Join our community of smart investors

When debt is at its most dangerous

Pay less attention to the weight of debt and more to the pace at which it rises, that’s what kills
February 28, 2019

Debt is dangerous. Perhaps more thought should be given to that cliché. Certainly it can be true, but that doesn’t mean it always is. So the question arises: when is debt dangerous and when isn’t it?

Ben Broadbent, a deputy governor of the Bank of England with responsibility for monetary policy, tackled this subject earlier this year in a speech at the London School of Economics (LSE) where he labelled worries about the rapid growth in consumer debt in the UK as “puzzling”.

True, consumer debt relative to household income – though still below the levels at which it peaked in 2008 – has been rising lately and remains far above the rate at which it began this century. But the key point of Mr Broadbent’s address to the LSE was that neither the level of debt per se nor its relationship to a stable base – say, to net income – are necessarily strong indicators of trouble to come. What matters more is the pace at which the amount of debt has been growing.

For example, households in the Netherlands went into the 2008 crisis with much higher levels of indebtedness than those in either the UK or the US. Yet it was in the US – the least indebted of the three – where debt defaults were most severe, while households in the Netherlands – and its banking system – escaped comparatively lightly.

Structural differences in lending markets played big parts in those outcomes. Mortgage interest is still tax deductible in the Netherlands; that encourages borrowing and means that Dutch consumers have a long history of borrowing liberally. Meanwhile, in the US many mortgages were non-recourse, making it easier for borrowers to walk away from their commitments, damaging the lenders. Yet simply examining the pace at which mortgage debt rose in the three countries during the early 2000s – while ignoring the level of debt – proved a reliable guide to what was to follow.

So if we can establish that the change in levels of debt matters much more than simply the level of indebtedness, we can also ask: does this apply to the fortunes of companies as much as it seems to apply to consumers? From that, more relevant questions spring (at least for equity investors): does rising indebtedness affect the performance of a company’s shares? Is there an inverse relationship that, as a company’s debt rises, its share price is likely to fall?

Before examining the data, it’s good to understand why there could be a relationship between a company’s debt and the value of its equity (ie, its share price). After all, a basic building block of finance theory says it does not matter whether a company finances its activities with debt, equity or any combination of the two – its value will be unaffected. This is the famous ‘M&M hypothesis’ – named after two US economists, Franco Modigliani and Merton Miller – which says that if the future cash flows of two companies are identical then their value today must also be identical. If they weren’t, there would be an opportunity to make risk-free profits buying and selling securities of the two until their values did equalise. In those circumstances, the financing of the companies between debt and equity becomes irrelevant.

But the whole point of M&M theory is to show that it only applies in perfect markets. In the real world, markets are never perfect. Particularly for companies, taxation affects the choice between debt and equity since interest on debt is generally tax deductible, while ‘interest’ on equity (ie, dividends) isn’t. That often makes debt cheaper than equity, which also means that, in effect, the holders of a company’s debt subsidise the owners of its equity. Yet that subsidy comes with a cost – by inserting an additional layer of fixed cost between a company’s operating profits and profits accruing to shareholders, debt makes shareholders’ profits more volatile and less reliable. Therefore when debt rises beyond a certain point – and that point varies from company to company – it makes a company’s shares less valuable since the present value of the estimated future cash flow accruing to shareholders has to be discounted at a higher interest rate.

This much is true in theory. However, it does not tell the whole story since a company’s bosses raise debt finance for a reason, and rarely is that reason because they are desperate to pay the day-to-day bills. More likely, they are borrowing to grow, which implies more value in the future, not less. Alternatively, they may borrow because the nature of their company’s operations is changing, perhaps becoming more predictable. In those circumstances, raising debt and buying in equity can make sense since that should also raise the per-share value of the residual equity.

If factors such as growth and reliability can increase a company’s equity value even as debt rises, current trends in corporate lending may depress it. These are linked to the growth in so-called leveraged loans, which are bundled into securities and sold as collateralised loan obligations (CLOs) to institutional investors.

