- What the MOVE index measures
- Treasuries are collateral, and collateral has a price of uncertainty
- Why bond volatility arrives before equity volatility
- Four volatility and stress gauges, and what each can actually tell you
- The chain from a Treasury option to an Indian share price
- How little of this reaches the score, and why that is worth knowing
- Five ways this tile gets misread
- Where this sits in the app
- Common questions
What the MOVE index measures
MOVE is the amount of movement the options market is pricing into US Treasury yields over the coming month. It is the same idea as the VIX, applied to government bonds instead of shares: not what rates have already done, but what protection against them moving is currently costing.
Start with the option, because the index is only an average of option prices. Somebody who owns a Treasury position and wants protection against yields jumping can buy an option that pays if they do. What they pay for it depends mostly on one thing — how far the seller thinks the yield might travel before the option expires. Wide expected range, expensive option. Narrow expected range, cheap one. Run that backwards across a set of short-dated options on Treasuries of different maturities, and the average expected range you extract is what MOVE reports. The general idea of reading expectations out of option prices, and where it goes wrong, is set out in option chain analysis.
Two properties of that definition do most of the work in the rest of this article.
The first is that it is implied, not realised. It is a price agreed today about a period that has not happened, which means it can move without a single yield changing — a rise in MOVE with a flat bond market means people are paying more for the same protection, which is information about willingness to bear risk rather than about rates.
The second is that it prices range, not direction. Protection against a yield spike and protection against a yield collapse both get dearer when the range widens. So MOVE going up does not say rates are heading anywhere. It says the market has become less confident about where they are heading, and that distinction survives every episode.
One practical caution about units. The VIX is a volatility measure on an equity index price; MOVE is a volatility measure on yields, conventionally quoted in basis points of yield — hundredths of a percentage point — annualised. The two are measured in different things, so setting one reading beside the other and calling one "higher" is a category error. Each is only comparable with its own history.
Treasuries are collateral, and collateral has a price of uncertainty
Here is why a bond-market volatility measure is not a bond-market topic.
A very large share of short-term borrowing between financial institutions is secured against government bonds. The lender does not assess the borrower so much as value the bond, then lend slightly less than it is worth. That gap — the amount by which the loan falls short of the collateral's value — is the haircut, and it exists to protect the lender against the collateral falling in price before it can be sold.
The size of the haircut is set by how far the collateral's price might move — and for a bond, a price move is a yield move scaled by the bond's duration, the translation worked through in duration and interest-rate risk. One multiplication apart, that is the quantity MOVE reports.
The arithmetic of a haircut, using an illustrative assumption. Suppose a book borrows against Treasuries at a 2% haircut — a stated illustration, not a measured market rate. Every $100 of bonds supports $98 of borrowing, so $2 of the borrower's own money carries $100 of position: fifty times leverage. Now the collateral's price becomes less predictable and the haircut goes to 4%. The same $100 of bonds now needs $4 of own money, and $100 divided by $4 is twenty-five times — half the position for the same capital. No bond defaulted. No yield moved. A buffer widened, and the borrower must either find more capital or sell half.
The same mechanism runs through cleared derivatives, where the collateral is called initial margin and is set by revaluing a portfolio across a grid of scenarios and charging the worst outcome. Widen the range of scenarios the grid considers and the margin requirement rises with no change in the position at all — the arrangement Indian traders meet as SPAN margin. A volatility measure is not a description of that process. It is close to being an input to it.
Now the consequence that gives this article its title. When a book must reduce, it does not sell what has fallen. It sells what still has a bid. A margin call is answered in cash by a specific hour, and the assets that convert to cash by that hour are the liquid, well-behaved, frequently profitable ones — which routinely means an emerging-market equity position that had nothing to do with the problem. That is the mechanism behind the familiar observation that everything falls together in a crisis. It is not sentiment spreading. It is one balance-sheet constraint being satisfied in several markets at once, which is why the selling looks indiscriminate: from the seller's point of view it is indiscriminate, and deliberately so.
Why bond volatility arrives before equity volatility
Because trouble that begins with the price of funding shows up first where funding is priced, and equities are downstream of that. Three separate routes run from Treasury uncertainty to share prices, and they are worth keeping apart because they operate on different timescales.
The collateral route is the one above, and it is the fastest, because a haircut or a margin requirement can change overnight and must be satisfied the next morning. It is also the only one of the three that is mechanical rather than interpretive: given a wider haircut and unchanged capital, a smaller position follows as arithmetic.
The risk-budget route is slower and reaches further. Institutions size positions against a risk measure, and a risk measure is built out of volatility. The arithmetic is proportional: a book's total risk number responds to a change in an asset's volatility in proportion to how much of that asset it holds, so the same percentage rise in bond volatility moves the total by more for a bond-heavy portfolio than for an equity-heavy one — and pension, insurance and central-bank reserve portfolios are bond-heavy by mandate rather than by choice. When the total breaches its limit, something gets sold — and the something is chosen for liquidity, not for guilt.
