With economic conditions heading downwards and the number of dividend cutters heading upwards, a company’s ability to generate cold, hard cash is becoming ever more important. And while reported earnings are useful when analysing the sustainability of a company’s dividend, I have finally realised there’s something better. And that something is free cash flow.
Earnings and free cash flows are not the same thing
The traditional way to measure the safety of a company’s dividend is to look at its dividend cover, which is the ratio between reported or adjusted earnings and the dividend. This is what I’ve always done. However, it’s far from perfect because dividends are paid out of cash, and earnings rarely reflect the amount of cash generated by a company.
For example, imagine that you own a small construction company. You sign a contract to build a house and the customer will pay you £1 million when the keys are handed over a year from today. It will cost you £600k to buy the land and build the house and all of your suppliers demand payment up-front. From a somewhat simplistic accounting point of view, you have £1 million in revenues, £600k in expenses and £400k in operating profits, and in theory you could pay out a large chunk of that profit as a dividend.
From a cash point of view, you have zero cash income and £600k of cash expenses. You’ll have to raise additional debt or equity just to cover the cash expenses, and the same goes for any dividend which you might (perhaps somewhat unwisely) decide to pay out before the house is completed.
For the most part, especially in large and established public companies, this sort of enormous discrepancy between earnings and cash flows doesn’t happen. There is always a difference, although it’s usually quite small, especially when measured over a period of several years. But that isn’t always the case, and sometimes earnings and cash flows can tell wildly different stories about how a company is performing.
Depreciation and capex: Two sides of an unbalanced coin
One distortion that crops up quite frequently is the impact of depreciation and capital expenses. Depreciation is a non-cash expense which is supposed to reflect the reduction in value of a company’s existing tangible fixed assets (trucks, property, machinery, etc.). Capital expenses are investments in new fixed assets and, despite the name, they’re not recorded on the income statement as an expense (they go on the balance sheet as an asset instead). In practice it works something like this:
Imagine you run a widget manufacturing company with one widget machine called Machine O. You bought Machine O ten years ago for £1 million and it depreciates in value every year as it becomes out of date and worn out. This deprecation of value is recorded as an annual expense, even though the deprecation doesn’t actually cost you anything each year. As per standard practice, Machine O is depreciated using the straight line method, which means you have an equal £100k non-cash depreciation expense in each of the last ten years. This reduced your reported earnings by £100k each year, even though in cash terms you spent £1 million in cash ten years ago and nothing since.
You can think of depreciation as an estimate of the amount of cash your company needs to put aside each year in order to replace Machine O when it eventually reaches the end of its useful life.
By coincidence, that happens to be this year. So, you decide to buy a new machine called Machine N, and because you’ve been a sensible CEO you did indeed save an amount of cash equal to Machine O’s depreciation each year (£100k per year, remember), giving you a £1 million cash pile to spend.
However, when you go to purchase Machine N you realise that £1 million isn’t nearly enough. Thanks to the wonders of inflation, a direct replacement for Machine O would now set you back £1.2 million rather than £1 million. Unfortunately, you don’t have the extra £200k because you paid out every penny of your earnings as dividends, and those earnings reflected the historic cost of depreciation rather than the future inflated cost of your next machine.
And it gets worse. Because you’re such a good CEO, your business has prospered and in order to maintain that growth, Machine N needs to have twice the production capacity of Machine O. That means it will cost twice as much, but twice £1.2 million, not twice £1 million. So, you find yourself with £1 million of savings because you based your cash allocation decisions (i.e. whether to save cash for a new machine or pay it out as a dividend) on depreciation costs and not the expected cost of future capital investments.
You have a few choices available, which you can mix and match: 1) Borrow the additional funds from your bank; 2) raise funds from shareholders via a rights issue; 3) cut or suspend the dividend (which won’t be nearly enough on its own); 4) sell off some of the company’s other assets.
None of these options are ideal and perhaps it would have been better if you hadn’t paid out all your reported earnings as dividends. The fact is that capital expenses (which aren’t deducted from earnings but do affect cash flow) are almost always higher than depreciation expenses (which are deducted from earnings but don’t affect cash flow), and that’s why reported earnings are routinely higher than the actual amount of ‘free’ cash a company is generating (free as in doesn’t have to be reinvested in the company, paid to the taxman, etc.).
This is one of the problems that free cash flow tries to resolve. Free cash flow starts with net cash from operations, which is basically operating profits with a few tweaks to add back in non-cash expenses like depreciation and take out cash expenses like interest and taxes paid. After that, we remove capital expenses and the end result is free cash flow. This means depreciation is effectively replaced with capex, and we’re left with a measure of ‘earnings’ that gives a much better picture of how much spare cash the company’s operations are throwing off.
