Value Drivers: The Handful of Inputs That Move Firm Value

A value driver is an input that changes the answer rather than a number that describes it. Four of them decide what a business is worth: what a fresh rupee of capital earns, how much profit goes back in, the rate the cash flows are discounted at, and how long that return stays above the rate. Growth falls out of the first two.

Four inputs is a surprisingly small list, and the smallness is the point. Anything worth arguing about in a business, the order book, the new plant, the competitor cutting prices, the supplier who raised terms, is real and matters. But none of it reaches the calculation directly. Each one has to travel through one of the four, and until the route is named, its effect on the answer cannot be stated.

The list is short because the calculation is short. A value is produced by discounting a stream of cash the business is expected to throw off. Only the numbers that go into that arithmetic can move what comes out of it. Koller, Goedhart and Wessels refuse to let growth be discussed on its own for exactly this reason: their formulation sets it beside what capital earns, in a single line, and once it sits there something becomes hard to miss. Growth is multiplying a distance that already exists between the return and the rate. A positive distance gets magnified. A negative one gets magnified too, in the direction nobody wants. Damodaran supplies the same discipline from the other side. Somebody has to pay for a growth rate out of reinvestment, so nobody may assume one on its own.

What makes one number an input and another one just a description?

A household running a small tailoring shop shows the distinction. The takings went up this month. Is that an input into the value of the shop, or a description of what happened? On its own it is a description. The rise in takings becomes an input the moment a reason can be given: the shop bought a second machine, and that machine earns more per rupee spent than the money would have earned sitting still. A machine that earns more per rupee spent is a statement about a return on capital, and a return on capital is one of the four.

Apply the same test to a listed manufacturer. An input is a value driver when the arithmetic cannot be done without it, and everything else reaches the answer only by first moving one of those inputs. The rule names which door a fact has to come through, and other facts matter enormously. A price war is devastating. The war reaches the answer by lowering what a rupee of capital earns, and that is the door it comes through.

The worked example throughout is Sankalp Industrial Systems Limited, invented, a manufacturer of industrial valves, precision castings and the aftermarket parts and service that go with them. Every figure attached to it was written for teaching. Its Year 0 revenue is Rs 12,00,00,00,000, its earnings before interest and tax are Rs 2,40,00,00,000, and its operating profit after taxTrading profit with tax deducted at the full rate and interest left out entirely, so the figure describes the business and not the way it is funded. is Rs 1,80,00,00,000 after the company's own assumed effective tax rate of 25.0 per cent, chosen for the example rather than drawn from any jurisdiction. Revenue, trading profit and profit after tax are the raw materials, and raw materials are not yet drivers.

Because these drawings use colour to carry meaning rather than decoration, here is the code, and it holds on every figure below.

Driver one: what does a fresh rupee of capital earn?

Every forecast year, Rs 1,00,00,00,000 of net new capital goes in at Sankalp, and profit after tax steps up Rs 18,00,00,000, five times over. Year 1 shows Rs 1,98,00,00,000, Year 2 shows Rs 2,16,00,00,000, and the ladder finishes at Rs 2,70,00,00,000 in Year 5. Eighteen over a hundred is 18.00 per cent, and that is the return this forecast asserts on money not yet spent.

Set it against what the money already spent earns. Invested capitalThe money genuinely tied up in trading: the working capital the business needs to run plus the fixed assets it operates, taken together. at Year 0 is Rs 12,00,00,00,000 and operating profit after tax on it is Rs 1,80,00,00,000, so the base in place returns 15.00 per cent. The forecast therefore asserts that fresh money will do better than the money already there, 18.00 against 15.00 per cent, and nothing in the record stands behind that. The 18.00 per cent is the single most load-bearing number in the whole model, and printing it without saying it is asserted quietly turns a choice into a fact.

A vague assumption is impossible to argue with and a specific one is easy, so the assertion is worth stating precisely. The profit margin does not move: operating profit after tax holds at 15.00 per cent of revenue throughout, whichever vintage of capital produced it. The difference lies in capital turnoverRupees of sales generated by each rupee of capital tied up. A stallholder who sells through his stock four times over in a year turns his capital four times., and setting the two side by side shows it in one look.

So the whole of the 18 against 15 assumption is a claim that new machinery spins faster than old machinery. Nothing about pricing, nothing about cost control. A reader can go and interrogate that specific claim; nobody can interrogate the vague version. How a return splits into a margin and a turnover is worked through separately, and is used here rather than taught.

