Excess Cash: What Counts as Surplus and What It Does to Returns

Excess cash is the part of a balance a business does not need to operate. Sankalp Industrial Systems Limited, an invented company, holds Rs 1,20,00,00,000, of which Rs 40,00,00,000 is operating cash and Rs 80,00,00,000 is surplus. The split between the two is a judgement, not a disclosure. Surplus earns a bank rate rather than the business return, and where it is counted decides the return on invested capital that gets reported.

All of that rests on a distinction the accounts refuse to make. A balance sheet gives one figure for cash. Nothing in it says which part is float moving through the business every week and which part is a store of value sitting perfectly still. So every measure that treats those two halves differently, and every valuation that adds one and not the other, is standing on a line somebody drew by hand. There are three common ways of drawing that line, and each one moves the return the company reports about itself.

What is actually sitting inside that one cash line?

Take a household first. The shape is exactly the same one, and the amounts are small enough to feel. Suppose there is Rs 3,00,000 in a savings account. Part of it is next month: the rent, the school fee, the electricity bill, the groceries. None of that part is spare. The rent and the fees are already committed, and the money is only sitting in the account because the payments have not fallen due yet. The rest is the sleep-at-night money, the balance kept back so that a hospital bill or a broken scooter does not become a loan. Both parts are in the same account. The bank statement shows one number. Nobody at the bank has any idea which rupee is which, and no statement anybody issues will ever say which.

A company is the same thing at a different scale. Operating cash is float and surplus cash is a store, and the balance sheet reports their sum without ever separating them. Sankalp Industrial Systems Limited, a manufacturer of industrial valves and precision castings used throughout this worked case, ends its base year with Rs 1,20,00,00,000 of cash and cash equivalents. The split used throughout, and it is an assumption every single time it appears, puts Rs 40,00,00,000 of that in operating cash and Rs 80,00,00,000 in surplus. No published statement anywhere says so. Somebody decided it.

What does the operating half have to be big enough to do?

Operating cash earns its name by being needed. Wages fall due on a fixed day whether or not a large customer has paid. A supplier who ships castings on 45 day terms expects the transfer on day 45 and not on the day the money arrives from somebody else. Electricity, freight, statutory dues and the hundred small payments that keep a plant running all land on their own timetable, and none of that timetable is negotiated with the timetable on which customers pay. Operating cash is the buffer that absorbs the difference between those two timetables.

The size of the operating half is therefore a question about the shape of the business rather than a question about preference. A manufacturer collecting in 66 days, holding 73 days of inventory and paying suppliers in 91 days is running a cycle with real gaps in it, and the buffer has to be large enough to cover the worst week of that cycle rather than the average one. The test for operating cash is whether the business could clear next month without it, and the honest answer is almost never a single clean number. The honest answer is a range, and the number that gets written down is a point somebody chose inside that range.

How big is Rs 40,00,00,000, measured against something real?

A rupee figure on its own is unarguable. Being unarguable sounds like a strength and is actually the problem. If a note simply asserts that Rs 40,00,00,000 is the operating requirement, there is nothing for a reader to push against. So the figure gets converted into something with a unit attached, and there are three conversions in common use. Each divides the same Rs 40,00,00,000 by a different flow, and each produces a completely different sentence.

Take them in turn, on the base year figures for Sankalp Industrial Systems Limited, all of them invented. Cost of goods sold for the year is Rs 7,20,00,00,000, so a single day of cost is Rs 19,72,603 and the operating balance is 20.28 days of it. Revenue for the year is Rs 12,00,00,00,000, so the same balance is 3.33 per cent of a year of sales. Net working capital stands at Rs 1,80,00,00,000, and the balance is 22.22 per cent of that. Three tests, three numbers, one unchanged rupee figure, and not one of the three settles whether the split is right. The three tests make the split checkable. Checkable is a smaller claim than settled, and a much more useful one.

Watch what the day count does that the rupee figure could not. Somebody who reads the business and concludes it needs a month of cost in the bank now has a place to stand. Thirty days of cost is Rs 59,17,80,822, or Rs 19,17,80,822 more operating cash than this case assumes, and the surplus shrinks by exactly that much. A reader who wants a month of cost is not being sloppy. Such a reader has a different view of how lumpy the payment calendar is, and the arithmetic lets them express it. The disagreement has become specific. A specific disagreement is the whole value of the exercise.

What does a surplus balance actually earn?

