Risk vs Volatility: Two Words That Are Not the Same Thing

Volatility is how much a price moves. Risk is the chance of not having what is needed when it is needed. Everyday speech uses the two as one word, and they are not the same thing: a price that moves a great deal is not risky to somebody who does not have to sell, and a price that never moves can be very risky to somebody who does.

The two definitions are not close variants of each other. Volatility is a statement about a price and nothing else. Risk cannot be stated at all until somebody says who needs what, and on what date. Volatility can be computed from a record of a price and handed to a stranger; the record does not contain the household, so risk cannot be computed from that record at all. Two households can hold identical things and be exposed to completely different risks, and no measurement taken from a price will ever tell them apart.

Every rupee below belongs to one invented household. The Bhosale household is Meghna Bhosale, who is salaried, Ashok Bhosale, who runs a tailoring counter, and Ira Bhosale, who is at school. Rs 42,770/- leaves the household in an ordinary month once the once-a-year items are spread across twelve. At the end of its second year it holds Rs 3,67,887/- in all and owes Rs 71,594/-. The difference is Rs 2,96,293/-. The household holds no shares, no fund and no monthly investment plan.

What is volatility, exactly?

Start with something that has no money in it. A vegetable seller at a weekly street market prices coriander differently on a Tuesday than on a Friday, and differently again after two days of rain. The price is not misbehaving. A price does one thing only. It settles wherever the day's buyers and sellers agree, and then settles somewhere else tomorrow. VolatilityHow much a price moves. Volatility describes the behaviour of the price and nothing else, and it can be worked out from the record of that price without knowing anything about who holds the thing. is a description of how much that settling wanders: not which direction, not why, not whether it will continue, simply how much. Three properties follow, and each of them matters later.

The first property of volatility is that it is computed from a price seriesThe record of what one price was, day after day or month after month. A price series is the only input a measure of movement needs. and from nothing else. Feed in the record of the price and a number comes out. Nobody has to be interviewed, no household has to be visited, no plan has to be disclosed. Volatility appears on documents for that reason: movement is the kind of fact that can be established at a desk by somebody who has never met the reader.

The second property is that volatility is backward looking by construction. The movement that has already happened is the only movement there is a record of. A number describing how much something has wandered is not a description of how much it is about to wander, and treating the first as the second is one of the commonest quiet errors in this subject.

The third is that volatility is symmetric: a sharp rise and a sharp fall contribute to the measurement in the same way. A household experiences a rise and a fall very differently, so the symmetry offends most people's intuition. The offence is an early sign that the word has been asked to carry something it was not built to carry.

The opposite of volatile is stableNot moving. Stability is a fact about a price, or about a balance that has no price at all, and it says nothing at all about a household.: the value in rupees is the same next month as this month. Nobody quotes a price for an amount in an ordinary account, so it is stable in exactly that sense. Rs 31,320/- kept aside is Rs 31,320/- in three months. Nothing said so far mentions a person, a date, a bill falling due or a hospital admission. Volatility is a complete idea, and it is complete with no household in it at all.

What is risk, exactly, and why does that definition contain a household?

RiskThe chance of not having what is needed when it is needed. Risk is a statement about a household and a holding together, and it cannot be worked out from a price on its own. is the chance of not having what is needed when it is needed. Three things are named in that sentence and all three are load bearing: an amount, a date, and the household that needs that amount on that date. Remove any one of the three and the sentence stops meaning anything.

Take the street market again. Two people stand at the same stall looking at the same wandering price. The first is buying coriander for a household meal and can come back on Friday if today looks dear. The second has a catering order to deliver at seven this evening and is buying whatever is in front of him at whatever it costs. The price and the movement are identical for both. The exposure is not, and the difference is not in the vegetable. The difference is in the date.

Risk is a property of a holding and a household together. No measurement taken on the holding alone can ever contain it. A number computed from a price cannot know about a date it was never told, or an amount nobody entered.

