Stress-testing a balance sheet
Every ratio so far describes a normal year. The years that matter aren't normal. Here's how to put a company through a recession on paper, step by step, and read what comes out.
Ratios describe the good years
Every ratio so far is measured on this year's figures, so if this year was fine they'll look fine. A stress test asks what happens if next year is bad: sales fall, margins get squeezed and maturing debt is refinanced at a higher rate. Each hurts on its own. Together they compound.
Choose a bad year that fits the business
There's no single right shock. A water utility's revenue might fall 3% in a recession, a chipmaker's 40%. The best guide is the company's own history. Find its worst year in the last fifteen (2009 and 2020 are the usual suspects) and start from something at least that bad.
Then think about costs. Rent, salaries and depreciation don't fall when sales do, so profit drops much faster than revenue. That's operating leverage, and the DL 230 course on margins and returns explains it properly. For this lesson, it's enough to know that a 20% fall in sales can easily halve operating profit.
Calder Packaging in a recession
Calder (illustrative) makes boxes and cartons. It owes $1,100 million and has $150 million of cash, so net debt is $950 million. Of the debt, $400 million is a bond at 3.0% that falls due next year; the other $700 million pays 5.0% and runs for longer. It pays a $90 million dividend. Its loans carry a maximum leverage covenant of 3.5×.
The recession we apply: revenue down 20%, EBITDA margin squeezed from 18% to 14%, and the $400 million bond refinanced at 7.5%. Management cuts capital spending from $110 million to $80 million, and $40 million of cash comes back from working capital as receivables and stock shrink with sales. Tax is 21% of profit after interest.
Revenue fell a fifth and EBIT fell by more than half. Coverage went from comfortable to 1.9×. Calder still made money and even paid down a little debt, yet leverage jumped from 2.6× to 4.2× because the thing it's divided by shrank. The covenant limit of 3.5× is broken.
Look at what happened to the cushion from lesson 2. Today, profit could fall 46% before coverage reached 3× (1 − 3 × 47 ÷ 260). With the refinanced bond, the same calculation gives 25%. The dearer bond had eaten almost half the cushion before the recession even started.
What Calder could do, and what it would cost you
Cutting the dividend to zero is the obvious move, and it isn't enough. Net debt would end at $843 million, still 3.8× EBITDA. To get back under 3.5× Calder would need net debt of $784 million or less, or EBITDA of about $267 million. So it would also have to sell a business, issue shares at a depressed price or pay its banks for a waiver. Every one of those comes out of the shareholders' pocket.
Now run the identical recession on Brookline Containers (illustrative): same sales and margins, but only $400 million of debt at 4% and $200 million of cash. Coverage starts at 16× and, with $150 million refinanced at 7.5%, ends at 5.8×. Net debt falls to $149 million, 0.7× EBITDA, with the dividend paid in full.
Same business, same bad year. The only difference is the balance sheet they walked in with. A fortress adds little in a good year. In a bad one it lets the company buy cheap assets while others are forced to sell.
Banks and insurers
A bank's debt is mostly its deposits, so net debt to EBITDA, maturity walls and interest cover tell you nothing about it. Insurers hold premiums they'll pay out as claims, years later. Both need their own tools.
Dividend Line scores banks with a separate model on the same 0 to 100 scale and the same five grades. Its six walls are capital (tangible common equity ÷ tangible assets, full marks at 9% and nothing at 4%), the hidden losses on bonds held to maturity, funding, the earnings buffer before loan losses, credit quality and the dividend. Its stress test for Premium members, the Loss line, runs a year of bad loans through profit and then through capital. Insurers, and car makers whose finance arms lend to their customers, get no Fortress Score until they have models of their own.
The Survival map, the bad year and the List
The fifth wall, the bad year, is worth 15 points and is a small stress test in itself. Eight points ask whether free cash flow still covered the dividend in the company's worst year of the last decade. Seven ask how far profit could fall today before interest cover reached 3×, the same calculation as above.
For Premium members, the Fortress tab's Survival map runs a full grid on the company's real figures: profit falls along one axis and the rate it refinances at along the other. Each point on the map is coloured by what would happen, from a dividend paid from cash flow to bonds at risk once cover drops below 3×. Presets include a repeat of the company's own worst year, a 2008-style recession (profit down 25%, refinancing at 8%) and a credit crunch (profit down 10%, refinancing at 10%).
The Fortress List ranks every company we score. Its lenses are useful for this course: Walls cracking shows companies whose score fell 10 points or more in three years, and One wall in ruins finds good scores hiding a weak tower.
Takeaways
Sources for this lesson
- Edward I. Altman, Edith Hotchkiss and Wei Wang, Corporate Financial Distress, Restructuring, and Bankruptcy (4th edition, 2019)
- Martin Fridson and Fernando Alvarez, Financial Statement Analysis: A Practitioner's Guide (4th edition, 2011)