Everything else hangs on this first driver, and the household case shows why. A tailoring household deciding whether to buy the second machine is asking exactly this question and nothing else: what will this rupee earn, compared with what the rupee costs to raise. Everything about the shop, its location, its regulars, its reputation for a good finish, matters only in so far as it changes that number.

Driver two: how much of the profit goes straight back in?

The second driver is the only one management sets directly, and it is refreshingly concrete: what share of a year's profit gets ploughed back into the business rather than left free. Sankalp puts back Rs 1,00,00,00,000 a year, all five years. Because operating profit after tax is growing and the reinvestment is not, the same rupee figure is a falling share of profit each year.

The other three are estimated from outside or simply assumed, so reinvestment is the driver a board actually votes on. A board cannot vote for a higher return on new capital; it can only vote for the projects it believes will produce one. A board cannot vote for a lower discount rate. The one thing it does vote on, this quarter, is how much of the year's cash goes into new capacity rather than out to the people who provided the money. The number in that middle column is a decision, and the share in the right-hand column is the consequence of that decision meeting a profit line that is moving.

How the net figure of Rs 1,00,00,00,000 is arrived at, out of capital expenditure, depreciation and the movement in net working capitalStock and money owed by customers, less money owed to suppliers. Growth ties up more of it, which is why an increase counts as money invested., is worked through separately. Here it is simply read off, as an input rather than a subject.

If growth is not on the list, where has it gone?

Most people asked to name the things that decide what a business is worth will say growth before anything else. Growth is not on the list, and its absence is deliberate rather than an oversight. Growth is what those first two drivers produce when they meet, so putting it on the list as well would count the same thing twice.

Multiply the share put back in by what that money earns, and the growth rate falls out. For Sankalp the share is 50.51 per cent and the asserted return is 18.00 per cent, so 50.51 multiplied by 18.00 gives 9.09 per cent. Does the forecast agree? Profit after tax moves Rs 1,98,00,00,000 to Rs 2,16,00,00,000, and Rs 18,00,00,000 on a base of Rs 1,98,00,00,000 is 9.09 per cent. The two agree to the second decimal, and they agree in every year of the forecast. The full verification across all five years is worked through separately. One check is enough to show the shape.

Driver three: where does the rate come from, and why is none of it built here?

A rupee arriving in Year 5 is not a rupee arriving today, and the rate that converts one into the other is the third driver. For Sankalp it is 12.00 per cent, and that figure is taken here as a given: how a company's weighted average cost of capitalEverything the providers of money want back, collapsed into a single rate and applied to each forecast year. Arriving at one takes a treatment of its own. is built, what goes into it, and where each input comes from is covered separately.

The rate acts on every rupee of every year, and it acts hardest on the years furthest away, so it matters more than almost anybody feels it does. Nobody should take that on trust, and estimating a cost of capital is a subject of its own. Only the rate's position on the list is worth carrying away: one of the four, estimated rather than observed, and published by no company. A reader will not find it in a filing.

Notice the asymmetry in where the four come from. The reinvestment is read straight off the forecast lines. The return on new capital is asserted by whoever wrote the forecast. The rate is estimated by whoever is doing the valuing, using their own judgement about risk. And the fourth is usually not stated at all.

Driver four: how long does the return stay above the rate?

Earning 18.00 per cent against a 12.00 per cent rate is worth something. Earning it for three years is worth much less than earning it for twenty. The distance from what a rupee earns down to what that rupee costs has a length as well as a size, and a treatment showing only the size has shown half the driver.

Sankalp's explicit forecast periodThe stretch of years a forecast writes out line by line, before it stops and treats everything afterwards as a single lump. runs five years and asserts 18.00 per cent on fresh capital in each. Year 6 is not recorded at all, so whether the return holds, slides or collapses is unknown. There is no fadeThe habit of assuming a strong return slides back towards the ordinary over the years, rather than holding at its opening level indefinitely. profile in the case record, no end date, and no statement that the return persists. The silence is worth pausing on. Inventing a decline profile would put a number into the model that nobody has ever justified, and inventing a permanent 18.00 per cent would be worse.

Think about a tuition class that has become the one everybody in the neighbourhood sends children to. The class earns well on every rupee the teacher puts into another room and another set of desks. Earning well now is not what decides the value of the class. The years that pass before a second good teacher opens two streets away decide it. Nobody can answer that precisely, and pretending otherwise is worse than saying so.