Surplus cash is not idle in the literal sense. The balance sits somewhere and it earns something. In this worked case the Rs 80,00,00,000 is placed on deposit at a deposit rateInterest a bank pays on money parked with it for a fixed stretch of time, quoted before any tax is taken off. of 6.00 per cent a year, an assumed rate rather than one read off any bank's board. Six per cent on Rs 80,00,00,000 is Rs 4,80,00,000 of interest for the year. Most readers stop at that figure, and stopping there is the mistake.

Interest is taxable. Sankalp Industrial Systems Limited carries an effective tax rateTax expense divided by profit before tax. It is the rate a company actually paid, which is rarely the rate written into the statute. taken at 25.0 per cent throughout this worked case, again an assumed rate rather than a statutory one. A quarter of Rs 4,80,00,000 leaves the company as tax, so Rs 3,60,00,000 stays. Measured against the balance that produced it, that is a return of 4.50 per cent. The company keeps the after-tax figure and nothing else, so the rate that matters is 4.50 per cent, not 6.00.

Now put that beside the rate the company's own funding costs. The weighted average cost of capitalA blend: the return lenders want on their money and the return shareholders want on theirs, mixed in the proportions the company is funded in. for this business is 12.00 per cent, a figure settled elsewhere in this worked case and used here without being rebuilt. The 12.00 per cent is already an after-tax figure, and an after-tax figure is the only thing 4.50 per cent can honestly be set beside. Lining up 4.50 against 12.00 is comparing two numbers measured the same way. Lining up 6.00 against 12.00 is not, and it quietly shrinks the gap by a quarter.

What is that shortfall costing every year, in rupees?

Percentages are easy to nod at. Rupees are harder to ignore, so convert. Funding Rs 80,00,00,000 at 12.00 per cent implies the capital ought to be producing Rs 9,60,00,000 a year. The deposit produces Rs 3,60,00,000 after tax. The difference is Rs 6,00,00,000 a year, and the identical answer follows the short way, by applying the 7.50 point shortfall directly to the balance. Two routes, one number. The check is worth running whenever a spread gets converted into money.

Rs 6,00,00,000 is a number that can be compared with something. Sankalp Industrial Systems Limited generates economic profitProfit left over once the whole capital base has been charged for, not merely the borrowed slice of it. of Rs 36,00,00,000 in the base year, a figure settled elsewhere in this worked case and restated here rather than rebuilt. The shortfall on the surplus balance is one sixth of the whole year's economic profit, or 16.67 per cent of it. Nobody should file that under rounding, and nobody should bury it in a footnote. The shortfall is a real number attached to a real balance, and it is the strongest argument for taking the classification question seriously in the first place.

The Rs 6,00,00,000 is a cost, not a verdict. Whether Sankalp Industrial Systems Limited should be holding this much is a separate question with a separate answer.

Does cash belong in the capital figure at all?

Return on invested capital divides a profit figure by a capital figure. The locked measure for this business puts net operating profit after tax (NOPAT)What the operating side earned once tax has been taken off it, and before a single rupee of interest is counted anywhere. of Rs 1,80,00,00,000 over invested capital of Rs 12,00,00,00,000, giving exactly 15.00 per cent. Look at what that denominator is made of: Rs 1,80,00,00,000 of net working capital, and beside it Rs 10,20,00,00,000 of net fixed assetsWhat is left of the productive base after accumulated depreciation has been knocked off its original cost.. There is no cash in it whatever. Not the surplus, and not the operating half either.

Leaving cash out is a choice, and a defensible one. The argument for it is that the measure is asking how well the money put to work in operations is working, and a bank deposit is not put to work in operations. The argument against it is equally clean: the company holds the cash, shareholders funded it, and a measure of return on capital that omits capital the company is holding is measuring something narrower than it sounds.

One question settles it, and taste plays no part in the answer. If the income the cash earns is inside the profit figure, the cash belongs inside the capital figure, and if it is not, it does not. NOPAT here is operating profit after tax, and interest on a deposit is not operating profit, so the Rs 3,60,00,000 is nowhere in the Rs 1,80,00,00,000. The numerator has already answered the question. Everything after that is just making the denominator agree with it.

Why do two careful analysts publish two different returns?

The control above is a faithful picture of what happens when two people who have both done the arithmetic correctly make different classification choices, and it produces the table below. The same company, the same profit, the same day, and a spread of 1.36 percentage points produced entirely by a line nobody discloses.

Two things are worth sitting with here. The first is the size. Most readers, asked to guess before they see it, put the effect at a few basis points, and it is over a full percentage point. The reason is arithmetic rather than judgement: the cash balance is a tenth of the capital base, adding a tenth to a denominator divides the ratio by 1.10, and 15.00 divided by 1.10 is 13.64. Once that is seen, the size stops being surprising and starts being predictable.

