ICR and EBITDA cover: the two tests that decide a London commercial mortgage

There are two caps on any commercial mortgage facility and the lower one binds. The first is the loan to value cap, which everybody models. The second is the affordability cap, which fewer people model and which decides more London cases.

ICR, for let investment property

The interest cover ratio applies where the security is an income-producing asset let to tenants. The lender divides the gross rent by the interest cost at a stressed notional rate, usually 1 to 2% above the pay rate, and wants a result of 140 to 160% depending on the asset, the sector and the tenant.

Worked through: a building producing £120,000 a year of gross rent, a proposed facility of £1.2M, a pay rate of 7.0% and a stress at 9.0%. Interest at the stressed rate is £108,000. ICR is 111%. That case fails a 140% test and it is not close. To reach 140% the facility has to come down to roughly £950,000, which is a different deal from the one the borrower thought they were doing.

Note what did not appear in that calculation: the LTV. If the building is worth £1.8M, the £1.2M facility is 67% LTV and comfortably inside any lender's ceiling. It still fails, because cover is the binding constraint.

EBITDA cover, for owner-occupier and trading business

Where the borrower's own business occupies the premises, the test moves from the rent to the business. The lender takes earnings before interest, tax, depreciation and amortisation, and tests them against the full mortgage payment, capital and interest, at a stressed rate.

Owner-occupier: 1.3 to 1.5 times. Trading business: 1.5 to 2.0 times, higher because the security is an operating concern.

A business with £680,000 of EBITDA against an annual mortgage payment of £466,000 covers at 1.46 times. That clears a 1.3 times floor with room. The same business against a £520,000 payment covers at 1.31 times, which technically clears and which no sensible underwriter will be comfortable with, because a single bad quarter takes it below the floor.

Why the affordability cap binds so often in London

Because yields on prime London stock are compressed. A building bought on a 5% gross yield produces £50,000 of rent per £1M of value. At 75% LTV that is a £750,000 facility, costing £67,500 a year at a 9.0% stress. ICR of 74%. The arithmetic simply does not support 75% leverage at that yield, no matter what the valuer says the building is worth.

This is the single most common cause of the gap between what a London borrower expects to be able to borrow and what a lender will actually advance. It is not the valuer being cautious. It is the rent not being large enough relative to the capital value.

The consequence is practical. On a compressed-yield asset, the LTV ceiling is decoration. Model the cover first and the leverage second.

The adjustments underwriters make

Both tests are run on figures the underwriter has adjusted, not the figures you submitted.

On ICR, expect the lender to disregard rent from a tenant in arrears, to discount rent under a lease with a break inside the fixed-rate period, to strip out any element of turnover rent, and to apply a void assumption on multi-let stock.

On EBITDA, expect the lender to add back genuinely non-recurring costs, to normalise directors' remuneration to a market salary, to strip out any income that does not recur, and to look at the trend across three years rather than the best of the three.

Adjusted figures are almost always lower than submitted figures. Building your own model on the adjusted basis before you submit is the single most useful hour a borrower can spend.

What we do with this

We run both tests before we approach any desk, using each lender's own stress assumptions rather than a generic one, because the stress rates and the adjustments genuinely differ between lenders and that difference sometimes changes which lender wins the case.

If a deal only works at the very edge of cover, we say so on the first call. A facility that clears at exactly 140% is one rent review or one tenant default away from a covenant conversation, and it is usually better to take less money at a comfortable ratio than the maximum at an uncomfortable one.

Send the rent roll or the accounts and we will model both tests before anything else happens.