Reinvestment Return

Growth is the part of a forecast buyers are most reluctant to pay for, and they have good reason. Growth costs money. A business that wants to sell more has to buy equipment, hire people, carry more stock and wait longer for more receivables. If the money spent to grow earns less than it costs, the business is bigger afterwards but worth less, and a buyer who paid for that growth has paid to destroy value.

Most owner-run businesses cannot say which kind of growth they have. They know they spent money on a new truck, a new location, a new product line, and that revenue rose. They do not know what each earned on what it cost. So the buyer treats growth spending as a cost and strips it out, draws the forecast's growth at a rate of its own, and gives the seller's growth story no credit at all.

Reinvestment Return shows what the business earns on the cash it reinvests. Growth capital spending and growth working capital are tracked to the revenue and operating profit they produced, so the forecast's growth becomes the product of reinvestment and return, something a buyer can check. It ranks second on the bridge, behind only working capital, because it is the one operating driver here that creates growth value.

Why the return is the whole question

One business at two returns shows why the return matters more than the growth.

In the illustration, a business earns $900,000 of NOPAT, grows at 3 percent a year and has a 12 percent cost of capital. If it earns 20 percent on what it reinvests, it needs to reinvest 15 percent of its NOPAT to grow at 3 percent, it produces $765,000 of free cash flow, and it is worth about $8.5 million. If it reinvested nothing and did not grow, it would pay out all $900,000 and be worth $7.5 million. Growth at a 20 percent return adds $1 million.

Now set the return to 12 percent, the cost of capital. The business has to reinvest a quarter of its NOPAT to grow at the same rate, its free cash flow falls to $675,000, and its value is $7.5 million, exactly what it would be worth without growing. Growth at the cost of capital adds nothing. Below it, growth takes value away, and the faster the business grows, the more it destroys. A buyer therefore cannot tell whether growth is worth paying for until it knows the return, and nobody knows the return unless someone has measured it.

A register of every growth project

The measuring happens in a reinvestment register: three years of growth capital spending and growth working capital, tracked by project or initiative, each with the incremental revenue and operating profit it produced, and reconciled to the classification of capital spending and the working-capital history.

The register works only if every growth project has a line before it is approved. The line states the spend, the incremental revenue and operating profit it is expected to produce, and when its results will be measured. The register's integrity comes from one rule: a project with no line is maintenance by definition, and is classified that way. That closes off the easy move of relabelling spending as growth after the fact, in a year when maintenance looked high.

Each project is then measured at fixed points after it starts, against what its line said it would earn. The finance lead records the actual incremental revenue and operating profit, and the project's sponsor signs the comparison. A sponsor who knows the project will be measured, and who has to sign the result, proposes the next one differently.

The register rests on two earlier pieces of work. It is a view of the growth half of the thirty-six months of classified capital spending built under Capex Discipline. Its growth working capital is the part of the working-capital build a project caused, which can only be isolated from the consistent history built under Working Capital Management. Tied to what it earned, project by project, growth spending stops being a claim a buyer has to take on trust in aggregate.

A return above the cost of capital, sized honestly

From the register comes the return on reinvested capital. It is computed on one definition across the three years, it sits above the cost of capital used in the forecast, and anyone can reproduce it from the register and the statements.

The definition is fixed once, with the outside accountant: incremental operating profit after tax divided by cumulative reinvestment, with the treatment of working capital and of timing written down and kept. A return whose definition changes from year to year can be made to say anything, and a buyer will assume it was.

The return then has to be sized by how many projects stand behind it. Three years of a small business's growth projects might be eight or ten projects, and a return computed from ten projects is an estimate with a wide margin. So the return carried into the forecast is pulled toward the industry base rate by a weight that rises with the number of measured projects and falls as their returns scatter. It is the same kind of weight used for the budget record, and the file shows how it was set. A register of ten projects that all earned close to 20 percent earns a heavy weight; three projects, one of which earned 60 percent, earn much less.

The register also changes decisions. It is reviewed every year, and a class of project that has repeatedly earned below the cost of capital is not repeated at the same size. Here the driver creates value directly, because stopping the projects that earn less than they cost raises the return on every dollar reinvested after them. A buyer can see from the same register that the growth it is asked to pay for earns more than it costs.

Growth computed from the register

The forecast's growth in each year is held to no more than that year's reinvestment rate multiplied by the return on reinvested capital. Any excess has to be attributed to a documented cause, such as a price increase or operating leverage.

