Independent vs Mutually Exclusive Projects
Accepting an independent project neither requires nor prevents another, so independent projects can all be taken together. Mutually exclusive projects cannot: accepting one makes the other impossible. Sankalp Industrial Systems Limited, invented, has two proposals that need the same factory floor, so only one gets built. The structure gets settled before any appraisal rule runs, and it decides whether the company is accepting projects or choosing between them.
The distinction is about the DECISION rather than about the money. Two projects whose cash flows have nothing to do with each other can still be mutually exclusive. Two projects chasing the same customers, with overlapping revenue and visible cannibalisationSales that a new project wins by taking them away from something the company already sells. between them, can be perfectly independent. A pair is exclusive when accepting one takes the other out of the set of things the company is able to do.
Exclusivity lives in the physical world. Exclusivity lives in a building, in a lease, in a licence condition, in a plan somebody drew two years ago. A forecast describes one project and says nothing about the existence of any other, so exclusivity never appears anywhere in a cash flow forecast. So the structure has to be established by asking, not by computing. The check gets skipped for exactly that reason: everything else in an appraisal has a cell to sit in, and structure has none.
What does it mean to call a project independent?
A project is independent when accepting it neither requires anything else on the list nor prevents anything else on the list. Take it and nothing else moves. Reject it and nothing else moves. Independence is a two way test, and both halves have to hold: the project must need nothing from the others, and it must take nothing away from them.
A household shows the same thing. Suppose one is considering putting a solar water heater on the roof and replacing the front door. Neither one needs the other. Neither one uses up something the other needs. Both can be done, or either, or neither, and the case for the door is exactly the same case whether or not the heater goes up. Independence feels like that from the inside: two questions that never have to be answered in the same room.
Two of Sankalp Industrial Systems Limited's five proposals are like that. Project 3 is a tooling upgrade. The upgrade costs Rs 30,00,00,000 and pays back Rs 12,00,00,000 in each of four years. Project 4, the regional warehouse, is a longer and larger thing: Rs 90,00,00,000 down against Rs 16,00,00,000 across ten. Neither one needs the other to exist. Neither one consumes anything the other needs. The tooling sits on machines the company already runs. The warehouse is on ground the company already holds, making it a brownfieldBuilt on ground a company already occupies, rather than on a fresh site bought for the purpose. build rather than a new site.
The consequence is the whole reason the word matters. When a project is independent, the only question is whether it is worth doing on its own terms, and the answer does not change when anything else on the list is examined. Judge it against the hurdle. For this company the hurdle is 12.00 per cent, its own weighted average cost of capital. If it clears, take it. The presence of four other proposals on the same sheet of paper is decoration.
What makes two projects mutually exclusive instead?
Two projects are mutually exclusive when accepting one makes the other impossible. Not harder. Not less attractive. Impossible. The test is not whether the two compete, but whether taking one removes the other from the set of things that can be done at all.
Back to the household. There is one wall in the living room and a choice between a bookshelf and a television mount. Both are good ideas. Both are affordable. The wall takes one of them. No amount of extra money creates a second wall, and no amount of persuasion makes the shelf and the mount coexist in the same square metre. Once the shelf is picked, the mount does not become expensive; it becomes impossible. Exclusivity works like that wall, and it has nothing at all to do with the price of either one.
Sankalp Industrial Systems Limited has exactly that on its list. Project 1 is a third valve line: Rs 2,00,00,00,000 down, then Rs 65,00,00,000 back in each of five years. The automation cell, project 2, needs Rs 50,00,00,000 at the start and brings Rs 20,00,00,000 back in each of the same five years. One floor in one building has to carry whichever of the two goes ahead. The building is real, the floor is already at full utilisationHow much of an asset's available capacity is actually in use at a given time., and only one of the two can be laid down on it. Approving both is not an expensive decision but an unbuildable one.
What makes two projects mutually exclusive?
Where does exclusivity actually come from?
Most people looking for exclusivity look for shared machinery, find none, and stop. The search misses three of the four places exclusivity comes from. Exclusivity has four distinct sources, and only one of them is visible on a factory walk.
