How Market Price, Value and Expectations Interact

A share price records what one buyer and one seller agreed, and it carries the assumptions that buyer holds about the future. Value is an estimate a person builds from evidence. Expectations sit between the two. A price therefore moves when assumptions change rather than when results arrive. A good year can be followed by a falling share price.

Three words get used as though they were the same word. A price is quoted on a screen. A value is worked out by somebody. Expectations are what the first of those two contains about the future and the second one has to argue with. Almost every confused sentence written about shares comes from sliding between the three without noticing, and the confusion is not a beginner's problem: it survives into professional work, where it appears as the belief that a number can be right in a way that obliges a market to agree with it.

Everything under this heading rests on work already done elsewhere: what an equity share confers and what it does not, and where a holder stands in the queue when the money runs out; what separates analysis from an opinion, namely that somebody else can walk the route and stop at the sentence where they disagree; what a research note is for, and how a peer set gets assembled. Valuation method sits in the layer beneath this one, covered separately and applied here rather than rebuilt.

A price can be read as a set of assumptions rather than as a verdict on a company. The move changes the question from one nobody can answer to one that can actually be worked on.

One caution belongs at the top rather than at the bottom. The subject drifts, and it drifts in one direction. A reader who has just understood the gap between a price and an estimate wants very badly for the next sentence to be a verdict about whether some share is dear or cheap. Naming an assumption says nothing about whether the assumption is right, so reading what a price assumes never produces that verdict. Every figure below belongs to Sarvani Coatings Limited, an invented paint maker whose numbers were constructed to tie.

What is a share price actually saying?

Start with the smallest true statement and refuse to add anything to it until it has earned its place. A share price is the record of one completed transaction: a quantity of shares changed hands at that number, between one buyer who wanted them more than the cash and one seller who wanted the cash more than the shares. That is the whole of it. Everything else people believe a price contains is a layer they added themselves.

Take it out of finance for a moment. A flat in a building of ninety flats sells for a certain sum. The next morning every owner in that building has a figure for what their own flat is worth, and it is the number that flat down the corridor fetched. Nothing happened to the other eighty nine flats. No survey was done, no one asked the other owners anything, and the buyer who set the number had never seen inside most of the building. One household needed to move cities, one household had saved enough, and the two of them agreed a figure. The agreed figure then got applied, silently and by everybody, to eighty nine flats nobody bid for.

A share price does the same thing on a much shorter cycle. On the stated date used throughout this guide, Sarvani Coatings Limited's shares changed hands at an illustrative Rs 486/-, and that number was then stretched across all 24.00 crore shares in issue, putting the whole company at Rs 11,664 crore. The trades that produced it were a small fraction of the whole. The number they produced was applied to every share, including the 52.4 per cent held by the promoter group. The promoter block was never offered, and would not have fetched Rs 486/- if it had been.

So the first correction to make is about scale, and it is easier to feel with the arithmetic set out than to accept as an assertion. Of the 24.00 crore shares in issue, 47.6 per cent is free floatWhatever is left of the share count once the blocks that never reach the market, a promoter group's above all, have been taken out. Free float is covered in full under market data., which is 11.424 crore shares. An average day sees about Rs 42 crore of value traded. At Rs 486/- that is about 8.64 lakh shares. About 32 per cent of that is taken to deliveryThe part of a day's trading that actually settles into somebody's holding, rather than being bought and sold again within the same session. The mechanics are covered under market infrastructure., which is about 2.76 lakh shares.

Worked against the float, that last figure gives a number worth sitting with. About a quarter of one per cent of the freely traded shares actually changes hands for keeps on an average day, and that sliver sets the number that gets written against all of them. Precisely, 2.76 lakh shares against a float of 1,142.4 lakh shares is 0.24 per cent of the float, and 0.12 per cent of every share in issue. The other 99.88 per cent of the register simply inherits the figure.

None of this makes the price useless or dishonest. A price is extremely good information about one thing: the terms on which shares actually moved. About almost everything else it is poor information, and the trouble starts when a reader promotes it from the first job to the second.