Two further terms make the definition usable. The horizonWhen the money is needed. Risk is about having something on a date, so without a horizon there is no way to say whether anything is risky. is the date the money is required, and a household with a date next month and a household with a date in twenty years are in different situations even if every other fact about them is the same. The shortfallNot having what is needed when it is needed. A shortfall is the event that risk is the chance of, and it is measured in what the household required rather than in how far a price moved. is the event risk is the chance of: the amount required minus the amount available, on the day it is required. A household that needs Rs 42,770/- on a Tuesday and has Rs 31,320/- is short by Rs 11,450/-, and that sentence is complete without any reference to a price. The units differ. Volatility is measured in movement, and risk is measured in shortfall.

Most confusion lives in one more fact. A shortfall does not require anything to have fallen in value. A shortfall requires only that the amount available on the day is smaller than the amount needed on the day. A perfectly stable holding that is simply too small produces one just as reliably as a holding that dropped, and with more certainty. Nothing about a stable holding is going to change in the household's favour either.

Criterion one: can either of the two be measured without knowing anything about the household?

Five criteria separate volatility and risk. The first explains everything that follows, including why the words merged.

Volatility is measurableComputable from data alone. A measurable quantity can be worked out by somebody who has never met the person reading the answer. from the price record alone. Somebody in an office who has never met the Bhosale household, does not know that Rs 42,770/- leaves it every month and has never heard of Ira Bhosale can compute how much a price has moved and print the answer. Risk is not measurable from that record at any level of effort. The obstacle is not that the calculation is hard, or that better data would fix it. The inputs are absent. A quantity cannot be computed from data that does not contain its inputs, and the household is one of the inputs to risk.

The consequence is uncomfortable and it is the reason the distinction is worth drawing. A number that can be produced for everybody must be one that needs to know nothing about anybody, and a number that is actually about a particular household requires somebody to sit with that household. The two requirements point in opposite directions, and an industry serving millions of people it has never met will always produce the first and always struggle with the second. The same asymmetry is why movement figures appear on the standard summary documents describing things people can hold and no equivalent figure for risk appears on them, ever. The absence is not an oversight and no rule forbids it. The document is written before the reader exists, and risk cannot be written before the reader exists.

Criterion two: what is each of the two a property of?

Volatility is a property of a price. Volatility belongs to the thing, travels with the thing, and is the same number for every person looking at that thing on the same day. Risk is a property of a pair: the holding is one half and the household is the other, and the quantity only exists where the two meet. Two households holding identical things face identical volatility. Their dates and their amounts differ, so they usually face completely different risk.

A ladder leaning against a wall has a steepness, and the steepness is a property of the ladder and the wall. Whether the ladder is dangerous is a property of the ladder and the person climbing it. A steep ladder is not dangerous to a roofer holding on with both hands; a gentle one is dangerous to somebody carrying a full paint tin at the end of a long day. The angle can be measured without meeting anybody. The danger cannot be stated without knowing who is on it.

A statement of the form "that holding is risky" is always incomplete, and a statement of the form "that price moves a great deal" is always complete. The second names everything it needs. The first has left out one of its two arguments, and the missing argument is the household. The Bhosale household holds gold at its own estimate of Rs 1,40,000/-, received at a wedding and chosen by nobody as an investment, and the movement in that price is exactly the movement every other household holding the same thing faces.

Criterion three: does the passing of time change volatility or risk?

Volatility does not change because a date changes. The measurement has no slot for a date to go into, so no date can affect it. The date is inside the definition of risk, so risk changes completely with the date, and it changes in a way that can flip the answer from serious to irrelevant.

A great deal of loose talk sits on top of the mechanism, so the mechanism is worth naming precisely. A movement in a price becomes a shortfall only for a forced sellerSomebody who has to sell on a particular date, whatever the price is on that date, because the money is needed then. For a forced seller, a movement turns into a loss.: somebody who must convert the holding into money on a particular day, whatever the price happens to be that day. If nobody has to sell on any particular day, the movement stays a fact about a price and never becomes a fact about the household. If somebody has to sell on the fourteenth, the price on the fourteenth is the only price that has ever mattered to them.