The first is a shared physical resource: one floor, one crane, one bay, one furnace, one plot with one approach road. The second is two ways of doing the same job. The two proposals may share not a single nut between them, and people miss that source most often. The third is a contract or a licence that permits one arrangement and not two. The fourth is a limit on people rather than on money, where the same handful of engineers would have to run both and cannot.
| Where it comes from | What the two projects share | Why a forecast cannot see it |
|---|---|---|
| A shared physical resource | One floor, one crane, one plot, one power connection | The forecast prices what a project uses, never what it displaces |
| Two ways of doing one job | Nothing at all, except the problem being solved | Two separate asset lists, two separate lives, one purpose that appears in neither |
| A contract or a licence | A single permission that admits one arrangement | The clause sits in a legal file, and the appraisal never leaves the finance file |
| A limit on people | The same few engineers who would have to run both | Salaries are costed in each project separately and add up perfectly well on paper |
The second source is where careful people still go wrong, so it deserves its own sentence. A proposal to rebuild an ageing production line and a proposal to hand the same work to an outside supplier are a classic make-or-buyThe choice between producing something inside the company and paying an outside supplier to produce it instead. pair. One buys steel and controls; the other buys nothing and signs a contract. Their forecasts look nothing alike. Once the line is dealt with there is no second line left to deal with, so the two are still mutually exclusive.
Two proposals would each replace the same ageing production line, one by rebuilding it and one by handing the work to an outside supplier. Are they independent?
Is there a third structure the two words miss?
There is, and it turns up more often than either independence or exclusivity. A project is contingent when it can only be done if another is done first. The site purchase, then the warehouse extension on that site. The power connection, then the furnace that draws on it. A contingent pair is one decision in two stages, not two decisions that happen to be related.
Treating a contingent pair as two decisions goes wrong in a way worth stating plainly, and the damage is more than untidiness. Approving the extension on its own approves something that cannot begin: there is no ground under it. Appraising the site purchase on its own values a plot with no stated use. The number that comes out has no meaning behind it. Neither half of a contingent pair means anything alone, and the value of the pair is not the sum of two values that were each computed as though the other did not exist.
The practical handling is to appraise the sequence as one project with two outlays at two dates. The sequence is ordinary phasingSplitting one piece of work into stages that begin at different times rather than all at once., and it is what a project sheet should have shown in the first place. The reason it often does not is that the two stages were proposed by different people at different times and arrived on the list as two papers.
A warehouse extension can only be built if a site purchase happens first. What structure is that?
Why does none of this appear in a cash flow forecast?
Put the two appraisal sheets for projects 1 and 2 side by side and read every line on both. Project 1 shows an outlay of Rs 2,00,00,00,000, an inflow of Rs 65,00,00,000 a year for five years, and a value of Rs 34,31,04,532 at the 12.00 per cent hurdle. Project 2 shows an outlay of Rs 50,00,00,000, an inflow of Rs 20,00,00,000 a year for five years, and a value of Rs 22,09,55,240 at the same rate. Both are after the company's own assumed effective tax rate of 25.0 per cent. Nowhere on either sheet is there any trace of the fact that the two cannot coexist.
The absence is not an oversight in the sheets. A cash flow forecast answers the question what does this project produce, and produces a number. Asking what else the project displaces is a different question with a different shape and no natural home in a spreadsheet. There is no row for it, no rate to apply to it, and nothing to total at the bottom. Exclusivity is a fact about the world, and the sheet is a model of one project's money.
Which means the structure has to be carried by something other than the appraisal. In practice it is carried by whoever remembers. Memory is a thin thing to rest a two hundred crore decision on, and that is exactly why the check belongs on a list of its own.
Both projects 1 and 2 clear the hurdle comfortably. Before reading on: does that make the decision about them easy?
What does the structure change about the decision rule?
Structure is the pivot the whole distinction turns on. With independence, the rule is an acceptance test applied one project at a time: does this clear the hurdle, yes or no. With exclusivity, the rule becomes a ranking: both may clear, and the company still has to pick one. Structure does not change the arithmetic at all; it changes which question the arithmetic is being asked to answer.