Tesco: Massive capital expansion led to hidden dividend dangers
Tesco (LON:TSCO) is a good example of the kind of problems this gap between deprecation and capex can cause. Before 2013, Tesco was an obvious example of a successful business. For more than a decade it had grown revenues, earnings and dividends by more than 10% every year. During that time the company had doubled in size and expanded from a primarily UK-based supermarket chain to a global giant. Its dividend had been consistently covered more than twice over by earnings and, from that point of view, there was no reason to think the dividend was unsustainable. As for management, some were considered geniuses and handed knighthoods.
But that only tells half the story, and the other half is far less positive. Looking at the decade to 2012, Tesco recorded a total depreciation expense of £10 billion, which was deducted from earnings. At the same time, it recorded total capital expenses of £31 billion, which was not deducted from earnings. This means that a) Tesco was massively expanding its tangible assets (i.e. supermarkets, warehouses, etc.) as part of the UK supermarket ‘space race’ and b) its reported profits were boosted by at least £21 billion over the decade compared to the amount cash generated and available for dividend payments.
From a free cash flow point of view, the company generated £32 billion of cash from operations after interest and tax over those ten years. Take the £31 billion of capital investments away from that and you’re left with free cash flows of just £1 billion. Over the same period the company made dividend payments of more than £8 billion, so how did Tesco manage to pay out £8 billion in dividends while generating only £1 billion in free cash?
There are a couple of primary reasons: 1) The company was able to raise several billion by selling off old assets as they were replaced by new assets. 2) Between 2003 and 2012, Tesco’s debts increased by almost £7 billion. This more or less covered the growing dividend – but borrowing to pay dividends is unsustainable. At some point the company won’t be able to afford the interest payments on its ever-growing mountain of debt.
And that’s pretty much what happened. After the financial crisis of 2008/9, Aldi and Lidl started turning the UK supermarket sector upside down. Shoppers cared more about bargain basement prices than stores with creches and luxury coffee bars inside, and Tesco found its operating cash flows falling fast. Capital expenses were slashed in half and the supermarket space race came to an end, but it wasn’t nearly enough. With an annual cash dividend of more than £1 billion, Tesco simply wasn’t generating anything like enough cash to cover it, so the dividend was cancelled.
How to spot dividends at risk from weak free cash flows
When you look at Tesco’s track record through the lens of free cash flows, capital expenses and tangible fixed asset growth, it’s possible to see all manner of problems which just don’t show up when you’re focused on reported earnings. For each problem there’s one or more metrics which can help to highlight it, and I’ll outline a couple of my favourites here.
Free cash flow dividend cover: This is an obvious one. If free cash flows don’t cover the dividend then the company’s operations weren’t generating enough net cash (net of interest, tax and capital expenses) to pay the dividend. However, this isn’t necessarily a disaster and most companies have the occasional year when free cash flow dividend cover is below one.
An obvious example is when a company has to build or buy a new manufacturing plant or other expensive fixed asset. For a year or two the free cash available to shareholders will be low because capital investment is high, but this should be a relatively short-lived state of affairs (unlike Tesco, where this was the norm for more than a decade).
When free cash flows are low it’s quite easy to cover the dividend by reducing cash on the balance sheet or raising a small amount of debt. That’s entirely normal and acceptable, as long as it’s just a year or three.
So rather than measure free cash flow dividend cover over a single year, I like to look at it over at least the last decade. Specifically, I’ve started counting how many times free cash flow dividend cover was greater than or equal to one over the last ten years. This gives me a number from zero to ten, with zero meaning the dividend was never covered and ten meaning it was always covered.
Most companies sit somewhere in the middle of that range. In the decade to 2012, Tesco’s free cash flows covered its dividend a grand total of once, which is a very large red flag.
Free cash flow growth: As a dividend-focused investor I’m generally looking for companies with a track record of consistent dividend growth. So, in addition to free cash flows which consistently cover the dividend, I also want to see free cash flows increasing over time. I do this because I want to avoid the situation where dividends are going up, but free cash flows are going down, and this can occur even when free cash flows consistently cover the dividend (although if those trends continue, the dividend won’t stay covered by free cash flows for long).
Free cash flow growth can be tricky to measure because it can be extremely volatile. A company may have generated £1 billion in free cash flows a decade ago and £0.5 billion last year, but that doesn’t tell me much about the long-term trend. Perhaps the company just built a new factory this year, so for a single year free cash flows were depressed, but only temporarily.
To get around this problem I’ve started counting how often a company’s free cash flows went up over the last decade. This produces a number from zero (free cash flow went down every year) to nine (free cash flows went up every year).
As with free cash flow dividend cover, most companies sit in the middle, with free cash flows going up and down in about equal measure. Tesco’s free cash flows grew four times in the decade to 2012, which isn’t terrible, but it is below average.
Of course, there are many more things you can and should look at, but I think looking at how often free cash flows cover the dividend and how often free cash flows have increased, should prove their worth in highlighting unsustainable dividend payers.
And from a personal perspective, if I’d used those metrics to review Tesco back in 2012, I would never have made the mistake of joining Warren Buffett as a shareholder.