Add-Back Cap: a Worked Example

A leverage covenant is meant to tell the lender when a business has borrowed more than its earnings can carry. Each adjustment added to EBITDA pushes that moment further away. Some adjustments are fair: a genuine one-off legal settlement, or the cost of closing a plant that will not reopen. Others are a forecast dressed as history, such as savings the business expects next year counted in this year's earnings. The lender cannot audit every claim at every test date, so it limits how much the claims can do in total.

The pattern the cap guards against is familiar to anyone who has read a run of compliance certificates. Restructuring costs appear in a bad year, again the next, and again the year after; each time they are one-off. Writing in 2018 about leverage measured when global leveraged loans are made, the Bank of England noted that add-backs assuming future earnings improvements may overstate EBITDA and so understate leverage. A cap does not judge each claim. It sets a ceiling on how far the claims together can move the ratio.

Four choices decide what the cap is worth. The first is which items fall under it. Cost savings, synergies and restructuring charges usually do; non-cash items such as impairments and unrealised losses are often outside it, and so add back in full. The second is the measure: a percentage of EBITDA, a fixed amount, or the greater of the two, which keeps a floor for a business whose earnings fall. The third is the base. A cap measured on EBITDA after the adjustments are added is larger than one measured before, because the adjustments raise the base they are tested against, as the last line shows. The fourth is time: expected savings commonly count only if the actions are taken, or expected, within a set number of months, and some agreements add per-item limits inside the overall cap.

A packaging manufacturer, constructed for this page like every figure in it, has net debt of US$84m and EBITDA before adjustments of US$22m for the year to its test date. Its covenant sets net leverage at no more than 3.75 times and caps add-backs, including all four items claimed below, at 10% of EBITDA before adjustments. Both levels belong to the example and are not offered as a market norm. Management claims four items:

On its trading alone the manufacturer is in breach. With everything it claims it has three-quarters of a turn to spare and the covenant looks comfortable. Under the cap it passes by 0.28 of a turn on the strength of US$2.2m of the US$5.8m claimed, and the lender's report to its credit committee shows the 3.47, not the 3.02. Had the cap been measured on EBITDA after the adjustments, the allowance would have been US$2.44m and the ratio 3.44 times. The drafting choice is worth a few hundredths here; in a tighter year it is the difference between passing and a waiver request.

Next year is where the cap earns its place. If the US$2.5m of savings arrive, they show in reported EBITDA and need no add-back. If they do not, and a second round of restructuring costs appears, the cap stops the business from passing on claims alone.

A cap limits the quantity of adjustments, not their quality. Many definitions add qualitative conditions: savings must be reasonably identifiable, supportable by facts and certified by a director or officer, and some lenders ask for an accountant's report when the amounts are large. A lender reads both, and reads the claims as a list, since the cap tells it nothing about which items made up the allowance.

The cash test sits beside it. An add-back that was real shows in the cash flow, and one that was not leaves EBITDA ahead of cash. A falling cash conversion ratio alongside a rising list of adjustments is the reading that matters. On the other side of the leverage ratio, the cash netting cap does for net debt what the add-back cap does for EBITDA: it limits how much the borrower can improve the ratio with something other than trading. A loan that caps one side and leaves the other open still lets the ratio be improved from the side left open.

DEFINED IN ONE PARAGRAPH → · ADJUSTED EBITDA → · PRO FORMA EBITDA →

The cap is negotiated in the term sheet, where the borrower's adviser asks for a higher percentage, a base measured after adjustments, a long period for savings to arrive, and more items outside the cap. Each concession widens the gap between the earnings the covenant sees and the earnings the business reports. Some bilateral facilities take a different route and allow adjustments only with the lender's agreement, which turns each claim into a conversation rather than a calculation.

After signing, the cap shapes every compliance certificate. The credit officer reconciles covenant EBITDA to the management accounts, lists each adjustment, and records how much of the claimed total the cap allowed. A borrower that uses the full allowance at every test date is telling the lender something about its earnings, whatever its explanation of each item. The headroom the lender reports is the capped figure, and a request to lift the cap near a test date is commonly read as a covenant reset by another name.

LOAN COVENANTS, EXPLAINED → · DEBT TO EBITDA →