Run it on Sankalp Industrial Systems Limited's own numbers and the difference is stark. Projects 3 and 4 are independent, so each faces the same question separately. The tooling upgrade is worth Rs 6,44,81,922 at the 12.00 per cent hurdle. The regional warehouse is worth Rs 40,35,685 at the same rate. The margin is thin and the case genuinely marginal, and the value is still positive. Both clear, both are accepted, and neither acceptance had anything to do with the other.
Projects 1 and 2 face a different question. Both clear, at Rs 34,31,04,532 and Rs 22,09,55,240, and clearing has stopped being the point. The point is which one, and on these figures the choice is worth Rs 12,21,49,291. The Rs 12,21,49,291 is the difference between the two values computed before either was rounded to the rupee. Subtracting the two rounded figures printed above gives one rupee more because each was rounded on its own. The relation, not the residue, is what matters here.
Notice what that says about the two kinds of mistake available. An independent project is got wrong by rejecting something worth having, and a mutually exclusive pair can be got wrong while both projects are approvable and the arithmetic on both is perfect. The second failure leaves no wrong number anywhere for anybody to find.
What does that do to the list as a whole?
Independence makes a decision separable. A separable decision can be settled without knowing the answer to any other. Exclusivity makes a decision joint. A joint decision has to be settled together with one or more others in one sitting. A list of independent projects is a list of separate errands, and a list containing an exclusive pair is a list containing at least one meeting.
The split settles how to organise the work rather than just how to name it, and that is worth more than it sounds. Separable decisions can be delegated, staged, and taken as the papers arrive. A joint decision cannot be split up without changing the answer: whoever settles the valve line has settled the automation cell at the same moment, whether or not they knew that is what they were doing.
The split also marks where to look for trouble. Trouble in a separable decision is contained: one wrong decision leaves the others unaffected. Trouble in a joint decision propagates. The wrong half was chosen at the same instant the right half was rejected, and by the time anybody notices, money has moved.
What does Sankalp's own list look like once it is mapped?
Sankalp Industrial Systems Limited has five numbered proposals in front of it. The map below can be drawn before a single one of them has been appraised, and drawing it that early is the whole value of it. Structure is a fact about what can be built and not about what anything is worth, so no rupee amount is needed anywhere on the structural map. The values are printed alongside only because they have already appeared above.
| Project | What it is | Value at 12.00 per cent | Its structure |
|---|---|---|---|
| 1 | Third valve line | Rs 34,31,04,532 | Exclusive with project 2, the same floor |
| 2 | Automation cell | Rs 22,09,55,240 | Exclusive with project 1, the same floor |
| 3 | Tooling upgrade | Rs 6,44,81,922 | Independent of everything on the list |
| 4 | Regional warehouse | Rs 40,35,685 | Independent of everything on the list |
| 5 | Effluent treatment plant | Minus Rs 16,74,88,849 | Neither: mandatory under the consent to operate |
Read the right hand column and the list rearranges itself. One choice, between projects 1 and 2. Two acceptances, projects 3 and 4, each judged on its own against the hurdle. And one item that is not being decided at all. Five items on the table contain three decisions, and two of the three are easy.
Which projects on Sankalp's list can be decided without knowing anything about the others?
Where does the mandatory project sit?
Project 5, the effluent treatment plant, belongs to neither structure, and the reason repays a moment of precision. Project 5 is not independent. Nothing is being weighed against it and nothing is being decided. Project 5 is not exclusive with anything either, and it removes nothing from the table. A mandatory project is an item on a list that is not on the table, and its category is not a third kind of relationship but the absence of a decision.
| Project 5, the effluent treatment plant | Figure |
|---|---|
| Outlay at the start | Rs 45,00,00,000 |
| Cost saving, each year | Rs 5,00,00,000 |
| Years of saving | 10 |
| Value at the 12.00 per cent hurdle | Minus Rs 16,74,88,849 |
Every appraisal rule turns that down, and it gets built anyway. The consent to operate makes it a condition of running the plant at all. The hurdle rate does not govern it. Project 5 needs a comparison of the available routes to compliance on cost instead. The comparison is a wholly separate exercise with a shape of its own.
