Section 24 for portfolio landlords: the mortgage interest restriction at scale

A landlord with ten mortgaged buy-to-lets generating £30,000 of real cash profit can owe £21,432 of income tax on it, an effective rate of 71%. Nothing in that sentence involves avoidance, error or bad luck. It is the designed output of Section 24 of the Finance (No. 2) Act 2015 applied to a leveraged portfolio in the 2025/26 tax year, and as of June 2026 it remains the single biggest structural force pushing portfolio landlords out of personal names and into limited companies.

Most Section 24 articles stop at the single-property example. This one starts there, then runs the same maths across a ten-property book, because the restriction does not scale linearly: it compounds, and at portfolio scale it produces cliff effects, the higher-rate threshold, the personal allowance taper, the child benefit charge, that a one-property example never reveals.

What Section 24 of the Finance (No. 2) Act 2015 actually says

Before April 2017, a landlord's rental profit was calculated the way any business calculates profit: income minus expenses, with mortgage interest fully deductible alongside repairs, insurance and management. Tax was charged on the genuine economic profit.

Section 24 of the Finance (No. 2) Act 2015 changed the computation itself for individuals. Finance costs, mortgage interest, interest on loans to buy furnishings, and arrangement and broker fees connected to the borrowing, are no longer deductible expenses. Instead, the landlord calculates taxable rental profit without deducting them, pays income tax on that inflated figure at their marginal rate, and then receives a tax reducer equal to 20% of the finance costs, capped at 20% of the lowest of the finance costs, the property profits, or adjusted total income.

The restriction was phased over four years and has applied in full since the 2020/21 tax year:

Two scope points matter. First, the rule applies to individuals and to partnerships of individuals, not to companies: a limited company deducts interest in full and pays corporation tax on real profit. Second, the last personal-name carve-out has gone: furnished holiday lettings kept full deductibility until the FHL regime was abolished from 6 April 2025, so from 2025/26 holiday let income in personal names sits under Section 24 too, a point that matters for anyone weighing holiday let mortgages in a personal name.

The single-property maths: how £1,200 of tax becomes £3,000

Take one property owned by a 40% taxpayer: £15,000 rent, £9,000 mortgage interest, £3,000 of allowable running costs. Real economic profit: £3,000.

The tax bill is 2.5 times higher on identical economics, and it now equals 100% of the real profit: the landlord works for HMRC on this property. The mechanical driver is the gap between the 40% rate charged on the interest and the 20% credit given back: 20% of £9,000 is the £1,800 difference. For a basic-rate taxpayer the charge and the credit are both 20% and the net effect is nil, which is why Section 24 is invisible to small, low-income landlords and brutal to leveraged higher-rate ones.

The portfolio-scale problem: fictional profit at ten properties

Now run the same shape of book at ten properties. A full-time landlord, no other income, with £150,000 of rent, £90,000 of mortgage interest and £30,000 of running costs. Real cash profit: £30,000, the income of a modest salary. Under Section 24, taxable rental income is £120,000.

Three separate mechanisms stack here, and this is what the single-property example hides:

• The fictional-profit push into higher rate. A landlord whose real income is £30,000, a basic-rate income by any measure, is assessed as if earning £120,000. £79,730 of essentially fictional income is taxed at 40% while the credit refunds only 20%. The taxpayer's marginal band is set by money that was never theirs: it was paid to mortgage lenders.
• The £100,000 personal allowance taper. Adjusted net income of £120,000 strips £10,000 off the £12,570 personal allowance (£1 for every £2 over £100,000), adding roughly £4,000 of tax. Between £100,000 and £125,140 the effective marginal rate on the fictional income is 60%. A landlord earning £30,000 in cash is paying the tax system's most punitive band.
• The child benefit cliff. The High Income Child Benefit Charge claws back child benefit on adjusted net income between £60,000 and £80,000 in 2025/26. Our ten-property landlord with two children loses the full £2,251 of annual child benefit, again triggered by income that exists only in the tax computation.