The discount-rate route is the slowest and the least visible. A share is a claim on cash a long way out, and the rate those distant rupees are discounted at is built on the government-bond yield — the arithmetic is in duration and interest-rate risk, and the leg from the US yield to an Indian valuation is traced in US yields and Indian share prices. Note what MOVE adds that the yield itself does not. The yield tells you the rate. MOVE tells you how confident the market is about it, and an uncertain discount rate widens the range of defensible valuations for the same unchanged earnings. The number in the model did not move; the honest error bar around the answer got bigger.
Be exact about the status of the headline claim. That the collateral channel exists is mechanical. That bond volatility therefore leads equity volatility is a judgement about the usual order of events, not a law — and it fails in an identifiable class of cases: when the shock starts in equity itself. A sector earnings collapse, an index event, a policy surprise aimed at shareholders and not at lenders will move equity volatility first and may never touch the Treasury options market. What the mechanism claims is narrower than the slogan: the lead runs from funding to equities and not the other way, and only when the shock started at the funding end.
Four volatility and stress gauges, and what each can actually tell you
These get quoted interchangeably as "fear gauges", which flattens away the differences that decide what a move means.
| Gauge | What it is the price of | Quoted in | What moves it other than fear | What it cannot tell you |
|---|---|---|---|---|
| MOVE | Expected movement in US Treasury yields, from option prices | Basis points of yield, annualised | Policy-meeting and data calendars, Treasury supply, dealer capacity to warehouse risk | Direction. A wider expected range prices a yield spike and a yield collapse identically |
| US VIX | Expected movement in the S&P 500, from index option prices | Percent of index price, annualised | Hedging demand around events, systematic option-selling flows | Whether the movement it prices is expected up or down — and nothing about funding |
| India VIX | Expected movement in the Nifty, from Nifty option prices | Percent of index price, annualised | Expiry-cycle flows, event premium around results and policy dates | Anything about global funding conditions; it is a domestic price |
| Credit spreads | The chance of not being repaid, over a government-bond baseline | Basis points of yield | Index composition, average maturity, liquidity premium | Which borrower is in trouble — an index is an average over a drifting basket |
Read the fourth column before the second. Every one of these has a driver that has nothing to do with stress, which is why a single reading is never self-explanatory. The pairing worth learning is MOVE against credit: MOVE prices uncertainty about the collateral, credit spreads price doubt about repayment, and the two answer different questions about the same funding system. Why lenders reprice before shareholders, and why the rate of widening carries more information than the level, are the subject of credit spreads and where trouble shows up first. The two equity rows are a separate comparison in their own right — the India VIX is not the VIX with a Nifty label, and where the two part company is set out in the VIX and the India VIX.
The chain from a Treasury option to an Indian share price
None of the above is about India yet. Here is the transmission in the order it runs, with the mechanism named at each step rather than assumed.
- Collateral buffers widen and leverage falls globally. The haircut and initial-margin arithmetic above, applied to every book funded against government bonds at once. Nobody has formed a view on India at this stage; a constraint has tightened.
- Positions are cut where they can be cut. Liquid emerging-market exposure is among the easiest things to sell in size on a given morning, so it absorbs a share of the reduction that is disproportionate to how much it caused. This is the step that makes the selling look unrelated to any Indian news, because it is unrelated to any Indian news.
- Dollar funding gets dearer for whoever borrows in dollars. Treasuries are the collateral behind a great deal of that lending, so uncertainty in the collateral raises the price of the loan — which reaches Indian issuers carrying foreign-currency debt, and reaches the rupee through the same allocation channel described in the dollar index and emerging markets.
- Crowded funding trades come under pressure at the same moment. Any strategy that borrows cheaply in one currency to hold risk in another is sized against volatility, so rising volatility shrinks it by construction — the unwind sequence followed through in the yen carry trade. The relevant point here is timing: this is a second forced seller arriving in the same window as the first.
- Domestic effects, which are real but slower. Indian government-bond yields respond partly to global rates, and the portion of a bank's bond book carried at market value transmits a yield move into reported capital rather than into a note nobody reads — which is why sharp global rate moves reach bank shares in a way they do not reach, say, a domestic consumer business. The underlying idea of what a rate is and what moves it sits in interest rates explained.
Two honest limits on that chain. The first four steps are about foreign capital and global balance sheets, so they reach Indian equities through the flow and the discount rate, not through anything a company did — which also means a domestic-funded business with no foreign borrowing meets this only at steps 2 and 5. The second limit is that India's own bond market is not funded the way the Treasury market is, and there is no comparable published measure of implied volatility on Indian government bonds, so the domestic leg of this story has to be read off yields and bank balance sheets rather than off an index.
