Beyond growth

But hang on - here in the UK we are in a mature economy and lots of companies don't grow. Their best days are behind them. It's a fact of life. For many companies, revenues are shrinking, not rising; profit margins are getting squeezed, not fattened; cash flows are a dribble, not a flood; new investment ends up returning less than the cost of capital. Corporate life goes into reverse.

So how do we value companies that are in decline, or worse? It's a problem because, as Aswath Damodaran, a finance professor at New York's Stern School of Business, says: "We are hard-wired for optimism." The solution, says Professor Damodaran, is "when valuing declining companies we have to go against the grain and estimate cash flows for the future that may be lower than cash flows today". What follows is a framework to do just that. It is based on a paper from 2009 by Professor Damodaran, whose explanations are worth taking seriously. He has written several books on valuation and is one of the 10 most downloaded authors on the academic website, Social Science Research Network.

The starting point is conventional enough. All company valuation begins with the axiom that a company is worth the present value of all the cash that it will generate in the future. That tells us everything because, as a principle, it's irrefutable. Yet it tells us nothing because we are still left with the challenge of putting quantities on those cash flows.

That challenge is addressed in two stages. First, to the extent that the near future - five years out, maybe a bit longer - is possible to forecast, we apply a standard discounted cash flow (DCF) valuation. Here, revenues, profit margins, profit and cash flow are all estimated and the cash discounted back to its value in today's money. Second, when the future gets really hazy, we guesstimate a terminal value - that's a lump sum derived from the growth and the profits that a company can sustain forever and it, too, is discounted to its present value.

The trouble is that when investors apply this process to declining companies, they make systematic errors, says Professor Damodaran.

■ They go into "auto-pilot optimism", where they assume growth rates that simply aren't realistic for geriatric companies; or they assume that the mighty force of 'mean reversion' will repair a company's profit margins.

■ They get into "discount rate contortions", where the interest rates they use for the cost of a company's equity and debt are based on historical rates that no longer apply; or they assume that poorly-performing companies in a healthy industry can raise capital at the cost of the sector average. Sometimes analysts compound the error by assuming that a barely-profitable company will somehow extract the full benefit of being able to offset interest charges against tax even though it won't make enough profit to achieve this.

■ They slip into "divestiture follies", where they make unrealistically optimistic assumptions about the proceeds that a company’s bosses can realise by selling assets. "In far too many valuations," says Professor Damodaran, “analysts seem to count on eating their cake and having it, too, by counting the cash proceeds from divestiture in the early years while not reflecting the loss of earnings in their forecasts in subsequent years."

These, and other errors, can produce serial overvaluation of declining companies with obvious implications for losses if investments are made on the basis of unrealistic figures.

So what's needed is a rigorous framework for dealing with declining companies. That starts with recognising decline and judging whether it is temporary or permanent. Then it moves to the question: is the company in distress or not? For discussion of these issues, see the box, 'When decline shades to distress'. In this part of the feature we are now going to focus on the nuts and bolts of constructing a valuation of declining companies that are not in distress. For this, download the interactive Excel spreadsheet - there are two tabs within the spreadsheet '1. Cost of capital' and '2. The cash-flow forecast'. Basically, you fill in the yellow cells and the spreadsheet does the rest. You can view the accompanying video for a full explanation.

For our working example, we are using publisher Trinity Mirror (TNI). It is best known as the owner of the Daily Mirror, but also owns a stable of provincial newspapers. Our assumption is that Trinity Mirror is a business in terminal decline; that is, its weaknesses are caused by secular forces, not cyclical ones. That's not necessarily as terrible as it sounds. It means that in its present guise, Trinity Mirror will never again be a growth company, but it is not about to disappear. It is not in distress because it can handle its debt load, which management has cut from over £400m four years ago to £120m net of cash today.

The first step is to put a cost on the company's capital. This is important because it will help drive the present value of the cash flows that the company should generate. Too low a cost and Trinity Mirror will look more valuable than it really is. We use the capital asset pricing model to propel the cost of equity. This is familiar enough. It assumes that investors get a risk-free rate of return (proxied here by the redemption yield on five-year gilts) and adds on an estimate of the risk premium that investors demand for holding the London market's equities. Then - and this is crucial - the risk premium is levered up or down by a measure of the propensity of the company's share price - Trinity Mirror's in this case - to bounce around, which is better known as its 'beta'.

To the extent that an extremely bouncy share price implies uncertainty, then beta can do a good job at highlighting risk. But it is notoriously unreliable. If, for example, a company's share price is moving wildly but with no clear relation to the market, then beta may end up with a low value, wrongly implying little risk.

