Fair Value vs Market Price: Two Different Objects

Fair value is an estimate a person produces from evidence and assumptions. Market price is a record of what two parties agreed. Estimate and record are different kinds of object, produced by different processes, and neither is a corrected version of the other. Expecting them to converge on a timetable is where most disappointment about this subject comes from.

Underneath that sits a familiar distinction. A price is a thing that happened. A value is a thing somebody thinks. A market price is a transaction record carrying a timestamp, and the timestamp is not decoration. Valuation methods are covered separately.

What is fair value, with no mention of price at all?

Fair value is an estimate of what a security is worth. A person produces it. The person picks a method, gathers the evidence they think matters, writes down the assumptions the method needs, and runs the arithmetic. The number that falls out the bottom is entirely downstream of those choices, and it carries them with it whether or not anybody writes them down.

A flat two doors down the corridor makes the point. Three brokers asked what it is worth give three answers, all confident, all in rupees, all different. One of them is comparing it to a sale in the next lane. One is starting from the rent it could fetch. One is starting from the builder's asking price for a new flat half a kilometre away. Nobody is lying and nobody has made an arithmetic mistake. The three brokers chose different evidence and different assumptions, so they produced different numbers. Every one of those numbers is a fair value estimate.

Two competent people working from the same published statements and different assumptions will produce different numbers, and neither of them has made a mistake, so there is no single fair value. A spread of estimates is not a defect that better analysts fix. The spread is what an estimate is. An estimate is a conclusion attached to a set of premises, and change a premise and the conclusion moves with it.

The useful consequence is that an estimate carries its own provenance. Because a person made it, that person can be asked how. Which years were used? Which assumption was made about the cost line? How much growth was carried, and for how long? Every step is exposed and every step can be attacked. Exposure is not a weakness of the estimate. Exposure is the most valuable thing about an estimate, and a price has none of it.

What is a market price, with no mention of value at all?

A market price is the recorded amount of the most recent transaction in a security. Two parties dealt with each other at a moment, the amount they settled on was written down, and that written amount is the whole content of the number. There is exactly one at any instant, and it belongs to that instant only.

A vendor outside an office gate sold a plate of food at twenty to two in the afternoon for Rs 60/-. Sixty rupees was the price. The figure is not a claim about what food is worth, not the vendor's opinion of his own cooking, and not an average of anything. The figure records one exchange between two people, one of whom was hungry and short of time. Had either of them been asked to justify Rs 60/- as the correct figure, neither could have, and neither was trying to.

A market price does not assert anything, it records something, and the difference between asserting and recording is where every confusion about the two numbers starts. Nobody voted on it. No committee approved it. The two parties were not describing the security to anybody. Each was transacting for reasons of their own, and those reasons may have had nothing to do with the security and everything to do with a redemption one of them had to fund that afternoon. The last person willing to deal at the level, the marginal buyerThe buyer whose willingness to deal at a level is what holds the level up. If that one buyer steps away, the next print happens somewhere else., sets where the next print lands, and that person changes through the day.

Who makes each number, and by what process?

The two production lines sit beside each other. One number is produced by reasoning. The other is produced by transacting. Everything else follows from that single split.

An estimate has a route that can be walked, and a price has no route at all, only a record, and that is the deepest difference between the two. An estimate can be taken apart at any joint. A reader can accept steps one through three, refuse step four, substitute a different assumption, and rebuild it. Two readers can do that to the same estimate and arrive at different places, and both can explain themselves.

None of that can be done to a price. There is no step four. There is no author to question, no method to substitute, no assumption to swap out. Asking a price why it is Rs 486/- is like asking a photograph why it was taken. The photograph records that light fell on a sensor at a moment. The photograph contains no reason, and staring harder at it will not produce one.

Why can the two sit far apart for years?

A price changes when the people transacting change their assumptions. An estimate changes when the person holding it changes theirs. Transacting participants and estimating analysts are two separate populations, doing two separate things, on two separate schedules, and nobody has arranged for the schedules to line up.

Nothing connects the timetable on which a price moves to the timetable on which an estimate moves, so a gap between the two can persist for years without either side being wrong in any checkable sense. An estimate may be revisited twice in three years, when a results season hands its author something genuinely new. The price prints thousands of times over the same three years, on flows, redemptions, a large holder rebalancing, a rumour, an afternoon with no news at all.

People expect convergenceTwo separate numbers drifting towards each other over time until they meet. Widely assumed, though no mechanism described here would produce it. because the word is comfortable and because in most other settings two numbers about the same object do eventually agree. There is a fact of the matter about how many sacks of cement sit in a yard, and both counters are trying to find it, so two people counting agree by the end of the afternoon. Here there is no fact of the matter for the estimate to be right or wrong about, so there is nothing for the two to converge on.

Does one of them correct the other?

No. Not slowly, not eventually, not in the long run. People believe the two directions for different reasons, so take the directions separately.

A price is not an estimate and therefore not a rival claim about the same thing, so a price is not evidence that an estimate is wrong. A price says two people dealt at a level. The record says nothing about what the security is worth, and a statement that says nothing cannot contradict one that says something.

An estimate is not evidence that a price is wrong either, and this direction is the harder one to give up. A price is a record of something that happened. A record is not the kind of object that can be wrong about itself. A record can be stale, it can be thin, it can be one small trade at the end of a quiet afternoon, and all of those are worth knowing. None of them makes it incorrect.