So what a horizon really changes is whether the household is a forced seller. Time does not smooth a price out; what a longer horizon does is remove the household from the position of having to accept whatever a particular day offers. That is a statement about the household, not a claim about what any price will do. The reverse case matters just as much and is discussed far less. A household whose date is close and whose amount is large is exposed even to a holding that never moves. The closeness of the date removes every option it might have had.

Criterion four: what actually reduces each of the two?

Volatility is reduced by holding things whose prices move less, or by holding a spread of things whose movements do not line up with each other. The formal treatment of that second idea, and the arithmetic showing that a combination of holdings can move less than the things inside it, is Harry Markowitz's, published in 1952. The Markowitz result is, strictly, a result about movement.

Risk is reduced by three completely different moves, and not one of them is a statement about a price. The first is changing the amount needed: a household that finds a way to need less on the day has reduced its risk without touching anything it holds. The second is changing the date, by moving a requirement from next month to next year. The third is changing what stands against the requirement, by holding more against it, holding something reachable against it, or arranging for somebody else to carry part of it.

Nothing on that list of three is an action taken on a price. No sharper demonstration is available that the two words are not one word. If they were one idea, the things that reduce them would overlap. The two lists barely touch. One honest overlap belongs in the open rather than hidden away. For a household that must sell a holding on a known date to meet a known amount, reducing the movement in that holding does reduce its risk. The movement is what could produce the shortfall. The case is real and narrow, and the trouble is that it gets applied far outside its own boundaries.

Criterion five: what does each of the two cost when it arrives?

Volatility, when it arrives, costs a household nothing directly. A price moved. If nothing has to be sold, no money has left, no bill went unpaid, and no plan changed. Volatility costs discomfort and nothing more. Risk, when it arrives, costs the household the shortfall: something needed did not happen, or happened by borrowing at a cost, or happened by selling something that should not have been sold, or happened by asking a relative. There is an actual event with an actual date and an actual amount attached to it.

Households make their worst decisions on the discomfort side, so it deserves its own sentence. Daniel Kahneman and Amos Tversky showed that people weigh a decrease far more heavily than an increase of the same size, and that people treat a recent run as evidence about the next stretch far more readily than the arithmetic supports. A household watching a price fall experiences something that feels exactly like a loss even when nothing has been sold and nothing is needed. Neither that reaction nor its mirror is stupid, and both are extremely common. Both are volatility being processed as though it were risk.

So the two words differ in their units, in what they are properties of, in whether time touches them, in what reduces them and in what they cost. Five criteria, five different answers. Two ideas that differ on every criterion available are not two names for one thing.

How can something volatile not be risky at all?

The gold is two bangles and a chain, valued by the household only by asking what such a thing would fetch, and it is the only thing this household holds that carries a price set outside the household at all. Is it volatile? Yes, by construction. The gold has a price, that price is set by people who are not in this household, and prices set that way move. A measurement of movement taken on it would return a number, and that number would be the same for every household in the country holding the same thing.

Is the gold risky to the Bhosale household? Risk is a different question, and answering it needs the two things that are not on any price record. The household has no date on which it must sell the gold, and nothing in its month depends on the gold's value: Rs 42,770/- leaves every month, and the gold is not part of how that is met. No school fee, no rent, no repayment and no plan of any kind is standing on it.

Because there is no date and no plan attached to it, a movement in the gold's price costs the Bhosale household nothing it needs. The number changes and nothing else does. The argument left two things out, and both are worth naming. No size was stated for the movement and none was needed. Nor was there any claim that the movement will be recovered. Nobody can promise that. The argument used one fact: nothing the household needs is standing on that value on any date. The useful question was never about the price, so a document describing the movement in that price would have been completely accurate and completely useless.

How can something perfectly stable be very risky?

The Bhosale household keeps Rs 31,320/- in an account set aside and not touched in the year. There is no price on it, so there is nothing to move, and a measurement of movement taken on it returns zero and would return zero every year. On the criterion that gets printed on documents, it is the calmest thing this household has.

Now ask the risk question properly: what does the household need, and when? Rs 42,770/- leaves in an ordinary month, so Rs 31,320/- is 0.73 months. Adding the Rs 10,567/- in the account the salary lands in gives Rs 41,887/- reachable the same day, or 0.98 months. The household ran a deficit in the year just finished, so the buffer is not being added to out of the month, and it owes Rs 71,594/-. When a hospital admission arrived, the share the household had to produce itself came to Rs 50,560/-. Rs 8,673/- had to be found somewhere else on the day.