City analysts, however, in their valuation models have a tendency to arbitrarily reduce high betas on the logic that mean reversion will perform that function soon enough anyway. That's a mistake. It has the effect of reducing a company's cost of equity and ignores the basic point that, with respect to a declining company, its cost of equity should be high for the very fact that the company is in decline. Indeed, with companies distressed by their level of debt there is a case for calculating a 'geared' beta. This ratchets up beta by the ratio of debt to equity and acknowledges that holding a company's equity must get riskier as its debt burden grows. Whether you apply the 'straight' or the 'geared' beta to the cost-of-equity calculation is a subjective choice. In Trinity Mirror's case, the straight beta seems fine.

Often, the mistake that investors make with calculating the cost of debt is to take a company's interest charges and express this as a percentage of the book value of its debt. That assumes that a declining company - or, worse, a distressed one - can raise new debt at the same price as maturing debt. That's unlikely. Where it's possible - and that's rarely - it is better to calculate the cost of debt as interest costs over the market value of debt. Alternatively, sometimes the notes to a company's accounts give sufficient information about interest rates on tranches of debt to work out the weighted average. With Trinity Mirror's fairly simple structure of sterling and dollar loan notes, that’s the case - and 7 per cent seems near enough the gross cost of debt.

Apply the effect of the tax break on interest costs - although only after first asking whether the company in question is likely to make enough profit to use this benefit (it should, in Trinity Mirror's case) - then average out the cost of capital by weighting the debt and equity in the capital mix and this stage of the job is done. Supply the inputs and the spreadsheet does all this, and for Trinity Mirror estimates a cost of capital of 10.9 per cent.

Phase two - working out the cash-flow forecasts and the terminal value - gets down to the nitty gritty. The table looks a bit complicated, but it's quite straightforward. As with most forecasting, much time is spent pondering the assumptions behind the estimates. Don't labour the process too much, though. The whole point of using computer spreadsheets is that several valuations, based on different assumptions, can be cranked out quickly. Then you might want to calculate the average-value figures based on the weighted likelihood of their outcomes. But more of that in a minute.

Start by inputting data from Trinity Mirror's latest full-year results. What's needed is figures for revenues, operating profit margins, the rate of corporation tax that the company paid and its return on capital employed. With these we have base figures from which we can start thinking about the company's future. What's going to happen to revenues, margins and so on in the coming five years?

In Trinity Mirror's case, forecasts from City analysts indicate that its revenues will fall in years one, two and three by the percentage amounts shown. For years four and five, let's assume that the pace of decline will come to a halt as the company retreats to a base from which it can grow slowly. This should be a standard scenario for a company that is declining but not distressed.

What about Trinity Mirror's future profit margins? In the past two years these have been unusually fat - 13.9 per cent - for a no-growth company. Most likely, at some stage in the next five years they will come under pressure, though who can say when? Let's allow for that by assuming an average margin of 12.5 per cent for the next five years. Obviously, the future won’t be that smooth, but, remember, this is modelling.

We make similar broad-brush assumptions about the rate at which the company will pay tax. From the outside, it's pretty well impossible to work out a company's corporation tax charge. Analysts depend on a steer from the company in question or, at the very least - and especially where a company has lots of tax losses to carry forward - a note in the accounts. In the absence of these - and assuming that the declining company will remain decently profitable - then the default position is that the company will pay tax at the standard rate. Input these rows of data and the spreadsheet will calculate forecasts for the company's after-tax profit for the coming five years. In Trinity Mirror's case, they start at £62m in 2013 and drift down to £56m in 2017.

Next we need to estimate future return on capital because the spreadsheet will - admittedly in a contrived way - use that estimate to value Trinity Mirror's capital invested and the amount of assets closed down or sold. This is a small part of the workings, but it acknowledges that as declining companies retrench they can throw off cash.

Our starting point is the 9.4 per cent return that Trinity Mirror made in 2012 and we assume a gradual straight-line improvement in the coming years to 9.9 per cent by 2017. Two things to note here. First - and pretty obviously - reality will not be so smooth. Second - and more important - Trinity Mirror's return on capital remains below our estimated cost of capital throughout. That implies Trinity Mirror will continue to destroy value. Sure, the company may well be throwing off cash and - as we will see - may create more value than the market price of its shares gives it credit for. Even so - and consistent with its status as a company in irreversible decline - it is a bad home for capital because it can't use it profitably.