How little of this reaches the score, and why that is worth knowing
A fully saturated MOVE moves the composite score by under four points out of a hundred, and the tile cannot trigger the stress state at all. The page carries MOVE as a scored input; here is where that ceiling comes from. The macro signals pillar sets out the whole composite; what matters here is one input's share of it.
The score is 100 × Σ(wi·ci) / Σ(wi). Each input's daily percentage change is divided by a per-input scale, clamped to the range −1 to +1, and signed according to whether rising helps or hurts Indian equities. MOVE carries a weight of 0.05, a sign of −1 — rising is modelled as a headwind — and a scale of 5.0, meaning a 5% one-day move saturates its contribution.
The denominator is worth stating precisely, because it is not a fixed number. Twenty market series carry a weight, and those weights sum to 1.33. A twenty-first input — the day's net foreign institutional cash flow — is folded in at a weight of 0.12 on days that figure is available, taking the divisor to 1.45. And an input whose feed fails is dropped from the top and the bottom of the fraction rather than scored as zero, which shrinks it again. The divisor is the sum of the weights that actually returned data that morning, so it sits at or below 1.45 and moves day to day.
Work out the ceiling on that basis. A fully saturated MOVE contributes 0.05 divided by a divisor between 1.33 and 1.45 — between roughly 3.4 and 3.8 of the hundred points, so under four either way. The VIX, at weight 0.10, contributes exactly twice whatever MOVE does, and that ratio holds on any day regardless of which feeds arrived, because the divisor cancels between any two inputs. Per single percent of daily movement the two sit much closer together: 0.05 ÷ 5.0 is 0.010 for MOVE against 0.10 ÷ 8.0, or 0.0125, for the VIX — four to five, not one to two. The gap between them is in the ceiling, not in the sensitivity.
The tile designed to arrive first is the one the score is slowest to hear. Two design choices compound. MOVE's weight, 0.05, sits below the five inputs at 0.06 and is a quarter of the dollar's 0.20. And the page's stress state — the override that can declare stress on a single input, without waiting for the weighted average to get there — triggers on a score at or below −35, a one-day VIX jump of 20% or more, or a one-day fall of 1.2% or more in USD/JPY. It does not trigger on MOVE at all. So the fast path watches equity volatility and the yen, and the early-warning tile reaches the number only through a weight that caps it under four points. The practical consequence is specific: if you think the bond-vol lead is real, you have to read the tile, because the composite is built not to raise its voice for it.
Every constant on this page — the weights, the signs, the scales, the cut-offs, the 0.12 on foreign flows — is FNOTrader's modelling judgement, not a measured property of the world. No regression produced them and no such constant exists to be measured. A different desk reading the same evidence would reasonably choose differently, and the numbers are published precisely so that a reader who disagrees can see what they are disagreeing with.
One sourcing note, since it changes how an unchanged reading should be read. The MOVE tile comes from a market feed rather than from FRED, so unlike the US 2-year and both credit-spread tiles it is not subject to the one-to-two-session publication lag. A flat MOVE tile means flat, not stale.
Five ways this tile gets misread
- Reading the colour as direction. Rising MOVE renders red because the model signs it as a headwind for Indian equities, not because red means "up". Across the page the colour is always the modelled effect on Indian shares — which is why a falling dollar index renders green and a rising USD/JPY renders green too.
- Comparing the MOVE reading with the VIX reading. One is basis points of yield, the other is a percentage of an index price. They are not on a common scale, and the sentence "bond vol is higher than equity vol" built from the two raw numbers means nothing. Compare each against its own range instead.
- Treating a low reading as reassurance. Calm is the modal state of every volatility measure, because most periods are calm — so observing calm is the reading you would expect on the great majority of days, including the ones shortly before the quiet ended. Its base rate of being low is exactly what makes the level uninformative. The informative event is a fast change and whether it persists.
- Turning co-movement into cause. The correlation panel computes Pearson correlation on daily returns over a 30, 60 or 90-session window, and a coefficient cannot distinguish "one moved the other" from "both responded to the same funding conditions", which is the ordinary case here. "MOVE is driving Nifty" is not a sentence a correlation can support. A sign flip between the 30-day and 90-day window is telling you about the window, not about the market.
- Treating a spike as a signal to act. The mechanism above constrains how much leverage the system can carry; it says nothing about whether an equity drawdown follows, and nothing in it should be read as claiming one does. Tighter collateral is a description of the environment a position sits in, and an environment read is not an entry. We have not measured a hit rate for this tile and do not offer one. What size of position suits a given horizon is a question about asset allocation, and no macro tile has a view on it.