People reach for both corrections constantly. The two numbers are denominated in rupees and printed to the same number of decimals, so they look like measurements of the same quantity. Shared units are a powerful illusion. The weight of a parcel and its worth to its owner are both numbers about the same parcel. The units differ and the difference is visible, so nobody expects one to correct the other. Here the units are identical, so the category difference is invisible, and the mind quietly supplies a relationship that is not there.

The gap has a name. Benjamin Graham named the deliberate cushion an investor keeps between an estimate and a price the margin of safetyThe cushion an investor deliberately keeps between what they estimate and what they pay, so that being wrong about an assumption is survivable. Named by Benjamin Graham.. The margin of safety is a rule about how much room an investor insists on before acting. The rule is not a forecast that the gap will close, and it was never sold as one.

Which number does which job?

Both are useful. Neither is a general purpose number. A market price is the number to reach for wherever a real transaction amount is required: what a holding could be sold for today, what it is carried at on a statement, what a fresh raise would be priced against, what a pledged shareholding would be liquidated at. Every one of those questions contains a transaction inside it, so only a transaction number can answer it.

An estimate is the number to reach for wherever a judgement is required and nothing has to be transacted at all. Should this be held? Is the cost line likely to behave? Which conditions have to hold for this to work out? None of those questions contains a transaction, so a transaction number has nothing to say about them.

Using one where the other is required is a category errorUsing a thing of one kind to answer a question that only a thing of another kind can answer. Not a small mistake that more care would shrink. The answer was never available from that direction. rather than an inaccuracy, which is why no amount of extra care improves the answer. An inaccuracy gets better with more work. Measure the room again with a proper tape and the figure improves. The number was never capable of answering the question in the first place, so a category error does not get better with more work. Asking a price how much something is worth to hold, and then being careful about it, produces a careful non-answer.

What do the two objects look like side by side on one company?

Sarvani Coatings Limited, an invented maker of decorative paints and industrial coatings, reports the figures that follow. Nothing on the record side requires anybody to think, so start there.

On the stated date the illustrative price is Rs 486/-. Sarvani Coatings has 24.00 crore equity shares, so the market capitalisation is Rs 11,664 crore. Sarvani Coatings holds cash and investments of Rs 312 crore against total borrowings of Rs 240 crore, so net debt is minus Rs 72 crore and the enterprise valueMarket capitalisation plus net debt. A company holding more cash than borrowings has negative net debt, and its enterprise value sits below market capitalisation. is Rs 11,592 crore, which sits below the market capitalisation rather than above it.

Notice what did and did not happen there. Nobody formed a view. Every one of those lines is the same record restated against a different published figure, and two readers who disagree about everything else would still produce that table identically. A record behaves exactly like that.

Now the estimate side, worked to show the process rather than to reach a number. Any estimate of Sarvani Coatings runs almost immediately into one question. Does the gross marginRevenue less the cost of materials, expressed as a percentage of revenue. Gross margin is the first profit line on the ladder and the one closest to what a business pays for its inputs. of 46.0 per cent hold?

The margin reached 46.0 per cent from 43.0 per cent two years earlier, as the cost of materials fell from 57.0 per cent of revenue to 54.0 per cent. The published statements record that move precisely and cannot say a word about why it happened. A pricing environment that let every maker price ahead of its input costs would produce that line. So would Sarvani Coatings pricing better than its peers. So would a shift in the sales mix towards industrial work. The statements do not separate the three.

Two people can hold opposite views on that one question, produce estimates far apart, and both be working properly. Meghna Iyer reads the mix shift as structural and carries 46.0 per cent forward. The reader at the next desk reads it as a window that closes and carries the cost of materials back towards 56.0 per cent. Same statements. Same evidence. Different assumption at step three, and two estimates that need not be anywhere near each other.

What follows when the two disagree?

The practical output is smaller and more useful than the one most readers arrive hoping for. The price of Rs 486/- and any estimate of Sarvani Coatings differ because of a disagreement about that margin. Naming the disagreement is the finding.

1. Write down what the price appears to assumeAt 42.0 times published earnings, something about the earnings line has to keep behaving. Say in one sentence what that something is.
2. Write down what the analyst's own estimate assumesSame line, same sentence structure, so the two are directly comparable rather than vaguely opposed.
3. Find the single assumption where the two part company, and name itNot three. One. If it cannot be got down to one, the statements have not been read to the end.

The output is a named disagreement rather than a conclusion about who is right, and naming a disagreement settles nothing about whether a price is correct.

Who reaches for which, on an ordinary working day?

A lender advancing money against a pledged shareholding uses the price, and uses nothing else. If the pledge is invoked, the shares get sold, and they will fetch what they fetch on the day. An estimate cannot be liquidated, so however carefully it was built it has no bearing on what the lender will recover. The lender then applies a haircut to the price. A haircut states how far the price might move before the sale completes, and says nothing about worth.

A research writer uses both, and the value of the note is neither number. The value sits in the named assumption in the middle, the only part a reader can act on, argue with, or watch for evidence about. A note that says the share is worth more than it costs has given the reader nothing to check. A note that says the whole case rests on the materials line staying near 54.0 per cent of revenue has given the reader a thing to watch every quarter.

A household does this without the vocabulary. The neighbours are selling. The flat that changed hands in the next building last month went for a registered amount, and that amount is a record. The household's own figure for the flat two doors down is an estimate, and that figure may sit comfortably above the registered amount for years. Nobody in the lane is wrong about anything. A loan is a transaction, so when the bank sizes one it starts from the record. Moving house is a judgement, so when the household decides whether to move at all it uses the estimate. Neither number has been misused, and nobody has had to declare a winner.

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