What is the spread, and why does this company have two of them?

Take what a rupee of capital earns and knock off what that rupee costs; the remainder is the spread, and Sankalp carries two of them at once. Money already sunk into the business returns 15.00 per cent where the rate is 12.00 per cent, so its spread is exactly 3.00 points. Money going in from Year 1 is asserted at 18.00 per cent, measured against that identical rate, so its spread is exactly 6.00 points.

The second spread is precisely twice the first, and what separates the two is one assumption nobody has evidenced. Striking out the 18.00 per cent and putting the observed 15.00 per cent in its place would make half of everything the model says about value creation on new money disappear, and not one line of the accounts would have changed. The single asserted number carries that much weight.

Why is the same growth rate valuable in one business and worthless in another?

Here is the part that catches experienced people out. Growth is not good. Growth is not bad. Growth is a multiplier on the spread, so it magnifies whatever sign the spread already has and it does nothing at all when the spread is zero. Koller, Goedhart and Wessels therefore refuse to let the two be handled apart. Kept separate, they read as independent virtues. Set down in one line, one of them turns out to be a multiplier on the other, and a multiplier does not care what sign it is applied to.

Work it in rupees rather than in symbols, on Sankalp's own figures. A year of reinvestment is Rs 1,00,00,00,000. At the asserted 18.00 per cent return against the 12.00 per cent rate, that money earns 6.00 points more than it costs, or Rs 6,00,00,000 a year of surplus. Put in twice as much and the surplus doubles. Growth is doing its work there.

Now suppose the same money earned 12.00 per cent, exactly the rate. The surplus is 6.00 less 6.00, or nothing at all, and putting in twice as much gives twice nothing. The business is bigger, its profit line is longer, its revenue chart looks handsome, and the answer has not moved by a single rupee. Suppose instead the money earned 9.00 per cent. Now each Rs 1,00,00,00,000 put in costs Rs 3,00,00,000 a year more than it earns, and doubling the reinvestment doubles that loss. More growth makes the answer smaller.

A working control for what happens as the return crosses the rate is provided where the measure crossing zero is defined. Because a static check lands on a rupee the forecast already shows, it is the stronger evidence on this claim.

Do the four actually add up to this company's own cash flow?

A list of drivers is only worth having if it reproduces the arithmetic it claims to describe. Koller, Goedhart and Wessels write value as operating profit after tax multiplied by one less growth divided by the return on new capital, all of it then divided by the rate less the growth. Only the front half of that expression, the numerator, is needed to test a set of drivers against a forecast, and Sankalp's own numbers run straight through it.

Year 1 operating profit after tax is Rs 1,98,00,00,000. Growth runs at 9.0909 per cent against 18.00 per cent on new capital, so growth divided by the return is 0.505051, and one less that is 0.494949. Multiply: Rs 1,98,00,00,000 by 0.494949 is Rs 98,00,00,000. The forecast's own Year 1 free cash flow to the firmMoney left for everyone who financed the business, counted once the growth it plans has been paid for and before any interest or repayment. is exactly that figure, reached here from the drivers instead of from the cash flow lines.

The unrounded rate matters here. Using the printed 9.09 rather than the exact 9.0909 gives Rs 98,01,00,000, a lakh out, so the check runs on the unrounded figure.

Landing exactly matters because it converts the four drivers from a teaching device into an audit tool. When the drivers quoted in a valuation do not reproduce the cash flows in its own schedule, one of the two is wrong, and which one can be established before anything else is argued about. The check takes a minute and it needs no software.

Dividing the numerator by the difference between a rate and a growth figure, and the number that comes out of that division, is worked through separately.

What is not a value driver, however loudly it is discussed?

The list of things that are not drivers is much longer than the list of things that are, and it contains most of what gets talked about. Being blunt about it is more useful than being polite.

Revenue is not a driver. The revenue line here gains the identical Rs 1,20,00,00,000 each forecast year, so the growth rate slides downward while the rupees never budge. Revenue reaches the answer only by changing what a rupee of capital earns, and the quality and durability of revenue growth is worked through separately. A business can double its revenue and be worth less than before, if the capital it took to get there earns below the rate.