Between them, three promotions account for most of the confused writing on this subject, and each is worth naming. The first treats a price as a poll. Nobody asked the holders anything. A shareholder who believes the shares are worth far more than Rs 486/- simply does not sell, and their disagreement leaves no mark on the number at all. The quiet conviction of the people not trading is invisible by construction. The second treats a price as an average. A price is not an average of the day's trades either: it is the most recent one, and each trade replaces its predecessor rather than being added into a mean. The third treats a price as something the company said. Sarvani Coatings' board has no ability to set it, may privately think it wrong in either direction, and finds out what it is the same way everybody else does.

Where does value come from, and who produces it?

Value is not observed, it is estimated, and every estimate has an author whose assumptions are part of the answer. That sentence sounds like a hedge and it is the opposite of one. Naming the author points straight at the place two estimates actually part company. The arithmetic is usually sound. The disagreement lives in the assumptions underneath it.

Think about how a household values a plot of land inherited three generations back. One person prices it against what the neighbouring plot fetched last year. Another prices it against the rent it could earn if a shop went up on it. A third prices it against what the same money would do in a fixed deposit. None of the three is being silly, and none of them can be shown to be wrong by looking harder at the plot. The disagreement was never about the plot but about which future to price.

Valuation method is covered separately and applied here rather than rebuilt. The shape of the exercise matters more than its machinery. Any method, whatever the arithmetic, takes published figures that are fixed and combines them with assumptions that are not. Sarvani Coatings' year three statements fix that revenue was Rs 2,415 crore, that earnings before interest, tax, depreciation and amortisation (EBITDA) was Rs 446 crore at an EBITDA marginOperating profit before depreciation, interest and tax, taken as a percentage of revenue. How it is built from the profit ladder is settled in the accounting layer. of 18.47 per cent, and that profit after tax was Rs 278 crore. The statements fix nothing at all about how long that margin lasts, how fast volume grows after this year, or what return an investor should require for bearing the risk of finding out.

So two competent analysts working from that identical set of statements will land in different places, and the reason is not that one of them slipped. The statements ran out before the question did. Meghna Iyer might read the two point gross margin gain as a durable change in what the company can charge, and price it as a level that holds. A colleague reading the same three years might read it as a pricing window that happened to open across the whole field and will close when input costs catch up. Both readings are consistent with everything published. The evidence available today does not choose between them, and pretending otherwise is the only actual error available here.

The honest output of the exercise is therefore a range with its assumptions attached, not a number, and the width of the range is information rather than embarrassment. A narrow range says the answer barely depends on what is assumed. A wide one says it depends enormously, and points precisely at the assumption to go and investigate. Collapsing that width into a single figure to look decisive throws away the most useful thing the work produced.

What sits between a price and a value?

Set the two side by side and something is obviously missing. A price is a fact about the past: a trade happened. A value is a claim about the future: this is what the evidence supports. Nothing yet connects them, and without a connection there is no saying why a price would ever move, or why it sits where it does rather than somewhere else.

The connecting layer is expectations: the assumptions about the future held by whoever was willing to trade at that number, and it is the only one of the three that is never written down anywhere. The buyer who paid Rs 486/- was not paying for the year already reported. The buyer was paying for a future they had in mind, and the price is the only trace that future left behind.

Which buyer, though? Not every holder, and this distinction does the real work. The people who matter for the number are the ones actually willing to transact at the edge of the market on that day: the last buyer who agreed to pay and the last seller who agreed to accept. A holder sitting on shares bought years ago at a quarter of the price has views, and those views do not enter the number unless they act on them. The price reflects the assumptions of whoever was at the margin. On any given day that is a small and shifting group, and its composition can change while the company does not change at all.

Here is what makes this layer genuinely hard rather than merely abstract. Sarvani Coatings publishes a great deal. The company publishes three years of statements that tie, a segment split, a shareholding pattern, cash flows, the notes behind them. Both exchanges publish the price continuously, along with traded quantity and the quoted spread. Everything on both of those lists is a matter of record. The assumptions inside the price are on neither list, are published by nobody, and cannot be requested from anyone. Inferring them from the number itself is the only route open.