Leaving it on the same sheet as the other four is what makes lists confusing. The four are candidates and the fifth is an obligation, and the format gives no way to tell them apart. Marking it as settled is a small thing that removes a whole class of argument later, when somebody asks why the company approved something with a negative value.
Change one structural fact and watch the count of decisions move
None of the controls below touches a cash flow, a rate or a rupee. Each control changes one structural fact about the list, and the map redraws to show how many decisions the five items then contain. The settings shown at the start are Sankalp's list as it actually stands.
Does a limited budget make independent projects mutually exclusive?
Sooner or later somebody says that money is the real constraint, and they are half right. A cap on investment for the year does make independent projects compete with one another. Rupees spent on one are rupees not available for another, and the competition is real. Competing is not the same as excluding, and confusing the two sends people to argue for more capital when the obstacle is a wall.
The two constraints differ in three ways worth holding on to. The two come from different places: the shared floor comes from the building and the cap comes from the board. The two respond differently to money. Raise more capital and the cap can be lifted. No amount of funding unshares a floor. And the two bite differently: the cap changes which combination of projects is best, and the floor changes which combinations are possible at all.
Choosing a set of projects under a fixed pot is capital rationingA ceiling the board puts on total investment for a period, so proposals have to fit inside a fixed pot., a real and separate exercise, and how it is done is covered elsewhere.
Which Indian body settles what, once a project stops being a list item
| The point made here | Who sets the rule | Where to read it, and it changes |
|---|---|---|
| A listed company telling the market it intends to build something | Securities and Exchange Board of India | sebi.gov.in, current text only |
| Any charge registered over the site or the machine two projects both want | Ministry of Corporate Affairs | mca.gov.in, current text only |
| Conditions travelling with money borrowed from a regulated lender | Reserve Bank of India | rbi.org.in, current text only |
| The 25.0 per cent effective tax rate sitting inside every project cash flow here | Nobody: it is the invented company's own assumption | Not a statutory figure and not offered as one |
Structure itself is settled inside the company and no outside body has a view on it. Only the commitment that follows touches any of the three named above.
Sankalp's board caps investment for the year. Does that cap make projects 3 and 4 mutually exclusive?
Where does the ranking difficulty begin?
Once a decision is a ranking rather than an acceptance, a new problem walks in that acceptance never has. Two appraisal rules can put the same pair in opposite orders. On Sankalp's exclusive pair they do exactly that: the value rule prefers project 1 and the internal rate of return prefers project 2, on the same cash flows, at the same 12.00 per cent hurdle, with no error anywhere. The company cannot escape that disagreement by taking both, and exclusivity is what makes it matter.
Notice the shape of that sentence. If projects 1 and 2 were independent, the two rules could disagree all day and nothing would turn on it. The answer would be to take both, and the ordering would be a matter of curiosity. The moment only one can be built, the ordering is the decision. Structure is what promotes a difference of opinion between two measures into a Rs 12,21,49,291 question.
Handling that disagreement, why one of the two rules is the defensible one, and the rate at which their answers cross are all covered separately. Exclusivity is why the disagreement cannot be sidestepped at all.
How does a practitioner actually use this?
Three people read the same list and each of them uses the structure differently. Structure is not an academic classification. Structure is the first thing a careful outsider looks for and the last thing an insider thinks to write down.
A lender assessing a term facility for the valve line wants to know whether the borrower's own plan is coherent. If the same floor appears in two capital sanctionThe formal approval that releases money against a project, usually given by a board or a committee. notes, the lender is being asked to fund a build that cannot happen as described, and the covenant testing behind it rests on cash flows from an asset that will not exist. A lender who spots that asks one question and saves a year.
An analyst modelling the company from outside cannot see the floor at all, and should therefore be sceptical of a capital expenditure line built by adding up every announced project. Announcements are made project by project. Some of those projects are alternatives to each other, and the sum of the announcements overstates the spend and overstates the capacity that will result from it.
