Depreciation, the sales-growth method and replacement cost on one borrower, and what each does to cover and cash leverage.
Maintenance capex is the spending needed to keep the business at its current size, and companies do not report it, so the lender has to estimate it. The most defensible estimate takes total capex and subtracts growth capex, measured as the average ratio of net fixed assets to sales times the year's increase in sales. On an illustrative borrower spending 14.0, that leaves maintenance capex of 7.19, against the 5.0 management states and 9.0 of depreciation, and the choice moves debt service cover from 1.94x to 1.69x.
Worked in full in Credit Analysis by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The number matters because it is the deduction from EBITDA that does not shrink in a downturn. Every cash-based ratio a lender uses, from the debt service coverage ratio to the years it takes to repay debt, depends on it, and it is the one input the borrower has most freedom to understate.
| Input | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Revenue | 180.0 | 195.0 | 212.0 |
| Net PP&E at year end | 72.0 | 78.0 | 85.0 |
| Net PP&E to sales | 0.4000 | 0.4000 | 0.4009 |
In year 3 the company spent 14.0 on capex and booked 9.0 of depreciation. Management's presentation says 5.0 of the capex was maintenance and the rest growth. EBITDA is 40.0, cash tax 4.0, interest 12.0 and scheduled amortisation 4.0 on 150.0 of debt.
The quickest answer is depreciation: 9.0. The logic is that depreciation spreads the cost of the existing asset base over its life, so replacing it at the same pace costs about the same. It ignores two things: assets bought years ago at lower prices, and assets whose accounting lives differ from their real ones.
The method associated with Bruce Greenwald works from the other end. If the business needs about 0.40 of net fixed assets for every unit of sales, then growing sales by 17.0 in year 3 (212.0 less 195.0, 8.7 per cent) needed about 0.40 × 17.0 of new assets. Whatever capex is left over kept the existing business running.
Growth capex = average (net PP&E ÷ sales) × change in sales = 0.4003 × 17.0 = 6.81
Maintenance capex = total capex − growth capex = 14.0 − 6.81 = 7.19
In Excel: =Capex-AVERAGE(PPE_1/Rev_1,PPE_2/Rev_2,PPE_3/Rev_3)*(Rev_3-Rev_2)
Average the ratio over three to five years, not the latest year alone, so that one lumpy project does not set it.
If the asset base is on average 6 years old and equipment prices have risen 3 per cent a year, replacing it costs 19.4 per cent more than its historical cost: 9.0 × 1.036 = 10.75. This is the estimate to use when the business is not growing, when the assets are old, or when the asset class has had real price inflation.
| Estimate | Maintenance | Cash available | Cover | Debt / (EBITDA − maintenance) |
|---|---|---|---|---|
| Management | 5.00 | 31.00 | 1.94x | 4.29x |
| Sales-growth method | 7.19 | 28.81 | 1.80x | 4.57x |
| Depreciation | 9.00 | 27.00 | 1.69x | 4.84x |
| Replacement cost | 10.75 | 25.25 | 1.58x | 5.13x |
On EBITDA the borrower is levered 3.75x whatever the estimate. On EBITDA less maintenance capex it is anywhere from 4.29x to 5.13x. Each 1.0 of maintenance capex is worth 0.0625x of cover and about 0.144 of a turn of cash leverage at the sales-growth estimate. Management's 5.0 is 0.56 of depreciation and 2.36 per cent of revenue: a business that books 4.25 per cent of revenue as depreciation and claims to sustain itself on just over half of that is either growing into new assets that will need replacing later, or deferring.
A useful cross-check. Total capex here is 1.56x depreciation. A company with stable sales and capex well below depreciation for several years is consuming its asset base, and the bill arrives in the downside, when the lender can least afford it.
| Sales growth in the year | Change in sales | Growth capex | Maintenance |
|---|---|---|---|
| 0.0% | 0.0 | 0.00 | 14.00 |
| 4.0% | 7.8 | 3.12 | 10.88 |
| 8.7% (base) | 17.0 | 6.81 | 7.19 |
| 15.0% | 29.2 | 11.71 | 2.29 |
The estimate is only as good as the growth year it is run on. In a 15 per cent growth year it attributes almost everything to growth and leaves 2.29 of maintenance; in a flat year all 14.0 is maintenance. In a year of falling sales it breaks altogether: a 5 per cent decline gives negative growth capex of 3.90 and a maintenance figure of 17.90, more than the company spent. Run it over several years and set it beside depreciation, rather than trusting any single year.
| Average asset age | 2% inflation | 3% inflation | 5% inflation |
|---|---|---|---|
| 3 years | 9.55 | 9.83 | 10.42 |
| 6 years | 10.14 | 10.75 | 12.06 |
| 10 years | 10.97 | 12.10 | 14.66 |
The common mistake is to accept the borrower's split between maintenance and growth because it is the only split anyone has published. Management has every reason to call capex growth: it flatters free cash flow, and growth capex can be cut in a downturn while maintenance cannot. A lender who takes 5.0 instead of 7.19 overstates cover by 0.14x and assumes 2.19 a year of discretionary spending that does not exist. The second mistake is quieter: holding maintenance capex at a percentage of revenue in the downside case. Machines do not get cheaper to maintain when revenue falls, which is exactly why operating leverage amplifies a credit downside more on cash than on EBITDA.
Never take a single maintenance capex number. Compute depreciation, the sales-growth estimate over several years and depreciation at replacement cost, and underwrite at the sales-growth estimate, averaged over several years, with depreciation as the floor whenever sales are flat and replacement cost when the assets are old. Here: 7.19 on the sales-growth method, 9.0 on depreciation, 10.75 at replacement cost, against management's 5.0. The walk from EBITDA to cash, with the three deductions that do not shrink, is laid out in the free workbook for this case.
It is a reasonable floor for a business with stable sales, but it is measured at historical cost. If the assets are on average 6 years old and equipment prices rose 3 per cent a year, replacing them costs 19.4 per cent more: 9.0 of depreciation becomes 10.75. For a growing business depreciation may overstate maintenance, so set it beside the sales-growth estimate.
There is no standard share; it depends on growth and asset intensity. In the illustrative case the sales-growth method attributes 7.19 of 14.0, or 51.4 per cent, to maintenance in a year when sales grew 8.7 per cent. In a flat year the same method would call all of it maintenance, so a fixed percentage taken from one year misleads.
Because calling capex growth flatters free cash flow and suggests spending that could be cut in a downturn. Management's 5.0 here is 4.0 below depreciation and 2.19 below the sales-growth estimate. Over five years that gap is 10.97 of underinvestment or of overstated cash flow, and the lender is the one who pays for it in a downturn.
This article is one calculation from Credit Analysis. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: credit, private credit and distressed debt → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.