The most stable thing the Bhosale household holds is the thing most likely to leave it without what it needs, and its stability does nothing whatever about that. Rs 31,320/- that never moves is Rs 31,320/- when the admission happens. The steadiness is real and it is not the point. The amount is the point, and the amount is 0.73 months against a month that costs 1.00.

Read that as a statement about what the word risk means and not as a verdict on the household. A buffer built out of a single salary and a tailoring counter, across two years in which the household also met an admission and carried Rs 71,594/- of borrowing, is a considerable piece of work. The number that describes how safe the buffer looks and the number that describes what it would actually do are two different numbers, and only one of them is 0.73.

The six readings run as follows. At a need one month out the flat holding covers what is needed on every shape. The first drawn shape sits below the line, the second above it and the third above it. At a need two hundred and forty months out the rising requirement has passed the flat holding and the flat holding has not moved to meet it, so the flat holding is short on every shape. The first drawn shape sits above the line, the second below it and the third far below it. The flat holding gives one verdict at each date whatever shape is drawn, and the moving holding gives two different verdicts at the very same date depending only on which shape happens to be drawn. The flat holding and the rising requirement meet exactly at sixty months on this drawing, and the flat holding is short from sixty-one months onwards.

Why did these two words merge into one in the first place?

Nobody decided to merge them. Start from the position of somebody who has to describe a holding to people they will never meet. Such a describer would love to tell each reader how risky the thing is and cannot. Risk requires the reader's amount and the reader's date, and there are millions of readers and one document. Describing the thing itself is possible, and how much its price moves is a genuine, checkable property of the thing.

Then a quantity that can be computed becomes a quantity that gets compared, ranked and eventually managed. Markowitz's 1952 work put the movement of a price at the centre of how holdings are combined, and the mathematics is excellent. One habit travelled outwards over the following decades, shorter than the mathematics behind it: using the movement of a price as the operational stand-in for the word risk, everywhere, including in sentences where the original word was doing a different job. The household then reads that description of movement as an answer to the question it actually has: whether it will be all right. Nobody lied and nothing was hidden. The word simply arrived carrying more than it can hold.

The merge lost precisely the household: the amount, the date and the person who needs them. All three were the only things that were ever going to decide the answer. A vocabulary with one word for both ideas cannot express the sentence "this holding barely moves and it is the most dangerous thing the household has", and that sentence happens to be true of the Bhosale household.

Which of the two can a household actually act on?

A household cannot act on volatility. A household can hold something whose price happens to move less, or hold a spread of things, but the movement in any given price is set by people the household has never met and will never influence. There is no lever in the house connected to it. A household can act on every single component of risk: it can change the amount it needs, it can change the date it needs it, and it can change what is standing against that date. The three levers are real, all of them inside the household, all of them things Meghna Bhosale and Ashok Bhosale can decide about on a Sunday afternoon without asking anybody.

The word a household can do something about is the one that gets less attention, and the word that gets all the attention is the one no household controls. The imbalance is not an accident of temperament. The imbalance follows directly from criterion one: the uncontrollable one is the one that produces a number, and a number is what a conversation gathers around.

For the Bhosale household this converts into a specific reading rather than a general encouragement. Nothing the household needs is standing on anything with a price, so the number that describes its position is not a movement figure of any kind. The number that describes its position is 0.73, and after that 0.98, and after that Rs 8,673/-. A single admission left that much over after everything reachable had been used. Whether the household should do anything about those numbers is a separate question, and there is no version of the answer that would fit every household.

How does somebody assessing a household actually read the difference?

A lender deciding whether the Bhosale household can carry a repayment does not look at how much any price moves. The lender looks at what comes in, what already goes out, what is already owed, and whether a payment falling due on the fifth can be met on the fifth. Every one of those is a statement about amounts and dates. A lender is reading risk in the sense of an amount, a date and a household, and it does not need a movement figure to do it.

