Replacement Cost: Valuing What It Would Take to Rebuild

Replacement cost asks what it would cost to build the same business from nothing today. For Sankalp Industrial Systems Limited, invented, the tangible asset base would cost Rs 17,90,00,00,000 to rebuild, against a traded enterprise value of Rs 22,40,00,00,000. The ratio of the two is 1.25, the measure Tobin set out, and the Rs 4,50,00,00,000 difference is everything a rebuild would not get.

Start where the question actually starts, and that is nowhere near a spreadsheet. A woman on a nearby street runs a small printing shop. One offset machine, one cutting machine, a folder, a computer, a rented ground floor room, and about four lakh rupees of paper stock sitting in the corner because paper is bought when it is cheap and used when the orders come. Somebody offers to buy the shop. She has no idea what to ask for, so she does the thing every human being does first. She works out what it would cost the buyer to start the same shop themselves.

Machine, so much. Cutting machine, so much. Deposit on a room, so much. Paper to open with, so much. She adds it up and the total is her opening position, and the reasoning behind it is not naive at all. The total is a threat, politely expressed. Anyone unwilling to pay this can go and build it themselves, and see how they get on.

The threat is the whole method. Everything that follows is that same sentence asked about a company with three divisions and a balance sheet, with the arithmetic tidied and the awkward parts made explicit.

What question is replacement cost actually asking?

Replacement costWhat it would cost to build the same productive capability from nothing today. asks one question and only one: what would somebody have to spend, starting today with an empty piece of ground and a bank balance, to end up with the same productive capability this business has? Not the same company. Not the same reputation. The same capability to make things and sell them.

Notice how different that is from the other questions in this subject. A cash flow model asks what the business will produce and what that stream is worth today. A multiple asks what the market pays today for businesses that look like this one. Both of those look forward, or sideways, at other people. Replacement cost looks at the thing itself and asks what it would cost to make another one.

The inward direction of the question is why the method feels reassuringly solid and why it is quietly limited in the same breath. Replacement cost is solid because the inputs are prices that can actually be found. Steel costs what steel costs. A computer numerically controlled machining centre of a given size costs what the maker charges for it. Construction of a shed of a given area costs what a contractor quotes. Nobody has to forecast anything, and nobody has to agree about growth or about the right discount rate.

Replacement cost is limited for exactly the same reason. Anything a rebuild could not buy at any price is simply absent from the figure, and the figure gives no hint that it is absent. The method is silent rather than wrong. The size of that silence is measurable even though its contents are not, and the measurement is the most useful thing the method produces.

What is the economic argument underneath it?

The method rests on an argument about entry, and the argument is short enough to say in one breath. If a business could be rebuilt for meaningfully less than it is worth, then somebody with money and patience has a choice. The buyer can take this business at its going value, or spend less and build a competitor. If building is genuinely cheaper and genuinely equivalent, more capacity arrives, the extra capacity competes, and over time what the existing business is worth and what a rebuild would cost should be pulled toward each other.

So wherever the two figures stay apart, something must be stopping the rebuild from being equivalent, and naming that something on a particular company is the entire use of the method. The gap is a question about barriers to entryAnything that makes rebuilding a business harder than buying its equipment., asked in rupees rather than in adjectives.

Take the printing shop again. If a buyer really could put the same shop on the next street for less than she is asking, and the customers would not care which shop they walked into, then her asking price is in trouble and she will find that out quickly. But suppose half her work is printing examination stationery for two schools that took three years to trust her, and suppose the security clearance for that work took a year of paperwork. Now the next street is not the same thing at all. The equipment is identical and the business is not. The gap between the shop's worth and its equipment cost is the value of the three years and the year of paperwork, and no invoice will ever show it.

Which cost is meant, because there are three?

Here is where most of the confusion in this method lives, and it is worth slowing down. Three different things get called the cost of the assets, people slide between them mid sentence, and only one of the three answers the rebuild question.

Historical costWhat the assets actually cost when they were bought. is what was actually paid, on the day it was paid. Historical cost is a fact about the past and it is the figure the accounts are built on. Sankalp Industrial Systems Limited, invented, has a surplus land parcel that cost Rs 12,00,00,000 when it was bought, and that Rs 12,00,00,000 is a historical cost. Nothing about it is wrong. Historical cost simply answers what was paid. Nobody costing a rebuild is asking what was paid.