CLOs are sketched as the corporate sector’s equivalent of the CDOs (collateralised debt obligations) that helped to create the sub-prime mortgage boom in the US, which splattered in the 2008 financial crisis. That’s an exaggeration. True, issuance of CLOs has risen fast in the US and – to a lesser extent – in the UK, and borrowing covenants on the underlying loans have tended to become looser. Yet CLO structures are, so far, much simpler than those of CDOs and the rate at which leveraged loans have been securitised remains much lower than securitisation rates of sub-prime mortgages.

Even so, the growth in CLOs may be symptomatic of a lowering in lending standards. Whatever banks may protest, they cannot rapidly expand their loan books without diluting their standards. To that extent, fast-rising corporate debt implies loans being taken out by the wrong companies for the wrong reasons. At some point, that must extract the price of less equity value.

So we return to the original proposition – that rising corporate debt correlates with falling share prices. In the famous phrase of Sam Goldwyn, it’s a definite maybe. Certainly, our findings, based on returns from companies whose shares are components of both the FTSE 350 index and the FTSE Aim All-Share index, suggest a link, in some ways a strong one.

The details are shown in the tables below. Table 1 takes a broad-brush approach. Table 2 narrows it down. Table 3 focuses on the findings from the Aim All-Share index.

Table 1: FTSE 350 big samples  
 Sample sizeSlope of lineR-squared
Change on 2 years:  
Net debt/Ebitda159-0.52.3
Debt/equity210-0.11.3
Change on 4 years:  
Net debt/Ebitda1520.20.4
Debt/equity194-0.10.3
Source: FTSE 350 and S&P Capital IQ

Even for Table 1, the sample size is far fewer than the 350 components in the FTSE 350 index. Investment trusts and banks were excluded before the start. Those companies with insufficient share price history could not make the cut and common sense said that those without debt – either at the start or the end of the periods reviewed – could not make it either.

Simultaneously there was the question of how to define debt. Simply to measure the change in debt from the start to the end of an observation period did not seem right. Measures were needed that would assess the affordability of debt. Among the metrics offered by our database were net debt as a fraction of ‘Ebitda’ (basically, earnings before interest, tax, depreciation and amortisation) and gross debt as a fraction of shareholders’ capital. The Ebitda measure assessed the strain that debt might exert on cash flow and the balance-sheet measure did the same for capital resources.

Then we quantified the changes in these two measures over a two-year and a four-year period to early February. It did not matter whether a company’s debt ratios rose or fell over those periods. So long as a company could supply the data it was in the mix.

On a scattergram where each data point juxtaposes one company’s change in debt ratios to changes in its share price over the same period, the ideal result was to see the regression line, which plots the average distance between all the points, sloping downwards. In three out of the four broad-brush regressions for the FTSE 350, we got that (see the column for ‘slope of line’ in Table 1). It was strongest – actually, ‘least weak’ is nearer the truth – in changes to debt-to-Ebitda over two years. From a sample of 159 companies, on average net debt rose a little as a proportion of Ebitda during the period and share price performance declined. The column in Table 1 for ‘slope of line’ shows a value of -0.5 for this particular sample, which means that for every unit that the debt ratio rose average share price performance fell by half a unit. See the chart for a visual interpretation of this.

Clearly, however, it may not do to get too excited because the relationship between the regression line and its data points is extremely loose. This is quantified by the ‘R-squared’ column in Table 1, which gives a value of 2.3 for two-year changes in debt/ebitda. In crude terms, that means changes in the debt ratio accounted for a mere 2.3 per cent of the changes in share prices. In a sense, that should not be surprising. After all, to expect a nice, neat linear correlation between changes in debt and changes in share price when so many other things could be going on within and around a company is to expect too much. A company’s debt to Ebitda might double, which would mean an extra charge on revenue, cutting profits and perhaps hitting the share price. But many other factors might simultaneously be working favourably for the share price. It is enough to know that the overall thrust was in the right direction – share price performance declined as debt ratios rose.

Table 2: FTSE 350 fine tuned 
 High & rising debt/ebitdaFast rising debt/ebitda
Sample size1223
Average price change (%)-6-6
FTSE All-Share (% change)0.70.7
Number of fallers914
Slope of linena0
R-squaredna0
Source: FTSE 350 and S&P Capital IQ

Table 2 shows what happened when we fine-tuned the logic. If rising debt hurts, then it might hurt most where debt levels start high and get even higher; alternatively, it might be worst where debt rises the fastest. The table shows that it does not much matter – either way, debt hurts.