The absence is not a complaint about disclosure, and no rule could fix it. There is no register of assumptions to file because assumptions are not held by the company. Assumptions are held, separately and privately, by every person who might trade the shares, and they change through the day as those people change.

Notice what follows for how a price behaves. If the price sits on the assumptions of the people at the margin, then the price moves when those assumptions move. Results do not move a price by arriving; they move it by differing from what was assumed. A company can publish an excellent year into a falling share price, and a poor year into a rising one, without anything irrational happening in either case. The mechanism is worth working through slowly.

First, though, the plainest possible arithmetic statement of how much of a price is expectation. Sarvani Coatings earned Rs 11.58/- a share in year three. The shares changed hands at Rs 486/-. Divide the first by the second: the year just published accounts for 2.38 per cent of what a buyer paid, the earnings yieldA year's earnings per share divided by the price per share, which is the price to earnings ratio turned upside down. Worked through under common stock. arrived at under common stock. Turned around, 97.62 per cent of the price refers to years that have not been reported, do not yet exist, and are entirely a matter of assumption.

What does the published position actually look like?

Everything so far has been mechanism. A figure seen being assembled is harder to misuse than one handed over finished, so the whole worked position goes on the table now.

Every number below belongs to Sarvani Coatings Limited. The quoted Rs 486/- is assumed rather than observed, and dated to the day it was set down; a live quotation would sit wherever the last trade left it.

Two things in that table are places where readers reliably slip. Both are worth stopping on.

The first is the rounding, and it matters more than it looks. The price divided by the published Rs 11.58/- gives 41.97 times. Divided instead by profit after tax over the share count, Rs 278 crore over 24.00 crore, or Rs 11.5833/- before anyone rounds it, it gives 41.96 times. Both round to 42.0 times, and 42.0 is the figure used below. The two are not the same calculation, and a decomposition that has to reconcile exactly will not do so if the rounded per share figure is quietly substituted for the unrounded one. The general rule underneath it is worth carrying away: never rebuild a published figure out of rounded readings, because the answer that comes back looks derived and is therefore trusted, and it is wrong.

The second is that enterprise valueThe value of the whole business rather than of the shares alone: market capitalisation with borrowings added and cash taken out. Built up under securities and valuation. here sits below market capitalisation, which is the reverse of what most readers expect. The reason is that net debtTotal borrowings less cash and liquid investments. When cash is the larger of the two the figure is negative, and the company is described as being in net cash. is negative: cash and investments of Rs 312 crore exceed borrowings of Rs 240 crore, so the adjustment is minus Rs 72 crore and it subtracts rather than adds. Six tenths of one per cent is a small difference and the direction is the whole lesson.

Why can a company report well and its shares fall?

Most people meet this question looking for a conspiracy, and there is no conspiracy in it. There is one mechanism, it is simple, and once it is held a large share of the price movements that look irrational from outside stop looking that way.

The price was already carrying an assumption about the result, so the result gets measured against that assumption and not against last year. Everything else follows from that one sentence.

The everyday version is a school report. A child who has been scoring in the sixties comes home with a seventy eight, and the house is delighted. Another child who has been scoring in the nineties comes home with the same seventy eight, and the evening is difficult. The mark is identical. The two households differ only in what they had already assumed, and nobody in either house thinks the arithmetic of marks has been suspended. A share price is the second household, permanently, with the assumption written in a number instead of in a habit.

Now make it concrete with figures that tie. Sarvani Coatings' year three was strong on every line of the ladder, and each line grew faster than the one above it: revenue up 13.9 per cent, gross profit up 19.1 per cent, EBITDA up 31.2 per cent and profit after tax up 41.1 per cent. Read against year two alone, that is close to the best shape a set of statements can have.