Current costWhat the same assets would cost to buy today. is what the same thing would cost if it were bought today. The same parcel of land is worth Rs 45,00,00,000 now. Same asset, different day, different figure, and the difference is not a gain anybody made. The difference is what happens to prices over the years an asset sits on a balance sheet. The work of marking the accounts from the first figure toward the second is covered separately.

Replacement cost is the third, and it is not the same as the second even though people use the two words interchangeably. Current cost asks what the same assets would cost. Replacement cost asks what the same capability would cost. The difference between those two questions is the entire reason line two of the build below is a deduction rather than an addition. A rebuild does not have to buy this equipment. A rebuild has to end up able to do this equipment's job, and today it would do that job with today's machines. Today's machines are not the same list.

How is the rebuild figure actually built?

In three lines. Three lines is genuinely all there is to it on this company, and three rather than thirty is worth saying out loud. The method's reputation for tedium comes from the underlying costing work rather than from the summary. Somebody has to price the sheds and the machines; once that is done, the summary is short.

All three lines describe Sankalp Industrial Systems Limited as it stands at the last completed year, called Year 0 throughout.

Line one: the gross block at current construction and equipment cost

Rs 18,50,00,00,000. Buying and installing the productive capacity this company has would cost that much today: the land, the sheds, the foundries, the machining lines, the test rigs, the handling equipment, the electrical infrastructure, the whole physical plant, priced at what a contractor and an equipment maker would charge for it now.

Two things this figure is not, and both are worth nailing down. The rebuild figure is not what the company paid, and it is not what the balance sheet carries. The accounts hold the gross blockThe full cost of the fixed assets before any depreciation is taken off. at what was paid, and then take depreciation off it year by year to reach a written down figure. Neither of those is a price anybody could pay today, so neither the historical gross figure nor the written down one is the input here. A costing exercise for a rebuild goes out and gets current quotations, and that is where the Rs 18,50,00,00,000 comes from.

Where does a real person get those quotations? From equipment makers, from construction cost indices, from the company's own recent capital spending on comparable equipment, and from whatever the plant engineers know about what a line of that capacity now costs. Costing a rebuild is legwork rather than modelling, and the legwork is why the method is slow to do properly and quick to do badly.

Line two: the allowance for what a rebuild would not reproduce

Less Rs 2,40,00,00,000. The deduction is the line people get backwards, and it is worth sitting with until it stops feeling strange.

Sankalp Industrial Systems Limited still runs two older production lines. They work. Both make saleable output. Both sit inside the Rs 18,50,00,00,000. The figure priced the physical plant actually standing there. But nobody starting from an empty site today would install those two lines. A new builder would meet the same demand with fewer, newer, faster machines. Those are what is available now, and what the capital cost per unit of output now favours.

So the allowance comes off, and it comes off because the question is what it would cost to achieve the same capability, not what it would cost to make an exact copy of an existing factory including its inefficiencies. A copy would faithfully reproduce two lines a new builder would simply not build. Reproducing them is a cost nobody would incur, so pricing them into the answer inflates it and quietly breaks the comparison the whole method exists to make.

Rs 2,40,00,00,000 is 12.97 per cent of the Rs 18,50,00,00,000 gross block, so this is not a footnote. The allowance is roughly one rupee in eight of the physical plant, removed on the grounds that a rebuild would spend those rupees differently or not at all.

The household version, if it helps. Consider the cost of setting up the same kitchen a household already has. The kitchen contains a large chest freezer bought fifteen years ago and now used twice a year. Costing the rebuild at today's prices, nobody would buy that freezer. Nor a smaller one. The rebuild would simply not have one, and would use the fridge. The freezer is in the kitchen and it is not in the rebuild, and leaving it out is correct rather than stingy.

Line three: working capital, because a rebuilt business still has to trade

Plus Rs 1,80,00,00,000. A factory with no working capitalReceivables plus inventory less payables, the money tied up in running the business. cannot trade. The factory can stand there, fully built and fully wired, and do nothing.