A board member has the easiest job and the one most often skipped. The question at the table is not whether each item clears. The paper already says so. The question is which items on this list cannot both happen, and it takes about ninety seconds to ask. A room with no answer to it is looking at a list that is not ready, whatever the arithmetic on it looks like.
A household version of the same discipline: before pricing a holiday and a new fridge, ask whether the two want the same fortnight, the same delivery slot or the same person to be home. Most of the time the answer is no and both go ahead. The one time it is yes, finding out in advance costs nothing and finding out afterwards costs the deposit.
The error that gets made, and what it costs
The failure is assuming independence by default, and almost nobody does it deliberately. A list arrives as a list. A list looks like a set of separate items. Separateness is what a list is for. Each item comes with its own paper, its own forecast and its own appraisal, and no document in the pack mentions any other document in the pack.
So the analyst appraises five projects, finds that four of them clear the 12.00 per cent hurdle, and recommends four. Projects 1 and 2 are both on that recommendation, and one floor is meant to carry both of them. The board sanctions Rs 2,00,00,00,000 for the valve line and Rs 50,00,00,000 for the automation cell in the same sitting, and every figure behind both approvals is correct.
The next stage is worse than a wasted approval. The constraint surfaces during commissioningThe stage at which a finished asset is tested, handed over and switched on for normal use., when somebody with a tape measure asks where the second line is going. One project is stopped after money has already been committed to it. The amount already spent then becomes an argument for continuing with the wrong one, and that argument is hard to kill because it feels like thrift.
The reason it happens is structural rather than careless: the exclusivity is a fact about a building and the appraisal is a document about cash flows, and the two live in different files with different owners. Nobody checked. Checking is not a calculation, and the sheet had no cell for it.
The defence is a question rather than a technique. For every pair on the list, the question is what physical thing, permission, site or team both of them need. If the answer is anything at all, the two are not independent, and no figure in either appraisal would ever have shown it.
Five separate project papers arrive, each with its own forecast and its own appraisal. What has that format already assumed?
How is the structure established before any figure is computed?
The check is four questions asked of every pair on the list, and it is deliberately not a calculation. Structure is established by asking, so it needs a place on the agenda rather than a cell on the sheet.
| Ask of every pair | What a yes means |
|---|---|
| Do these two need the same site, floor, machine or connection? | Mutually exclusive, and no funding decision changes that |
| Are these two different ways of solving one problem? | Mutually exclusive, even though nothing physical is shared |
| Does one permission, licence or lease cover both? | Check the clause before either appraisal is believed |
| Would the same handful of people have to run both? | Treat as exclusive until somebody names who runs the second |
| Can one of these only start after the other finishes? | Contingent: appraise the pair as one sequence |
Two habits make the check stick. Ask it of pairs rather than of projects. Exclusivity is a property of a pair, and a project on its own has nothing to be exclusive with. And ask it before the appraisals are read rather than after. Once a set of values is on the table, the conversation is about the values and nobody goes back.
The output is a one-sheet map like the one drawn above: items grouped into decisions, with the settled items marked as settled. The map takes ten minutes for a list of five, and it would have taken ten minutes for the list that approved two builds for one floor.
What is the question to ask of every pair on a list before appraising any of them?
Classify three pairs in order: two projects needing the same crane; a project that can only start once another finishes; two projects sharing nothing at all but bidding for one year's money.
Where the rules behind these figures come from
| Used for | Source | Where |
|---|---|---|
| Reading a list of projects before pricing any of them | Aswath Damodaran, corporate finance teaching material | pages.stern.nyu.edu |
| Treating an investment programme as one set rather than as many papers | Koller, Goedhart and Wessels, Valuation | named by title |
| Every rupee figure, every project and the 12.00 per cent hurdle | The invented case record behind these notes | restated in full above |
Sankalp Industrial Systems Limited, its five projects and its factory floor are invented.
Educational material. Not advice on any investment, tax, budget or market position.