For comparison, the identical book inside a limited company: £30,000 of real profit, interest fully deducted, corporation tax at the 19% small profits rate (profits under £50,000 in 2025/26) of £5,700, leaving £24,300 retained for the next deposit. The personal-name landlord keeps £8,568; the company keeps £24,300 before extraction. That gap, repeated every year, is the arithmetic behind the structural shift in our industry: most new purchase applications we see from portfolio landlords are now through SPVs, consistent with industry-wide figures from lenders' own reporting since 2023.

Who Section 24 does not hurt

Precision matters here, because the restriction is often described as a tax on all landlords. It is not. Four cohorts are untouched or barely touched in 2025/26:

• Basic-rate taxpayers who stay basic-rate after the fictional profit is added. The 20% charge and the 20% credit cancel. The trap is the phrase "after the fictional profit is added": the test is salary plus rent minus running costs against the £50,270 threshold, not real profit.
• Limited company landlords. Companies deduct interest in full. This is the entire premise of limited company buy-to-let and the reason lenders built a dedicated SPV product market.
• Low-leverage landlords. Section 24 taxes interest. A landlord with 25% LTV across the book has little interest to restrict; one at 75% LTV has the maximum exposure. Identical portfolios, radically different Section 24 cost.
• Landlords with no other income and modest books. Two or three properties with the personal allowance and basic-rate band to absorb the inflated figure often see little or no net effect.

The four structural responses, honestly assessed

1. Move income to a lower-rate spouse. Where one spouse is a 40% taxpayer and the other basic-rate, shifting beneficial ownership through a declaration of trust, with HMRC Form 17 where the property is jointly held, moves rental income into the lower band. It is cheap, quick and reversible, and for a couple with headroom in one basic-rate band it can neutralise Section 24 entirely on part of the book. Its limits are equally real: it only works while the receiving spouse has band headroom, it transfers genuine ownership with genuine consequences on divorce or death, and the lender must consent where the property is mortgaged. Full mechanics in our guide to declarations of trust on rental property.

2. Incorporate the portfolio. The complete fix: a company deducts interest in full and pays corporation tax at 19-25%. The honest assessment is that the cure has an entry price, capital gains tax (CGT) at up to 24% on accrued gains, stamp duty land tax (SDLT) at surcharged rates, and a full refinance of every mortgage, unless the reliefs in Section 162 TCGA 1992 and the SDLT partnership rules apply, and they apply to fewer landlords than the seminar circuit suggests. For a genuinely active, high-leverage, higher-rate portfolio with a long horizon, incorporation usually pays back within three to five years. The numbers, reliefs and traps are in our guide to transferring property to a limited company.

3. Deleverage. Section 24 is a tax on interest, so reducing interest reduces the tax. Selling one property to clear debt on three, or simply amortising, shrinks the problem without restructuring. The honest assessment: it works, but it surrenders the leverage that makes property returns attractive, and it crystallises CGT on anything sold. It suits landlords near retirement who want income over growth; it rarely suits landlords still building.

4. Sell up. The response HM Treasury arguably designed for. Disposal triggers CGT at 18% or 24% on residential gains with only a £3,000 annual exempt amount in 2025/26, and surrenders the income stream. For a marginal one or two property landlord with thin yields, exiting is sometimes the rational answer; for a profitable portfolio it is usually the most expensive possible reaction to a tax that restructuring can largely solve.

What this means at the mortgage desk

Section 24 is a tax rule, but its fingerprints are all over mortgage underwriting. Lenders stress personal-name higher-rate taxpayers at a 145% interest coverage ratio (ICR) against 125% for companies, precisely because the post-Section 24 tax leakage on a personal-name property leaves less rent behind to service debt: the stress test is the tax rule translated into credit policy. That asymmetry means the structural decision and the financing decision are the same decision. A higher-rate taxpayer choosing a personal name pays twice, once to HMRC annually, and once in borrowing capacity on every application, as our guide to portfolio stress tests and ICR shows in pounds. Run your own figures through the limited company vs personal calculator, and if a restructure is on the table, the refinancing is as material as the tax: we package the portfolio mortgage and SPV mortgage tranches that incorporations and partial restructures depend on, priced and sequenced before the transfer is committed.

Frequently asked questions