The rule is built into the model. The growth line computes growth from the two inputs, and an analyst who overrides it has to write the reason next to the override. If the owner wants the forecast to show faster growth, the model asks the obvious question: reinvested where, and earning what? The buyer gets a growth rate it can trace back to the register.

Growth that adds profit at the margin

The register also reports the incremental margin on growth: for each of the last three years, the change in operating profit divided by the change in revenue, computed by line and reconciled to the split between fixed and variable costs the budget uses. Any growth year whose incremental margin fell below the reported margin is explained in writing.

Incremental margin answers a question the reported margin hides. A business whose revenue grows by $1 million and whose operating profit grows by $200,000 earned 20 percent on the new revenue. If its reported margin is 12 percent, the growth added profit at a higher rate than the business already earned, usually because fixed costs were spread over more sales. If the new revenue earned 9 percent, the growth diluted the business even though both revenue and profit rose. That is growth bought with margin, through discounts, cheaper customers or costs that rise faster than sales, and the reported margin may not show it for a year or two.

The report becomes routine once every cost line in the budget model is split into fixed and variable, one classification per account, kept with the chart of accounts. Incremental margin is then reported beside reported margin each year, by line, with a one-line explanation wherever the two differ by a material margin. The framework sets a default for that margin, which the lead may override with a recorded reason. A buyer reading the report sees whether growth adds profit at the margin or dilutes it.

How long the growth lasts

The forecast also has to be honest about time. It names the year in which growth settles to a mature rate, capped at the risk-free rate used in the cost of capital. It sets the period during which the business earns more than its cost of capital, its competitive advantage period, from a named base rate for how long businesses of this size sustain such returns. The register or the map of competitors may shorten that period, and nothing lengthens it beyond the base rate.

During the fade, the return on new capital falls in equal annual steps from the shrunk register return to the cost of capital. At and after the convergence year it equals the cost of capital, and the model reinvests only what the mature growth rate requires at that return. Every year of the fade, and the cap, appears as a line in the model.

As the first section showed, growth after the convergence year adds no value, because it earns exactly what it costs. All of the growth value sits in the advantage period and the fade, and a buyer can see how long that is claimed to last and why. The period comes from the base rate, the register and the market the business serves, and is never set to make the valuation work.

What it does to value

The framework counts three things. The growth value itself, the discounted value of reinvestment at the shrunk register return above the cost of capital over the advantage period, moves from no credit, which is what a buyer gives a growth rate without evidence, to full credit. Projects earning below the cost of capital are stopped, which raises the return on the next dollar. And any growth year whose incremental margin fell below the history is corrected before the forecast is valued. None of these is a price effect, which is why the driver ranks second.

It also defines the portfolio lever on this bridge. An acquisition is tested as reinvestment at its own return, on the price paid plus the cost of integrating it, and a divestment of a line earning below the cost of capital is tested by the cash it releases. Both are tagged on the register and ranked beside the operating levers. Crediting an acquisition with the organic return, or with a multiple step-up for being larger, counts price, and the framework does not value it.

What has to be in place first

The register is built from the classified capital spending and the working-capital history, so both come first. So does the integrated forecast from Forecast Model, because reconciling growth to reinvestment and return happens inside the model.

Knowing that it works

Every dollar of the year's growth spending should belong to a project that had a register line before approval, and that share is measured. The return on reinvested capital over the trailing three years is reported each year against the cost of capital. The forecast is checked for the share of each year's growth that reinvestment and return produce without an override. And incremental margin by line is set against reported margin.

The owner can be tested directly. Asked to name the best and worst growth projects of the last three years with their returns, from the register and without the finance lead, an owner who has to guess has shown the gap. Asked where the forecast's growth rate comes from, the owner should be able to say, in one sentence and before anyone asks, which reinvestment rate and which return produce it.

Why it slips, and what holds it

Growth spending stops being tracked to what it earned. Growth gets bought with margin. And the forecast's growth outruns the year it was meant to settle, as each new forecast pushes the convergence year a little later.

Each has its own safeguard. The register is the route to approval, so a project cannot be funded outside it, and it is part of the annual review. The growth line in the model stays derived from reinvestment and return, so a growth rate the two inputs cannot produce is caught when the forecast is built, before it has a chance to miss. And the incremental margin and the convergence year are re-derived from the register and the cost split with every annual forecast, so a drift in either shows up at the next budget, long before the next sale.