Against what was assumed, it reads differently. Nine forecasts of profit after tax were already on paper when the year closed. Their average sat at Rs 268 crore, the lowest of them at Rs 255 crore, the highest at Rs 284 crore. Sarvani Coatings then reported Rs 278 crore. Held against the average, that is 3.7 per cent more than looked for. Held against the spread of the nine, Rs 278 crore sits roughly four fifths of the way up a band those forecasters had already committed themselves to, and the second reading says something quite different. An outcome landing within limits people wrote down in advance carries almost no information, whatever language greets it. The same holds on earnings per share, where a mean consensus estimateThe average of the published forecasts several analysts have made for the same figure, along with the highest and lowest of them. How it is compiled and used is covered under earnings. of Rs 11.17/- met an actual Rs 11.58/-.

Now push it one year forward, and label what follows clearly: the year four figures below are hypothetical, and Sarvani Coatings has published nothing of the kind. Suppose the company reports another year of 13.9 per cent revenue growth. Hold the margin flat and hold everything below the operating line still, meaning no change in depreciation as a proportion, in finance cost, in the tax rate or in the share count. On those assumptions, earnings per share rises 13.9 per cent too, from Rs 11.58/- to about Rs 13.19/-.

Carrying a revenue growth rate across to earnings is exactly the step that goes wrong silently, so the bridge has to be stated rather than assumed. Revenue and earnings grow at the same rate only if margin holds and nothing beneath it moves. In Sarvani Coatings' actual year three neither of those held: margin rose two points and profit grew 41.1 per cent on revenue up 13.9. So the illustration below is a deliberately flat case, and saying so is the whole of the discipline.

A price of Rs 486/- against Rs 13.19/- is 36.8 times, and a reader who had assumed the price contained 13.9 per cent growth would find nothing to react to. But suppose the buyer at the margin had been carrying an assumption of 20 per cent. A buyer carrying that assumption needed Rs 13.90/- and got Rs 13.19/-, a shortfall of 5.1 per cent. The report improved on the year before by 13.9 per cent and fell short of the assumption by 5.1 per cent, at the same instant, from the same sheet of paper.

Nothing about the result was bad, and a falling price on that day is a repricing of the assumption rather than a verdict on the year. Two things changed places without either of them being wrong: the company delivered growth, and the market discovered that its assumption had been too high. The price movement says nothing about the business, and everything about what people had been assuming.

The same mechanism runs in reverse, and it is the half people forget. If the assumption priced in had been worse than what arrived, a company can report a fall in profit into a rising share price. The reversal looks perverse and is the identical arithmetic pointed the other way.

One more version of this catches out even careful readers. Sarvani Coatings' management gave guidanceA company's own published indication of what it expects on some measure in a coming period. The weight guidance carries, and how it is tested, are covered under earnings. at the start of year three: high single digit volume growth, with gross margin expected to hold near the previous year's level. Volume came in at 6.0 per cent, at or just under the bottom of that range. Gross margin rose two points against a guide of holding. So the company missed on the measure it had chosen to emphasise and beat on the one it had been cautious about, and which of those a price responds to depends entirely on which one the buyer at the margin had been weighting. There is no rule that settles it in advance, and looking for one is a way of avoiding the actual question.

How is what a price already assumes worked out?

Here is the move this whole guide exists to hand over, and it is a reframing rather than a technique. The instinct, on meeting a price, is to compute what the shares are worth and compare. The comparison is unanswerable in any final way, for the reason already given: there is no single value, and any one estimate is one of many that the evidence permits. The exercise then ends with a figure whose authority cannot be established, and the analyst is no better off than at the start.

Run it backwards instead. Take the price as given, and ask what would have to be true for it to make sense. The question nobody can settle becomes a short list of assumptions somebody can test. The exercise is not deciding whether the price is right, but reading what the price says.

The everyday shape of this is familiar. A shop in a busy market changes hands for a sum that seems high, and rather than arguing about what a shop is worth, the question becomes what the buyer must be assuming: that the rent will not jump when the lease renews, that the road works finish, that the footfall in that lane holds. The buyer's assumptions give three things to go and check, none of which required anybody to decide what a shop is worth. The price showed where to look.

Apply it to Sarvani Coatings. The price of Rs 486/- is 42.0 times the year just published, and only 2.38 per cent of it is covered by that year. So the assumption inside it concerns earnings that have not happened. Which earnings, and resting on what?