A rebuilt Sankalp Industrial Systems Limited would need the same cycle the real one runs: the same castings and valve bodies sitting on the floor as inventory, the same invoices outstanding with industrial customers who pay on their own terms, and the same balance owed to its own suppliers working the other way. On this company that cycle nets to Rs 1,80,00,00,000, being receivables of Rs 2,16,00,00,000 plus inventory of Rs 1,44,00,00,000 less payables of Rs 1,80,00,00,000.

The rebuilder has to fund that cycle out of their own pocket on day one exactly as the existing business funded it years ago, so it is part of what the rebuild costs and it goes in as an addition. Leaving it out is the second most common error in this method after getting line two the wrong way round, and it understates the answer by 10.06 per cent of the finished figure.

The printing shop had the same line and its owner would have found it obvious. Four lakh rupees of paper in the corner is not optional. A printing shop cannot open with no paper and tell the first customer to come back next month.

The table read backwards is a check, and the habit is worth building on any three line build. Rs 17,90,00,00,000 less Rs 1,80,00,00,000 is Rs 16,10,00,00,000, and adding back Rs 2,40,00,00,000 gives Rs 18,50,00,00,000, where line one started. The build reconciles in both directions, and a build that does not is saying something before anybody has looked at what it means.

What does the Rs 17,90,00,00,000 describe, and what may it sit beside?

The rebuild figure describes the operating assets. All of them, before any question of who paid for them. A rebuilder facing an empty site has to fund the whole thing whatever mix they eventually choose, so none of the three lines asked whether the sheds were funded by shareholders or by a bank, and none of them should.

Funding-blind arithmetic makes Rs 17,90,00,00,000 a firm-level figure, and a firm-level figure belongs beside an enterprise value rather than beside a market capitalisation. The pairing is not a stylistic preference. Pairing correctly is the difference between comparing two answers to the same question and comparing two answers to different ones.

And this company has two enterprise values, an inconvenient and honest fact. The traded enterprise value is Rs 22,40,00,00,000, built from what the shares actually change hands at plus what the lenders are owed and the other bridging lines. The discounted cash flow enterprise value is Rs 21,28,13,79,094, produced by a model of the company's own cash flows. How each of those is constructed is covered separately.

Against the traded figure, Rs 22,40,00,00,000 over Rs 17,90,00,00,000 is 1.2514, written 1.25. Against the model figure, Rs 21,28,13,79,094 over the same Rs 17,90,00,00,000 is 1.1889, written 1.19. Two enterprise values are available for this company and any ratio must name which one it used, in the same sentence, every time. The two answers differ by about six hundredths of a turn, enough to make two people comparing notes think they are discussing different companies.

Why is the market capitalisation the wrong partner for this figure?

Now for the trap on this particular company, and it is the sort of trap that catches careful people rather than careless ones, because it is spotted by looking closely.

The replacement cost is Rs 17,90,00,00,000. The market capitalisation of Sankalp Industrial Systems Limited is Rs 18,00,00,00,000. The two figures sit Rs 10,00,00,000 apart, or 0.56 per cent of the market capitalisation. Put them side by side and the ratio between them is 1.01, the tidiest finding anybody has ever produced: the market is valuing this business at almost exactly what it would cost to rebuild.

The tidiness is an accident, it means nothing whatsoever, and the two figures are not comparable in the first place.

Here is why. Replacement cost describes the whole asset base, funded by whoever funded it. Market capitalisation describes what is left over for shareholders after the lenders have been satisfied, a smaller and later claim on a different thing. Comparing them is comparing the cost of a house with the size of the deposit somebody has in it. The two numbers can land near each other and the nearness carries no information at all.

Each line of the walk from one to the other is a real claim on the business, so the walk set out below is worth reading as a walk rather than as a formula. Start at the market capitalisation of Rs 18,00,00,00,000. The lenders have a claim on the same assets, so add gross debt of Rs 6,00,00,00,000. A quarter of the consolidated subsidiary Sankalp Coatings Private Limited, invented, belongs to somebody else, so add minority interest of Rs 60,00,00,000. Take out cash of Rs 1,20,00,00,000 and non-operating assets of Rs 1,00,00,00,000. Neither is part of the operating plant a rebuild would put up. The walk lands at Rs 22,40,00,00,000, and how those lines are chosen and ordered is covered separately.