Somebody analysing a business for a living uses both and knows which is which. The movement in a price is used as a fact about the price, in the places where movement is the actual question, and whether the business will have money to meet what falls due is a separate exercise built out of amounts and dates. A household reading its own position is doing what a lender does, for itself: what does it need, when does it need it, what is standing against it, and how many months does what it can reach actually cover. For the Bhosale household those answers are Rs 42,770/- a month, Rs 41,887/- reachable and 0.98 months of cover, and not one of them required knowing anything about any price.

There is one more reader worth naming, and it is the reader who arrives with somebody sitting across a table waiting for an answer. If a description of a holding is being offered, two questions separate the words. The first is what movement figure is being described. The question is fair and has a real answer. The second is what amount this household needs and on what date. No document can have answered that in advance. A conversation that has answered the first question and not the second has described a holding and has said nothing about the household.

When does the distinction stop mattering?

Separating the two words takes most of the argument, and the other half of the answer is the set of conditions under which arguing about which word to use changes nothing at all. Three of them are visible in the Bhosale household's own figures.

The first is that nothing held against what the household needs carries a price. Every rupee this household can reach on the day it needs money, Rs 41,887/-, sits in balances that nobody quotes a price for. A measurement of movement taken on that money is zero and would be zero every year. One of the two words has nothing to describe here, so they do not disagree, and the reading is 0.73 months of cover whichever word is used to arrive at it.

The second is that the gap is too large for the choice of holding to close. The admission needed Rs 50,560/- and everything reachable came to Rs 41,887/-. No rearrangement of where that Rs 41,887/- was sitting produces the missing Rs 8,673/- on the day, and no movement figure attached to any holding changes that subtraction. Where the amount available is short of the amount needed before anything has moved, both words return the same answer, and the answer is short.

The third is that the household is not a forced seller on any date. The Bhosale household stands there today. Nothing it needs is standing on the gold, so a movement in the gold's price is a fact about a price that never becomes a fact about the household. While that holds, a movement figure and a risk reading agree that the gold is beside the point, and it makes no difference which of the two words is used to say so.

The third condition is the one that expires without announcing itself. The condition held on the day of the admission only because Rs 41,887/- was reachable without touching the gold. Had the missing Rs 8,673/- needed to come out of the gold that day, the household becomes a forced seller, the gold's price on that one day is the only price that has ever mattered to it, and how much that price moves turns into a direct input to whether the household has what it needs. In that position, reducing movement genuinely does reduce risk, and the two words point at the same action. Nothing notifies a household that it has crossed into it. Dates appear on their own: an admission, a fee, a job ending, a repayment falling due earlier than expected. The condition that made the distinction safe to ignore lapses on the day the money is needed, and that is the same day the household finds out it has lapsed.

One thing here is not a condition and should not be dressed as one. Where the distinction bites hardest is the household's actual position: a still holding that is simply too small. There the two words give opposite answers, the reassuring one is the wrong one, and no condition in this section softens that or is meant to.

What is the shortest version of the whole comparison?

Volatility answers the question "how much does this price move", and it answers it completely, from the price record, for everybody at once. Risk answers the question "will this household have what it needs on the day it needs it", and it cannot be answered at all until somebody says what is needed and when. One of the two can be printed on a document and handed to a stranger, and the other cannot. The gap is not a defect in anybody's paperwork but a consequence of what the word means. The measurable one ended up wearing the other one's name, and the cost of that is that a household can read a document, understand every figure on it correctly, and still have learned nothing about itself.

For the Bhosale household on the last day of its second year, the reading is two sentences. The gold at Rs 1,40,000/- moves, and because nothing the household needs is standing on it, the movement costs it nothing it needs. The buffer at Rs 31,320/- does not move at all, and because it is 0.73 months against Rs 42,770/- a month in a household that ran a deficit and once had to produce Rs 50,560/- on a day, it is the thing most likely to leave the household short. Volatile and not risky, then stable and very risky, from one household on one day.

The Bhosale household, Meghna Bhosale, Ashok Bhosale and Ira Bhosale are invented.
Educational material. Not advice on any investment, tax, budget or market position.