Any assumption about repeating year three's profit has to rest on where that profit came from. Work back through it. EBITDA rose from Rs 340 crore to Rs 446 crore, an increase of Rs 106 crore. Split that increase into its two parts. Had gross margin stayed at year two's ratio, the year three revenue of Rs 2,415 crore would have produced about Rs 1,063 crore of gross profit against the Rs 1,111 crore actually reported, so roughly Rs 48 crore of the increase came from the margin gain alone. The remaining Rs 58 crore is everything else: more revenue at the old margin, less the Rs 72 crore by which employee and other expenses rose. The two parts sum to Rs 106 crore exactly.

So a little under half of the year's EBITDA increase rests on a two point gross margin gain, and that gain is precisely the part the published statements cannot attribute. The earlier decomposition shows why. Across that single year the materials line moved down two points of revenue, from 56.0 to 54.0. Underneath the ratio, volumes ran 6.0 per cent higher, each unit sold fetched roughly 7.5 per cent more, and each unit cost roughly 3.6 per cent more to make. Inputs grew dearer rather than cheaper. Realisation climbed faster than cost, and that is what opened the margin. Three stories fit the shape equally well: a price move that swept the entire field, a price move the company made on its own account, or a tilt in the sales mix towards industrial. The sector evidence narrows the field: all three makers gained margin over the two years, and the segment data rules mix out by arithmetic. Two explanations are left, the field wide environment against the company's own pricing, and nothing published separates them.

The assumption is now laid bare, and that is the output. The price of Rs 486/- is carrying a view about whether a margin of 18.47 per cent is a level the company can hold or a peak it happened to reach, and it is carrying a second view about whether volume growth of 6.0 per cent continues. Neither of those is settled by anything published. Both can be tested against evidence that will arrive.

Before the simulation, the reading it reproduces, written out so it survives without it. Take the published Rs 11.58/- a share and carry it forward five years at 13.9 per cent, the rate revenue actually grew in year three, holding margin flat so that revenue growth and earnings growth are the same number. Earnings per share reaches about Rs 22.20/- in the fifth year. An unchanged price of Rs 486/- against that figure is 21.9 times, rather than the 42.0 times it is against the year just published. The five years of earnings between here and there add up to about Rs 87.01/- a share, or 17.9 per cent of the price. Even five years of double digit growth, all of it delivered, accounts for under a fifth of what a buyer pays today.

Running the price backwards required no discount rate, no choice between methods, and no single figure to defend against anybody. The output is not a conclusion at all but two written sentences, each naming the evidence that would settle it, and another person can pick those up, disagree with them, and take them further. A multipleA price expressed as how many times some figure it represents, such as 42.0 times earnings. The arithmetic and the choice of denominator are covered under valuation. is not a fact about a company: it is an assumption about its future written in the form of a price.

What actually closes the gap, and what does not?

Suppose the work has been done honestly and the estimate that results sits below the price. The natural next thought is that the price will come down to meet it. Nothing in the mechanism supports that thought, and the belief that it does costs more than any other error in the subject.

Go back to what a price is. A price is made by whoever is willing to trade, so it moves when what those people assume moves, and for no other reason. An analyst's estimate is a fact about that analyst's spreadsheet. The estimate has no route into the number unless it changes what somebody trading believes, and there are only a few such routes.

Evidence can arrive that nobody had. A result can land against what was assumed, the mechanism worked through above. A disclosure can change what is visible: if Sarvani Coatings began reporting segment margins separately, a question that is currently unanswerable from the outside would become answerable, and everybody's assumptions would move at once. The set of things it gets compared against can shift, and the price changes without anything happening to the company at all. Each of those works by changing other people's assumptions, and being correct is not among them.

Two consequences follow, and both are unwelcome on first meeting.

The first is that a gap does not close because it is large. There is no restoring force in the mechanism, no tension that builds. A gap can widen for years while the analysis that identified it stays correct the whole time, and the analyst who assumed that a wide gap must close soon has borrowed a law of physics that markets do not contain.