So the rebuild figure goes beside Rs 22,40,00,00,000, not beside Rs 18,00,00,00,000. And once that is done, the Rs 10,00,00,000 near equality that looked so satisfying turns out to be two unrelated quantities happening to land close together, in the same way that a person's age in years and the temperature in their city can coincide on a warm afternoon without either explaining the other.

What is the ratio of market value to replacement cost?

The ratio is what dividing one by the other produces, and the idea belongs to Tobin, who set out the comparison of the market value of a set of assets with what it would cost to replace them.

On Sankalp Industrial Systems Limited the ratio Tobin set out is 1.2514, written 1.25, taking the traded enterprise value of Rs 22,40,00,00,000 over the replacement cost of Rs 17,90,00,00,000. Read plainly, it says the market currently values this operating business at about one and a quarter times what somebody would have to spend to build an equivalent one.

Tobin was interested in the behaviour that follows. When the ratio is high across an industry, building new capacity is attractive relative to buying existing capacity, so capacity tends to get built. When it is low, building is unattractive, so capacity tends not to get built and sometimes gets shut. The high-ratio, low-ratio story is an argument about investment behaviour in aggregate, and a very different animal from a claim about any one company on any one day.

What does a ratio above one actually say?

Less than people want it to, and the reason is almost tautological once it is seen. A business is worth more than its equipment whenever it is worth more running than it would be as a heap of machines. A business worth less running than its parts is a business somebody is already thinking about stopping, so the condition describes nearly every business that is still operating.

A reading above one is therefore the normal state, not a finding, and treating it as a finding is the single most common way this method gets misused. Sankalp Industrial Systems Limited at 1.25 against its traded enterprise value is unremarkable in that sense. The reading gives the direction, predictable enough, and the size, the part worth having.

The size is worth having because it is a measurement of something that is otherwise very hard to measure. Rs 4,50,00,00,000 separates the traded enterprise value from the rebuild cost, being 20.09 per cent of that traded enterprise value. Whatever is stopping a rebuild from being equivalent, on this company, is worth Rs 4,50,00,00,000 to the people setting the price. Rs 4,50,00,00,000 is a real quantity even though its contents are not itemised anywhere. Run the same subtraction against the model enterprise value instead and it is Rs 3,38,13,79,094, being 15.89 per cent of that figure, and the base has to be named because the two answers are not the same.

What sits inside the Rs 4,50,00,00,000?

A guess made before reading on is more useful than the answer arrived at without one.

On this company, three specific things. Not three categories, not the word intangible waved at a total, but three nameable items that a rebuilt factory would not have on its first morning and could not go out and buy.

One, the approvals. The plant operates under consents that took time to obtain, and the time was the cost rather than the fee. A new entrant would have to obtain them again, in their own name, on their own site, from the beginning. Nobody can sell a consent for a plant that has not been built. Which consents a plant of this kind needs, and what conditions attach to them, is a matter for the authorities that set them.

Two, the installed base. The aftermarket parts and service division of Sankalp Industrial Systems Limited earns Rs 54,00,00,000 of earnings before interest, tax, depreciation and amortisation (EBITDA) on a 30.0 per cent margin, and it earns that because valves the company sold years ago are running in customers' plants right now and need parts, seals, servicing and eventual replacement. The installed baseEquipment the company has already sold that is still running in customers' plants. is the accumulated residue of every sale the company ever made. The only way to have equipment running in a customer's plant is to have sold it to them earlier, so a rebuilt factory would have none of the installed base on day one and could not acquire it at any price. Why aftermarket work carries the margin it carries is a question of business analysis and is covered separately.

Three, the customer qualifications. Industrial buyers do not simply order from whoever quotes lowest. A buyer qualifies a supplier first. Qualification typically means audits, trial batches, documentation and time, and only then does the supplier become orderable at all. The customer qualificationsApproval by a buyer that a supplier may be ordered from at all. attach to the company, not to the machines. Buying the machines gives a buyer machines. Owning them still does not make the buyer able to quote.