The trade-off in watching this at all, stated plainly: an early signal is early precisely because it fires before the outcome is settled, which is the same sentence as saying it fires on occasions where no outcome follows. You buy earliness with false alarms — that is definitional, not a finding. There is no version of this measure that is both early and rarely wrong, and a reader who wants one is asking for a forecast rather than an indicator.
Where this sits in the app
The value of a macro read is in seeing the channels separately rather than in one number that has already blended them.
FNOTrader's Options Analytics app carries the macro page described here: MOVE grouped with the rates tiles alongside the US 2-year, 10-year and 30-year yields and both credit-spread series, each with its transmission mechanism written out on the tile, the composite score with its largest contributors named in words, and the rolling correlation heatmap across 30, 60 and 90-session windows with each window shown separately rather than blended. Every tile opens its own history chart from one month to five years, so a reading can be placed against its own range instead of against a memory of it.
The weights, scales and regime cut-offs are the ones stated in this article, and they are ours. FNOTrader is not a SEBI-registered investment adviser. Nothing here is a recommendation about any security or market, and nothing here is a forecast — it is the mechanism and the arithmetic, so the interpretation stays with the reader.
Common questions
What is the MOVE index?
A measure of how much movement the options market is pricing into US Treasury yields over the coming month — implied volatility on government bonds, the same idea as the VIX applied to rates instead of shares. It is derived from short-dated Treasury option prices and quoted in basis points of yield, annualised. It prices the expected range, not the direction: protection against a yield spike and against a yield collapse both get dearer when it rises.
How is MOVE different from the VIX?
Two ways that matter. MOVE prices expected movement in Treasury yields; the VIX prices expected movement in the S&P 500. And they are quoted in different units — basis points of yield against a percentage of an index price — so the two readings are not on a common scale and comparing them numerically is a category error. Each is only comparable with its own history.
Why does bond volatility affect share prices at all?
Because US Treasuries are the collateral most short-term institutional borrowing is secured against. A lender against that collateral holds a buffer sized by how far its price might move, which is precisely what MOVE prices. Widen the buffer and the same capital supports a smaller position, so leverage falls across every market a borrower holds — and the selling that follows lands where assets can be sold quickly, not where the problem started.
Does a rising MOVE index mean equities will fall?
No, and nothing here forecasts that. A rising MOVE says the market has become less confident about where Treasury yields are going, and there is a channel — tighter collateral, lower system-wide leverage — by which that condition reaches risk assets. But the channel constrains how much leverage the system can carry; it does not say an equity drawdown follows, and we have not measured a hit rate for it. It characterises the environment a position sits in. It does not establish what happens next.
Why is a rising MOVE tile red on the macro page?
Because a tile's colour shows the modelled effect on Indian equities, not the direction of the number. MOVE carries a sign of −1 in our model, so rising is treated as a headwind and renders red. The same convention makes a falling dollar index green and a rising USD/JPY green, since a weaker yen means the yen-funded carry trade is intact. Reading the colour as direction inverts half the page.
How much does MOVE move the composite macro score?
Less than its billing suggests. It carries a weight of 0.05, and the divisor is the sum of the weights that returned data that morning — twenty market series summing to 1.33, plus 0.12 for foreign institutional cash flow on days that figure is available, less anything whose feed failed. So a fully saturated MOVE contributes between roughly 3.4 and 3.8 of the hundred points: under four either way, and exactly half of whatever the VIX contributes at weight 0.10, since the divisor cancels between them. Its scale of 5.0 means a 5% one-day move already saturates it. Every one of those constants is FNOTrader's design choice, not a measured fact.
Does the macro page's stress alert trigger on a MOVE spike?
No. The stress state triggers on a score at or below −35, a one-day VIX jump of 20% or more, or a one-day fall of 1.2% or more in USD/JPY. MOVE reaches the page only through its 0.05 weight in the average. So the fast path watches equity volatility and the yen, and a reader who thinks bond volatility arrives first has to read that tile directly rather than wait for the composite to flag it.
Is the level of MOVE informative, or the change?
The change, over days rather than months. Calm is the modal state of any volatility measure, because most periods are calm — so a low reading is what you would expect on the great majority of days and tells you very little. A fast rise, and whether it persists, is the part that carries information. That is the same reading discipline that applies to credit spreads.
Does bond volatility matter for an Indian investor with no foreign exposure?
It reaches them, but indirectly and through other people's balance sheets. Foreign investors sizing positions against volatility reduce emerging-market exposure when collateral tightens, which shows up as flow rather than as news about any Indian company. Domestically, global rate moves feed Indian yields, and the portion of a bank's bond book carried at market value transmits that into reported capital. A domestic-funded business with no foreign borrowing meets this only through those two routes.
Continue reading
More in Macro & Markets · App: Options Analytics · Definitions: glossary · Free tools: calculators · All: every article