The second is that a gap can close for causes wholly unconnected to the reasoning behind it. Suppose the price does fall and the estimate turns out to have been the better guide. A raw material may have moved, a shareholder with a large holding may have needed cash, or attention may have shifted somewhere else entirely. Being right and being right for the stated reason are separate events, and only the second says anything about the method. Judging the work by the outcome rather than by the route is a specific failure named and set aside under research discipline; it applies with full force here.

Benjamin Graham's margin of safety belongs here, and his name is not detachable from it. His point was not that a gap will close, and reading him as though it were is the common misuse. His point was narrower: an estimate rests on assumptions that can be wrong, so the size of the gap required before acting is a way of surviving being wrong, and the gap makes no promise at all about when or whether anything corrects. The gap is protection against the analyst's own error, and not a prediction about anybody else's behaviour.

Once that is held, one more thing becomes visible. A view that needs some particular event to happen before other people notice it is not really a view about a company. Such a view is about other people, held with a company attached, and it should be described that way when it is written down.

Do prices already reflect what is known?

The question is unavoidable at this point and it has an author, so name him rather than describing the idea anonymously. Whether a price has already absorbed everything the public can find out about a company, and if so how promptly, is the question at the centre of Eugene Fama's work on market efficiency, and his name belongs with the idea wherever it is used. The literature it produced, the arguments made against it and the evidence on both sides are large enough to need their own treatment, and they are covered separately.

The general answer changes less about the day's work than it appears to.

Suppose the strongest version were true and prices did absorb everything public immediately. The future is not public information, so absorbing everything public would still not turn a price into a value. Efficiency of that kind says only that everything knowable and published is already in the number, a statement about the inputs rather than about whether the assumption built on them will hold. The buyer who paid Rs 486/- had access to every published figure in this guide and still had to decide whether an 18.47 per cent margin was a level or a peak, and no amount of information absorption decides that for them.

Suppose the weakest version were true and prices were routinely slow or wrong. The analyst gains nothing there either. The work would still be to identify which particular assumption in which particular price was mistaken and to demonstrate it from evidence, exactly the work in hand anyway.

So the practical position is narrower than the theory and considerably more useful: whatever the general answer, the task is to find the specific assumption inside a specific price and test it against evidence. That task is the same size on Monday whichever way the argument came out over the weekend.

How is all of this actually used?

Three steps, and the third is the one that takes discipline.

The price is read as a set of assumptions rather than as a verdict. Against Rs 486/- and 42.0 times, the useful question is not whether that number is correct. The claim inside the number is the useful question, and the arithmetic above answers it: 2.38 per cent of the price is covered by the year reported, so the rest is a claim about years that have not happened.

The assumptions are written down, in sentences somebody else could argue with. An assumption left unwritten stays comfortably vague, and cannot be tested or attacked. Most work fails quietly at exactly that point. Written out, it commits its author to something. For Sarvani Coatings the two are on the card below, and next to each sits the evidence that would settle it.

Then test the one that carries the most, and stop. Of the two, the margin assumption carries more. A little under half of the year's EBITDA increase rests on the two point gain, and the sector evidence has already narrowed the explanations without closing them. So the errand is specific: four more quarters of gross margin, read against input cost per unit rather than against the margin alone.

Two written assumptions with the evidence that would settle each one is the finished output rather than a partial one. It feels incomplete because it does not end in a verdict, and the feeling is worth examining rather than obeying. A verdict would be one person's assumptions compressed into a figure that displays none of them. The list displays all of them, and somebody else can disagree with a line of it instead of with the whole thing.

Reading a price this way yields no rating, no valuation and no target of any kind. None of those follow from naming an assumption; what a target would additionally require is set out under targets for a share price. The method builds the material a view rests on and stops just short of the point where the view would be stated.

Who outside a research desk reads a price this way?

Sarvani Coatings Limited, Nandivarman Paints Limited, Kesaria Surface Solutions Limited, Thottam Chemicals Limited, Meghna Iyer and Ravindra Setlur are invented.
Educational material. Not advice on any investment, tax, budget or market position.