Why is no figure put on any of those three?

Because the split cannot be known, and pretending otherwise would be the single most damaging thing to do with this arithmetic.

Here is the shape of the temptation, and it is a strong one because it produces work that looks finished. There is a total of Rs 4,50,00,00,000 and there are three named items. The next move writes itself: assign a figure to the approvals, a figure to the installed base, and let the qualifications take whatever is left. The three add to Rs 4,50,00,00,000 exactly. The table looks like a valuation of intangible assets. The table can be put in a note, defended in a meeting, and carried into the next month's work by somebody who was not in the room when it was made up.

The total was fixed by a subtraction before the split existed, and nothing whatever establishes the split itself, so every rupee of that split is an invention dressed as a finding.

Think about what the arithmetic actually produced. The subtraction produced a residualWhat is left after a subtraction, whose size is known and whose contents are not.: a number known only because both ends of a subtraction are known. A subtraction gives the size of what is left over. A subtraction says nothing at all about what is inside the leftover, in the same way that knowing a sealed box weighs four kilograms says nothing about how much of that is the packaging.

The third figure is made whatever is required, so any three figures chosen will add to Rs 4,50,00,00,000. The perfect addition is not evidence and it is not a check. Perfect addition is a consequence of how the last row was built.

So the work stops there. The three contents are named, the size of the total is stated, and the total is not divided among them. Naming what is left over and not pricing it is the correct treatment of that Rs 4,50,00,00,000.

The dissatisfaction that follows is the feeling of wanting a number nobody has, and noticing that feeling is worth more than any of the three figures that might have been written down.

When is replacement cost the right method, and when is it empty?

The condition is checkable before the work starts, unusually kind as conditions go. Ask whether the assets could be priced line by line from equipment catalogues and construction rates, and let the honest answer decide whether to do the work at all.

A valve factory passes that test easily. Sheds have an area and construction has a rate. Furnaces, machining centres, test rigs and handling systems have makers who will quote. Land has a market. Working capital is arithmetic off the accounts. Every line is priceable, and priceable lines are what make Sankalp Industrial Systems Limited a fair worked instance, along with capital-heavy businesses generally: cement, steel, shipping, hotels, warehousing, power generation, sugar, paper. In all of those, somebody can walk the site with a costing sheet and come back with a defensible figure.

Now the other side. Consider a business whose main asset is a research team of forty people, or one holding a licence that is not currently being issued to anyone new, or one whose value is a reputation built over thirty years of not letting customers down. Costing the tangible assets of any of those produces a small figure: some computers, some leased offices, a little inventory. The figure is arithmetically correct and it is useless.

The danger is that the small figure looks conservative, and it is not conservative, it is empty. Conservatism means choosing the cautious end of a range that has actually been measured. An empty figure is what comes of never measuring the asset that carries the value, and reading it as caution mistakes an absence of information for a margin of prudence. A number invites comparison and an absence does not, so the mistake is worse than having no figure at all.

What is replacement cost not?

Three things, and each of them is something the figure gets mistaken for often enough to be worth naming.

Replacement cost is not a floor. Nothing in the arithmetic prevents a business from being worth less than what it would cost to rebuild, and there is a whole industry state in which exactly that happens. When an industry has built more capacity than its customers need, businesses inside it can trade well below what their plant would cost to put up, and they stay there until enough capacity leaves. A reading below one is unusual but it is not impossible, and calling replacement cost a floor turns an observation about normal conditions into a rule that fails precisely when things stop being normal.

Replacement cost is not a target. Nothing about the number says a value should move toward it, or away from it, over any period. Tobin's argument is about the incentive to add capacity across an industry over time, a statement about aggregate behaviour and not a statement about where any single figure is heading.

And it is not a measure of anything mispriced. The Rs 4,50,00,00,000 does not say the market is paying too much for these assets and it does not say the market is paying too little. The figure says a rebuild would not obtain three specific things, and that the people setting the price are attaching Rs 4,50,00,00,000 to the difference between having them and not having them. Description is where the arithmetic ends.