The second is what does not move. All four returns clear the 12.00 per cent cost of capital comfortably, so the classification changes the measure without changing the conclusion this particular case supports. The clearance will not always be that comfortable. A business earning 12.60 per cent on the narrow definition would drop below its cost of capital on the wide one, and then the classification choice would be doing the entire work of the answer. Whether the choice matters depends on how close the ratio already sits to the rate it is being compared with.

What does the surplus buy that none of this arithmetic prices?

Everything above measures the cost of holding a surplus and nothing above measures what it is for. The asymmetry is not an oversight, and it should not be closed with an estimate. Go back to the household for a moment. The sleep-at-night balance earns almost nothing and it is not there to earn. The balance is there so that a hospital admission at eleven at night is a phone call rather than a loan at whatever rate is available at eleven at night. Nobody in that household has ever put a number on it, and nobody needs to in order to know why the balance is kept.

A company's version of that balance is optionalityRoom to act on short notice without needing anybody else to agree first, which is a real advantage that rarely carries a price tag., and it shows up in a small number of very specific situations. A major customer disputes an invoice and stops paying for two months. A supplier who takes payment in ten days offers better terms for doing so, and the arithmetic of that trade is covered separately. A furnace fails in a way the insurance covers slowly. A parcel of land next to the plant comes up and the seller wants an answer in a fortnight. In every one of those situations the surplus buys the ability to act before asking anybody, and this worked case attaches no rupee figure to any of them.

Where does the cash go when the business is valued?

A discounted forecast values the operating business: revenue, margin, tax, capital spending and the movement in working capital. A balance sitting on deposit produces none of those, so it is not inside the forecast at any point. The cash therefore joins exactly once, on the walk from the value of the operating business to the value of the equity, and it joins there rather than anywhere else. The full line by line walk, with every other item on it, is covered separately.

Once surplus has been defined, one question follows naturally and is worth settling here. If the surplus is Rs 80,00,00,000, does that later step pick up Rs 80,00,00,000 or the whole Rs 1,20,00,00,000? The answer here is the whole Rs 1,20,00,00,000. Preference has nothing to do with it, and the deciding fact is what the projection already contains, set out below.

Read the last row, because it settles the question on its own. Not a rupee of cash sits inside that base, so picking up the entire balance afterwards duplicates nothing. Had the operating Rs 40,00,00,000 been folded into the projection instead, picking it up again later would count it twice, and only the surplus could be picked up.

Some houses do fold it in, and they are not wrong to. Doing so means picking up the surplus alone and finishing Rs 40,00,00,000 lower. Spread over 20,00,00,000 shares, that is Rs 2.00 each. Either pairing holds together on its own terms, and what fails is a treatment that folds the cash into one side and picks it up again on the other. The treatment used here keeps every rupee of the balance outside the projection, and that is precisely why every rupee of it is picked up later.

Who actually uses this, and what do they do with it?

Four readers pick up the same balance and do four different things with it, and watching them is the fastest way to see why the classification is not academic.

A lender looks at the operating half and mostly ignores the rest. A working capital facilityA revolving bank line a business draws on and repays as its cycle turns, secured on the stock it holds and the money its customers owe. is sized against the cycle it is financing, so what a credit officer wants to know is whether the business can meet the next quarter's payments from what it collects plus what it already holds. A surplus balance is comforting, but a surplus balance can be paid out as a dividend on a Tuesday and an operating balance cannot. So the lender discounts the surplus heavily and sizes on the float. A valuation does the opposite with the same two numbers.

An analyst building a return measure has to declare the treatment before anything else. The whole point of a return on capital is comparison, across years and across companies, and a comparison between a narrow measure at one company and a wide measure at another is not a comparison at all. The practical discipline is small: pick one treatment, apply it to every company in the comparison, and print the treatment beside the number. An analyst who does that can be disagreed with. One who does not cannot even be checked.

Somebody valuing the equity wants the cash counted exactly once, and does not much care which convention gets it there. The failure that costs real money is not choosing the wide convention over the narrow one. The costly failure is adding the balance on the walk while also carrying its interest inside the forecast, or deducting borrowings net of cash while separately adding the cash back. Both count one asset twice, and both produce answers that look entirely reasonable.

An operator inside the business reads the split as a question about sleep rather than about return. A finance director who has once had to ring a bank on a Friday afternoon holds more operating cash afterwards, and the arithmetic in this guide will never talk them out of it. The extra cash is not irrational. The finance director is making a judgement about how bad the worst week can get, formed by having been through one, and what an analysis of this kind can honestly do is price the judgement and then leave it where it belongs.