Input figures for return on capital and the spreadsheet will automatically capitalise a value for Trinity Mirror's invested capital, which also represents its assets. As the capital invested declines (a function of capitalising falling profits), then the model assumes that the declining company has closed down or sold assets. So the question arises: how much cash might be raised from closures or disposals? To answer that a guesstimate is needed of the proceeds as a proportion of the assets sold - and it's very much a guess. This affects the ultimate value to investors because cash raised from asset sales will contribute towards net present value (and that applies whether disposals are made at a book profit or a loss). Often it may be best simply to input 'zero' into the relevant spreadsheet cells. In Trinity Mirror's case, we have assumed it can generate a small amount of cash from disposals over the years.

That completes stage one of the cash-flow forecast and we have figures for net profits and other cash generated for the coming five years. Stage two takes these and expresses them in today's money values. In addition, it calculates a present value for the cash that the company might throw off in the years beyond - its terminal or 'perpetuity' value.

Very little additional inputting is needed. The model assumes that in the years one to five the company's cost of capital will be as calculated earlier. Arguably, we might finesse this by cutting the cost as the company gets itself onto a more stable footing. Certainly, for the cost of capital that goes into the perpetuity-value calculation it makes sense to reduce the figure

Then we need figures for the long-term rate at which the company can grow from a stable base. Recall that we are dealing with companies in decline and we are shooting far into the future, so anything above 2 per cent is likely to be unrealistic. The stable return on capital is likely to be the same figure as we used in year five of the cash-flow forecast. And the reinvestment rate - the proportion of net profits that are reinvested for the company to maintain its pedestrian growth - is purely an arithmetic function of the stable growth rate as a fraction of return on capital.

These three figures - growth rate, return on capital and reinvestment rate - play the key roles in calculating the perpetuity value, the theoretical value of all the cash that the company will produce after the forecastable years run out. Essentially, we find this value by capitalising the company's net profits in year five - or as far out as we feel we can forecast - in much the same way as a 'constant growth' discount model works. That means the perpetuity value will be higher or lower depending on the following:

■ Net profits - the higher, the more value.

■ Stable growth rate - the higher, the more value.

■ Reinvestment rate - the higher, the less value

(although, arguably, the higher the reinvestment rate, the more we can push up the stable growth rate).

■ The cost of capital - the higher, the less value.

Apart from this, filling in the other cells is self-explanatory. We need the latest values for the number of shares in issue, the gross cash and the gross debt. The spreadsheet will do the rest. Using the chosen cost-of-capital figures, it will express the cash flows and perpetuity value in today's money values and judge whether the shares are cheap or dear. That depends on whether the current price is 80 per cent of the estimated value or less. True, 80 per cent is an arbitrary level, but it leaves that crucial margin of safety between price and estimated value.

Happily, it judges Trinity Mirror's shares cheap. But what if we reckoned that Trinity Mirror might not be in irreversible decline, but was a reversible situation - how could we filter that proposition into the valuation model? Generically, Professor Damodaran suggests that a company's decline is more likely to be reversible if:

■ The company has a track record of recovering from decline.

■ It is part of a healthy industry.

■ It is in a cyclical industry that should follow macroeconomic trends.

Unfortunately, that probably does not help Trinity Mirror. But, for argument's sake, what if we decided that, yes, the company could slot into that segment of the 'matrix of decline' table labelled 'low distress' and 'reversible' decline? How could we estimate its share value?

As alluded to earlier, we run two or more valuation exercises, where we change the assumptions. For Trinity Mirror, we will keep the valuation shown here but run another where we assume that sales stabilise sooner and return to growth by year five; that profit margins hold up better; that, as a result, return on capital rises a touch and cost of capital falls a little. Re-enforcing this positive feedback, we also assume that the stable growth rate will rise by a further half percentage point. Factor in these changes - none of which seems outrageous - and we have an estimated present value 12 per cent higher at 194p a share.

As to which we choose - a bit of both. The low-ball proposition looks more likely so we give it a 0.67 probability and 0.33 to the higher figure. So the weighted value of the two scenarios is 180p. That makes the shares look cheaper still. Not by a lot - although, given Trinity Mirror's array of businesses, we would not expect profound changes to value - but by enough to make the two-minute exercise worthwhile.

And that's how it is with valuing declining companies. Make lots of assumptions - all of them cautious, all of them sensible. Resist the temptation to use the discount rates by which growth companies are valued. Don't twist your inputs to get the output that you want. Do all that and the estimate of value that drops onto the bottom line has to be useful. At the very least it provides good food for thought, which is worthwhile because good investing is likely to stem from good thinking. And our little spreadsheet makes the process a whole lot easier.