Almost by definition, risks are never static; they evolve and change over time. And when risks do materialise, it can all happen very quickly. This is especially true of credit risk. Seemingly solvent companies can get into serious trouble inside a single reporting period.
One of the most popular metrics to assess credit risk is financial gearing, often defined as net debt divided by earnings before interest tax, depreciation and amortisation (EBITDA). The theory is that this measures financial indebtedness as a multiple of gross cashflow, giving an idea of how long it would take to repay the debt. Gearing of up to 3.0x is often considered reasonably safe for industrial companies. Unfortunately, the theory is flawed in several important ways.
The pitfalls
The first problem is that EBITDA is a deeply unsatisfactory number, as I have discussed previously. Treating depreciation as a non-cash cost is simply wrong. I have never met a company with a material depreciation charge that could take tangible capital expenditure down to zero. One can argue about the proportion, but it is very rare indeed to see tangible capex below 50% of depreciation. I don’t like the concept of splitting capex into maintenance and growth – it is almost always an inseparable blend of both – but companies cannot switch maintenance off completely.
It is not easy for auditors to challenge these numbers
The problem is even more acute with capitalised development costs. Unlike tangible assets, both the amount capitalised and the amortisation charge are inherently subjective. Management can in effect choose what to capitalise and the length of the amortisation period. It is not easy for auditors to challenge these numbers.
To make things worse, companies that get into trouble often find that customers delay payments and suppliers want paying up front. Working capital can easily become a major cash drain. The problem is even worse for companies with large contracts and advance payments.
EBITDA is not a good proxy for gross cashflow, yet it remains widely used
The net result is that EBITDA is not a good proxy for gross cashflow, yet it remains widely used for reasons that I find hard to fathom.
Case in point
Aston Martin is a classic recent case of how quickly things can go south. It launched a rescue rights issue on 31 January 2020, before the 2019 results were due. Gearing was 2.7x at the end of 2018, which is high but not generally considered scary. During 2019 it almost doubled to 4.76x due to lower EBITDA and higher net debt. From safe-ish to scary in 12 months.
Looking at the numbers in a bit more detail, the non-scary 2.7x leverage at the end of 2018 was calculated after capitalising £202m of development costs, with an amortisation charge of £61m. If we take a cautious view and back both these out, gearing was already 5.47x at the end of 2018. The EBITDA or ‘gross cashflow’ for 2018 was in effect boosted by 139% in this way. I cannot think of a better illustration of why EBITDA is an unreliable metric.
This was not a case of bad accounting but simply a business getting into difficulties
I am not criticising Aston Martin’s management or its auditors. This was not a case of bad accounting but simply a business getting into difficulties due to over-expansion and poor execution.
For equity investors it’s even worse. Most companies are valued by markets on an enterprise value (EV) basis, with the share price being in effect the EV less the net debt. When profits fall and debt rises, the equity gets crushed.
Lots of simple ratios have their uses. I always looked at EBIT as a percentage of revenues to gauge pricing power and operational efficiency. But if you are worried about financial risk, you need to look deeper. A misleading snapshot can be actively dangerous.
