personal investment blog

Dogs and tricks – new light from accounting changes?

Dogs and tricks – new light from accounting changes?

13 Jan 2018

The following paragraph is not true.

A neat way to value a company is to divide the share price by the earnings per share (EPS) which gives you something known as a P/E (price/earnings) ratio. A low P/E ratio (say <10x) implies that a share is cheap and a high P/E ratio (say >20) suggests expensive.

Many people, some of them claiming to be investment professionals or financial journalists, still promote P/E ratios (which came to be the standard valuation method in the 1970s and the 1980s). Here are some reasons why they are wrong.


The price of a share is a measure of one of a company’s liabilities (the equity owned by shareholders) but not the value of the company. The equity is what is left over after all other obligations have been met. The value of the equity is known as the market capitalisation of the company.

EI Group (formerly Enterprise Inns) has 479.5 million shares trading at 143p giving it a market capitalisation of £685 million. Also with a market capitalisation of £685 million is Go-Ahead Group with 43.2 million shares at 1586p.

Their earnings per share last year were 20.5p (EI Group) and 207.7p (Go-Ahead) giving them P/E ratios of 7.0x and 7.6x respectively. How cool is that? Are they both cheap and are they almost equally cheap?

You will not be surprised to read that it’s not as simple as that. The balance sheet of EI Group reveals that it the business is carrying more than £2000 million of net debt whereas Go-Ahead has £200 million of net cash. Consequently, the enterprise value of EI Group is £2700 million (market capitalisation plus net debt) and Go-Ahead’S enterprise value is just £485 million (market capitalisation minus net cash). On that basis the pub leasing business is worth 5.6x as much as the bus and train operating business.

This doesn’t tell us which share is more likely to go up but it gives us plenty of ideas about what might influence their prices. None of which involve reported EPS.


Another reason why P/E ratios are nearly useless is that reported EPS are prone to influence by accounting and other decisions taken by managements. Goodhart’s Law states that any measure will start to misbehave once it becomes a target. A long time ago company managements typically decided that if investors want growth in EPS, then growth in EPS is what they shall get. Their personal bonuses were often set according to exactly that target.

You can manipulate the “E” of EPS by declaring that certain costs are “extraordinary” even though they seem to occur every year and that certain income is “ordinary” even though it results from something that does not appear to be part of the company’s normal operations. You can even tamper with the “S” by using shareholders’ money to buy back shares. More “E” for fewer “S” and, magically, growth results. Bonuses all round! (I hate share buybacks).  

The first question that a sensible investor should ask is “how likely is this company to go bust?” I’m not going to say anything more about EPS other than to offer for your amusement the EPS of Carillion plc for the years 2014-16. 33.7p, 35.0p and 35.3p: how’s that for modest but stable-looking growth? Despite this, in 2016 the stock market began to realise that something might be wrong with the company and the shares have fallen from 300p then to 14p now as it continues negotiations with its creditors. The 2016 P/E ratio of 0.4x doesn’t offer much comfort, to be frank.


To judge whether a company is going to go bust it is necessary to be aware of its financial liabilities and the probability that it can meet them. This can mostly be done by looking at the balance sheet and the cash flow. When I look at a balance sheet I look for financial liabilities like bank borrowing and fixed term bonds and also pension liabilities. I net these off against cash and liquid investments. That gives a net debt figure.

There is no scientific method of judging if a company has too much debt but the key question is whether the business generates enough cash, at a minimum to cover its interest payments but also to pay for the capital expenditure on fixed assets and software that allows it to carry on as a competitive concern.  

This may sound daunting but it is really quite easy if you are basically numerate and willing to commit ten minutes. And it is much easier than trying to understand most businesses from the outside. If you are comfortable with a company’s financial strength you can sleep in the knowledge that, if it makes some mistakes or just hits a cyclical downturn, its viability will not immediately become an issue. I will illustrate how this applies to UK retail shares, though it is a sector about which I have very little understanding or natural interest.   


Now the story gets slightly exciting. From 2019 there will be a new accounting rule (IFRS 16) that brings liabilities known as operating leases from the notes at the back of the annual report on to the balance sheet itself. When a company decides to lease rather than buy an asset that is necessary to its operations – buildings or vehicles, for instance – it signs a lease that is generally compatible with the time that it judges that the asset will be needed and fit for use. Company cars are usually leased for three years, buildings typically for ten or fifteen.