The operating margin is not a driver here, and this company proves it. The margin on earnings before interest, tax, depreciation and amortisation (EBITDA) holds at a flat 24.0 per cent across all six years, being Rs 2,88,00,00,000 on Year 0 revenue of Rs 12,00,00,00,000, and taking Rs 48,00,00,000 of depreciation off that leaves earnings before interest and tax at 20.0 per cent of revenue. Taxed at the invented company's own assumed 25.0 per cent, the profit after tax margin is a flat 15.00 per cent. Not one of the three moves by a basis point across the forecast, and the value still depends on all four drivers. Note also how those margins were reached: from the rupee lines, never from a percentage label. Run that check whenever a summary and a schedule disagree. A margin matters when it changes what a rupee of capital earns; sitting still, it is a description of the income statement. The meaning of an expanding or compressing margin, and how to read one that does move, is worked through separately.

Size is not a driver. Sankalp's invested capital rises from Rs 12,00,00,00,000 to Rs 17,00,00,00,000 across the five years, and its return on invested capital rises from 15.00 per cent to 15.88 per cent. Fresh money is asserted to work harder than the base it joins, so the rise has nothing to do with getting bigger and everything to do with the mix. Nallamala Components Limited, invented, sits in the same industry at Rs 5,00,00,00,000 of revenue, being 41.67 per cent of Sankalp's revenue, and earns 24.0 per cent on invested capital against Sankalp's 15.0. Smaller, and earning far more per rupee.

Earnings per share is not a driver, and it is the most dangerous of the four impostors. Sankalp's is Rs 6.90, being profit attributable to owners of Rs 1,38,00,00,000 over 20,00,00,000 shares. Nothing about it touches the return on capital, the reinvestment, the rate or the duration. A company can lift earnings per shareProfit belonging to shareholders, spread across the share count. Two businesses trading identically can print very different ones purely through funding. by substituting borrowing for equity, and it will have moved none of the four while the headline goes up. How leverage does that, and what it costs, is a subject of its own.

How does a lender, an analyst or a household actually use this list?

The four drivers are not an academic tidying exercise. Each kind of reader picks the list up by a different handle, and those three handles show better than any definition what the four are for.

A lender reads driver two backwards. Where an equity reader sees reinvestment as the price of growth, a lender sees it as cash that has left the building before any interest could be paid out of it. Sankalp's Year 0 interest cover, earnings before interest and tax over interest, is exactly 5.00 times, and Rs 98,00,00,000 remains in Year 1 once the Rs 1,00,00,00,000 has gone back in. A lender asked to approve a step-up in capital expenditure is being asked to accept a smaller number in that second column, and will want to know what driver one is expected to do in return.

An analyst uses the list as a discipline before opening any file. Write the four down. Beside each, write where it came from: read off, asserted, estimated, or not stated. For Sankalp that produces read off, asserted, estimated, not stated, in that order for drivers two, one, three and four. The moment that column is written, two of the four inputs to the answer are plainly standing on nothing, and the conversation changes. Standing on nothing is not a criticism of this forecast. Every forecast ever written does the same, and the fault lies only in not saying so.

A household saving for a shop extension is running the identical test with smaller numbers and better instincts. Will the extension earn more than the money would earn if it stayed where it is? If yes, doing more of it is better. If it earns about the same, doing more of it changes nothing except how tired everyone is. If it earns less, doing more of it is actively worse, and the fact that the shop will be bigger and the takings higher does not rescue it. Most people feel that asymmetry correctly about their own money and lose it entirely when a spreadsheet is involved.

None of those three readers concluded that Sankalp is cheap, expensive or worth buying. Naming the inputs that move a value does not produce a value, and a list of inputs is never a view about a security.

Which of the four are evidenced, and which are simply asserted?

Being honest about this is more useful than any of the arithmetic above, so here it is in one place.

The two inputs carrying the most weight are the two with the least behind them, and that is the ordinary condition of valuation rather than a defect in this particular case. The capital that is easiest to observe is the capital already spent, and it is the capital not yet spent that decides most of the answer. Naming which inputs are evidenced and which are asserted is a large part of reading any model honestly, and it costs nothing but the willingness to write four words in a column.

One caution about the fourth. Forever is what a formula quietly does when nobody intervenes, so a reader who finds no duration stated will often assume the answer is forever. Forever is an assumption too, and a much larger one than 18.00 per cent, and it deserves the same red ink.

Sankalp Industrial Systems Limited, Sankalp Coatings Private Limited, Aruna Tooling Private Limited and Nallamala Components Limited are invented.
Educational material. Not advice on any investment, tax, budget or market position.