The first column shows results for the sample of just 12 qualifying companies – from a pool of 164 – where net debt started at 2.9 times Ebitda two years ago (then the average level for the whole index) and simply rose during the period to early February. The second column shows results for the sample of 23 companies where debt/ebitda rose fast during the period (by at least 40 per cent). This sample is bigger, but the result is the same. In both cases, the average price change was a drop of 6 per cent during a period when the All-Share index nudged higher (up 0.7 per cent). Also in both cases, there were many more fallers than risers. Nine out of 12 for high-and-rising debt; 14 out of 23 for fast-rising debt.

We can explore further. If debt hurts, then it is likely to be most painful for really vulnerable companies. To a large extent, that will exclude companies whose shares are in the FTSE 350 index. These are well-established corporate beasts that may well be bruised but are unlikely to be felled.

To find the weaker ones, we need to go where they gather – in the Alternative Investment Market (Aim); in particular, to the FTSE Aim All-Share index. Starting from an index of 786 components, within the Aim All-Share there are just 115 that can provide the data we need to show a change in net debt to Ebitda over the past two years and a price history to match. For these 115, debt levels have fallen in that period from an average of 4.7 times Ebitda to 3.6 times now. On average, the price of the 115’s shares has fallen 8.6 per cent and – as with the big sample for the FTSE 350 index – the regression line of changes in debt to share price changes points the ‘right’ way. That is, it slopes slightly downwards, indicating that, on average, share prices are punished as companies take on more debt.

Table 3: Aim All-Share fine tuned 
 High & rising debt/ebitdaFast rising debt/ebitda
Sample size1536
Average price change (%)-34-20
Aim All-Share (% change)-0.1-0.1
Number of fallers1326
Slope of line-0.01-0.01
R-squared7.41.5
Source: FTSE Aim All-Share350 and S&P Capital IQ

However, the fine-tuned data is much more revealing (see Table 3). Using similar logic to that applied to Table 2, we distilled the sample to companies that either had high-and-rising debt to Ebitda over the past two years or those that simply showed a big rise in the debt ratio over that period regardless of the starting level two years ago. To accommodate higher overall levels of debt among Aim-traded companies, the starting point for high-and-rising debt is 3.6 times Ebitda; for fast-rising debt companies, again the bar was set at an increase of at least 40 per cent.

The results are more startling than for the FTSE 350 index. Of the 15 companies whose debt ratio started at 3.6 times Ebitda and rose, the price of 13 fell during the two years and, on average, dropped 34 per cent compared with no change for the Aim All-Share index. The sample of companies whose debt ratio rose by at least 40 per cent was bigger – at 36 – so some regression to the mean was perhaps predictable and the outturn was an average price fall of 20 per cent where 26 of the 36 ended the two years with a lower price. Once more, however, the progression was not really linear – the values for the slope of the regression line and for the fit between the lines and the data points are insignificant.

So what are the conclusions? One is to do more number crunching. Suggestive though the findings are, we have not done enough research to have anything statistically significant. That said, it looks unlikely that a money manager could construct a money-making exchange-traded fund from large-scale portfolios based solely on the rate at which debt ratios have risen. Yes, the stats indicate that, on average, share prices are penalised as debt rises. But that ‘average’ is so close to the market’s index as to make a portfolio unprofitable after costs.

However, fine-tuning the exercise looks more hopeful. That such a high proportion of share prices fell among those companies whose debt ratios were high and/or rose fast suggests there are lessons on offer. First, those companies that take on debt faster than their ability to cope with it are heading for trouble whatever their bosses may say. Second, that before too long – within two years anyway, maybe sooner – the extra debt will erode the value of equity. In other words, the share price will fall. Which, in a way, brings us back to Modigliani’s and Miller’s famous hypothesis – any company only has a given amount of value, so the more of it that’s absorbed by debt, the less that’s left over for equity. That’s another way of saying, debt is dangerous.