This can clearly make sense. It saves the cost and inconvenience of buying and managing assets in which the company may have no direct expertise and it offers visibility and some flexibility. Under the old accounting standard it was considered that the financial liability (to pay the rent every year) could be excluded from the balance sheet because the company did not carry all the risks of ownership. For example, if the value of the asset declines more than expected over the lease period, the lessee can just hand back the keys and the loss belongs to the lessor or owner.

Needless to say, there could be a temptation to indulge in arguably unwise behaviour. For instance, businesses (and regrettably government departments and NHS hospitals) that owned buildings could engage in a sale and leaseback. They would sell the building for cash and agree to lease it from the new owner for many years. It became yet another example of jam today and liabilities dumped on the managers (and taxpayers) of the future.

For the supermarket group Tesco, realising the value of its property portfolio in this way became a regular feature of its annual results and in the first decade of this century it raised more than £5 billion through sale and leasebacks. Tesco became the object of much admiration as its real estate holdings effectively funded its retail growth, with UK revenues doubling from c£20 billion to £40 billion from 2000 to 2010.

The downside to this policy becomes apparent when the economy starts to slow down. How uneconomic does a store have to become before you are willing to buy yourself out of a ten year lease commitment in order to close it? Leasing becomes a burden rather than an agent of flexibility – a burden, moreover that could prove terminal in some cases.

It is for this reason, I imagine, that the gurus of accounting standards have decided that shareholders are being denied the true picture if operating leases do not appear on the balance sheet. And I think they have a point, but there might be some shocks to come.


When I look at the market value of a company I don’t look at the market capitalisation but at the enterprise value, as with the examples of EI Group and Go-Ahead given earlier.

Viewed by business sector it provides a swift overview of how much we are paying for each company’s revenues. If you click on the table below you will see a striking range, from just 0.22x for the middle-class comfort of Sainsbury supermarket to 1.50x for the sausage roll heaven of Greggs. (Valuations are in theory driven by results which are measured by numbers: Greggs makes 15p of cash flow for every pound of revenue while Sainsbury scores just 5p).



What really attracts my attention is the apparent cheapness of Debenhams, an ailing department store chain that, if it doesn’t cut its dividend (a gargantuan “if”) yields 11%. Yay!

But how will that table look if we add in operating leases? I have done the work and the result is quite shocking (click on the next table). I should say that I have used gross operating lease liabilities, regardless of their duration. In practice, all the figures will be adjusted to net present value (NPV) – the NPV of a long-term liability is lower than that of one that is due within a year.

Specifically, Debenhams is now the most expensive of the 13 shares rather than the second cheapest. Its net debt has grown by 24 times! It’s not so much Yay! as Nay!




On the 4th January Debenhams warned that Christmas trading had been disappointing and its shares fell by 17%. The market is beginning to understand that the management is constrained in what it can do by its operating leases. Debenhams has 166 stores in the UK plus 80 (mostly franchised) in the rest of the world. Back in October it said that it might close ten stores over the next five years. 25% of its operating leases mature within five years so presumably the stores will be among these: (52% have more than ten years to run).

My supposition is that the management would like much more flexibility to review its store portfolio. Instead it has resorted to rather desperate attempts to turn its stores into “destinations” by introducing food and drink franchises and even gym facilities.

I have written about the supermarkets before, nearly four years ago. At the time I suggested that Morrison had better asset backing than Tesco and Sainsbury for the reason that it had retained ownership of its store portfolio and was consequently better placed to ride out challenging competitive conditions. It has been a rocky ride for the supermarket shares since then but it is a small triumph for my view. Since March 2014, Morrison shares are +10% while Sainsbury are -18% and Tesco -30%. Given that Tesco has improved its balance sheet considerably, I have carefully started to buy its shares this year.   


According to work done by PWC and generously published on the internet, the median increase in net debt for the retailing sector will be 98%. The second biggest impact will be on airlines (47%).

EBITDA will typically be higher because lease payments will not be expensed as before. Instead the costs will be recorded as depreciation (D) and interest (I) (because the assets will be treated for accounting purposes as if they are owned).

This is potentially important because many company’s debt covenants are based on the ratio of net debt to EBITDA. In the case of companies with big operating lease liabilities, this ratio is likely to grow substantially meaning that debt covenants will be broken and need to be re-negotiated. This is unlikely to be a comfortable time for those companies and their shareholders. 

Leave a Reply