wait and they will wait for ever. The stronger version grants that a wait can be made to end and asks what it costs. If the charge on an illiquid holding is pushed back until the asset is sold, the revenue is pushed back with it, and a government that needs the money this year has not been paid this year. Liquidity has been protected by handing the Exchequer a timing problem.

The strong version deserves its due. A budget is set a year at a time, and a charge that nobody pays this year is, for this year's purposes, a charge that does not exist. A Treasury told that the money will arrive when the founder sells has been told something about a date it cannot control.

That is a fair description of what deferral does. Whether it is a problem depends on two things the objection usually leaves out: what the debt earns while it waits, and what makes it come due.

What the debt earns while it waits

A deferred charge is a debt. It has a creditor, a due date and a rate, and whether the state is worse off for waiting turns on the rate. Price the wait at or above what the government pays to borrow and the Exchequer is paid later and not paid less; it stands where it would have stood had the tax arrived on time and reduced the government's borrowing by the same amount. Price it below that and the deferral is a subsidy, paid to exactly the people the tax was aimed at. The rate is therefore the whole of the design, and a scheme that leaves it out has not been designed.

Norway's scheme, in force from the 2026 income year, runs for up to three years, applies where the amount deferred is at least 30,000 kroner, and charges the Norges Bank key policy rate plus five percentage points on the balance. The tax authority's own worked example takes a deferred 50,000 kroner to 65,244.19 kroner by the time it falls due. The qualifying assets include shares, share savings accounts, partnership interests, and the fixed assets and commercial property of a sole proprietorship. All of that is current as verified on 12 August 2026, and Norway resets its wealth-tax parameters every year.

Three years is a short wait and the margin makes it an expensive one. That is the design choice. An owner who can raise the money will pay now, because waiting costs more than paying. One who cannot gets time, at a price the state does not lose on.

What makes the wait end

The second thing the objection leaves out is the trigger. A deferral with no end is an exemption with better manners. A deferral that ends on sale is secured on the one event certain to produce cash.

Britain already runs one. Where inheritance tax falls on qualifying property, section 227 of the Inheritance Tax Act 1984 allows the tax to be paid in ten equal yearly instalments, on a written election, and Finance Act 2026 extended the qualifying class to relevant business property. If the property is sold, the unpaid tax falls due at once. The sale that produces the cash is the event that ends the instalments, which is the security mechanism doing its work. The election gives a timetable, and the timetable is cancelled the day the money arrives.

Two limits travel with that precedent. It is a transfer tax, charged once at death, so a ten-year timetable applied to an annual charge would produce overlapping streams, each year's bill still being paid down as the next arrives. The interest treatment of those instalments sits in a separate section of the Act and is not asserted here. The provision is current as verified on 12 August 2026.

The alternative is a reduction, and a reduction is permanent

Spain shows what the other route costs. Its wealth tax does not wait for a sale. Article 31 of Ley 19/1991 caps the combined wealth and income tax bill at 60 per cent of the sum of the income-tax bases; where the cap bites, the wealth-tax charge is cut to fit, but the cut may not exceed 80 per cent of the charge, so at least a fifth is always payable. That too is current as verified on 12 August 2026.

The cap solves the same cash problem a different way, and the difference is the whole of this objection. A deferred charge is collected later. A capped charge loses the part above the cap for good, and loses it every year the cap applies, so the revenue given up is not a one-off either. The cap also carries a cost that a deferral does not, because it is linked to income: a taxpayer who can choose when to recognise income can choose how much wealth tax to pay. Spain's cap protects liquidity by giving up revenue. Norway's deferral protects it by waiting, at a price.

Where the objection is right

Deferral does move revenue into later years, and nobody can say how much. The yield profile of a deferral regime depends on how many people elect to use it, and for a tax that does not yet exist that number cannot be known in advance. The OECD's 2018 review of net wealth taxes lists liquidity provisions among the features of a well-designed one, alongside high exemption thresholds (which it makes conditional on broad capital-income taxes existing alongside), a comprehensive base, consistent valuation and limited exemptions; it puts no figure on what liquidity provisions cost. And no operating regime achieves accurate annual valuation, low administrative cost, neutrality between asset classes and full liquidity protection together; each gives one of them up.

So a UK design that offered deferral would be collecting part of each year's charge in later years for as long as the scheme ran, and anyone costing it should say so.

What a workable design contains

Security comes first: the deferred charge is registered against the asset itself, so the debt is enforceable against the asset whoever holds it. Interest runs at or above the government's own borrowing cost, which is the line between a timetable and a subsidy, and it compounds, so that the longest waits carry the largest bills. A sale triggers settlement, as section 227 already provides, so the moment cash arrives is the moment the balance is paid. And the term is capped, so the case of an owner who never sells is answered by the calendar.

Put together, that is a payment timetable. It costs less than an exemption because nothing is given away, and it costs less than Spain's cap because the charge survives the wait. A payment timetable is a cheaper concession than an exemption. See what deferral costs once it is priced properly.

Sources and evidence status

Skatteetaten, Defer payment of net wealth tax calculated on business assets (the Norwegian deferral from income year 2026, its cap, minimum, interest margin, worked example and qualifying assets), verified 12 August 2026. Inheritance Tax Act 1984 s.227 as amended by Finance Act 2026 Sch. 12 (ten yearly instalments by written election, the extension to relevant business property, the balance falling due on sale), verified 12 August 2026. Ley 19/1991 del Impuesto sobre el Patrimonio, Art. 31 (the cap on combined wealth and income tax and the ceiling on the reduction), verified 12 August 2026. OECD, The Role and Design of Net Wealth Taxes in the OECD, 2018 (the five design features). Comparative design position on valuation, cost, neutrality and liquidity.

The Norwegian, UK and Spanish provisions are T8 Legal or Regulatory Statements, current at the date given; Norwegian rules are reset annually, the Norwegian deferral is months old and the UK provision was amended by Finance Act 2026, so none should be relied on without checking its date. The UK provision belongs to Inheritance Tax and is cited as a transfer-tax precedent for instalment payment, not as a wealth tax; its interest treatment is not asserted. The OECD list is T5 Economic Interpretation from an official institutional report. The statement that no operating regime settles valuation, cost, neutrality and liquidity together, and the argument that interest at the government's borrowing cost leaves the Exchequer whole, are T5 Economic Interpretation drawn from those provisions. The design in the last section is a proposal in the brand's voice. No figure for take-up of a deferral, for the revenue it would move, or for the cost of liquidity provisions is asserted here, and none should be inferred.

Nothing here is tax advice.