Lesson 14
IV Rank (IVR) vs. IV Percentile (IVP)
One reading, two questions: where is IV within its past range, and how often was it lower?
Same 18%. Two different scores.
Prior annual IV readings · %
Hypothetical NIFTY 50 IV · 12 prior readings, oldest to newest · today's 18% excluded from the count and the range
Give today’s IV a reference
I’m Mr. Chartist. Let’s place one NIFTY volatility reading against its own history, and see what two different scores can and cannot tell us.
By the end you will be able to answer
- Why is 18% not enough on its own?
- How does IV Rank place today on the range?
- How does IV Percentile count the readings?
Your screen shows an implied volatility (IV) of 18% for NIFTY 50. The last lesson showed that IV is a yearly volatility input read from option prices. Is 18% high or low? That depends on what you compare it with. Two simple scores place it against its own history: IV Rank and IV Percentile.
IV Rank (IVR) shows where today's IV sits between the historical low and high. IV Percentile (IVP) shows the share of past readings that were lower than today's. Both describe history only. Neither says whether NIFTY will rise, what volatility comes next, or whether an option trade will make money.
Updated
Why is 18% not enough on its own?
A reading needs its own history to mean something.
My screen shows an IV of 18% for NIFTY. Is that high or low?
That depends on what you compare it with. Two simple scores place it against its own history.
IV Rank shows where today sits between the historical low and high. IV Percentile shows the share of past readings that were lower.
So a high score tells me what NIFTY will do next?
No. Both scores describe history only. They do not say where NIFTY goes, what volatility comes next, or whether a trade will make money.
The IV level still matters. An IV of 18% and one of 30% price different amounts of movement, even if each sits mid-way in its own history. The scores add context to the level. They do not make different instruments equally risky. A quiet index at a high score can still have a lower IV than a wild stock at a low score.
First, define what the history measures. Our example is a constructed series of NIFTY IV: annualised, at the money (strike near the index level) and always 30 calendar days ahead. "Constant maturity" means the series keeps describing a 30-day horizon as dates pass. It does not follow one contract whose life keeps shrinking. Every reading uses the same selection, quote and model rules. A real data provider must explain how it builds such a series.
We compare today's 18% with twelve prior readings, oldest to newest: 14%, 15%, 16%, 16%, 17%, 18%, 18%, 19%, 20%, 21%, 22% and 24%. Today is kept out of both the high-low and the count. Some services use longer windows such as 52 weeks, and the window and number of valid readings belong next to the score. We use twelve so the whole history stays on screen.
How does IV Rank place today on the range?
Rank shows where today sits between the low and high.
In one line: IV Rank is today's distance above the low, as a share of the full low-to-high range. The formula is (current minus low) divided by (high minus low), times 100. Think of a ruler. Only the two ends and today's marker matter. The readings in between change nothing.
Our low is 14% and our high is 24%. Today's 18% is four points above the low, in a ten-point range. So IV Rank is 4 ÷ 10 × 100 = 40. The IV is still 18%. The 40 is a position score, so never put it into a pricing model or a movement-scale formula in place of volatility.
Where is 18% on the ruler?
Compare the current annual IV with the lowest and highest prior readings.
Hypothetical NIFTY history · current IV 18%
(18 − 14) ÷ (24 − 14) × 100
5 of 12 readings strictly below · 2 equal · 5 above
Inspect the prior observations and calculation
Current IV 18% is separate from these 12 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 14% | Below current |
| 2 | 15% | Below current |
| 3 | 16% | Below current |
| 4 | 16% | Below current |
| 5 | 17% | Below current |
| 6 | 18% | Equal to current |
| 7 | 18% | Equal to current |
| 8 | 19% | Above current |
| 9 | 20% | Above current |
| 10 | 21% | Above current |
| 11 | 22% | Above current |
| 12 | 24% | Above current |
IV Rank = (18 − 14) / (24 − 14) × 100 = 40.0000%
IV Percentile = 5 / 12 × 100 = 41.6667%
Range position is a score, not annual volatility or a probability. Current IV is excluded from the reference history.
An IV Rank of 50 means the midpoint between the two ends. It does not mean half the readings were lower. Most readings could cluster near either end. When today sits inside a range that has some width, the score falls between 0 and 100. We look at what happens outside that range, and when the ends coincide, later.
How does IV Percentile count the readings?
Percentile counts the past readings that were lower.
In one line: IV Percentile is the share of prior readings that were lower than today's. Count those readings, divide by the number of valid prior readings, and multiply by 100. This lesson counts strictly below. A reading equal to today's stays in the total but does not count as lower. Counting ties as lower would answer a different question.
Five of our twelve prior readings are below 18%: 14%, 15%, 16%, 16% and 17%. Two equal 18%. Five are above. The percentile is 5 ÷ 12 × 100, which is 41.67 after rounding. Repeated values are separate readings, so never merge them. We keep full precision in the working and round only what we display.
Five below, two equal, five above
Every prior observation stays visible. Only readings strictly below today’s IV count in the numerator. It starts at 18%; move it to see both scores respond.
12 prior observations · current IV 18%
(18 − 14) ÷ (24 − 14) × 100
5 of 12 readings strictly below · 2 equal · 5 above
Inspect the prior observations and calculation
Current IV 18% is separate from these 12 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 14% | Below current |
| 2 | 15% | Below current |
| 3 | 16% | Below current |
| 4 | 16% | Below current |
| 5 | 17% | Below current |
| 6 | 18% | Equal to current |
| 7 | 18% | Equal to current |
| 8 | 19% | Above current |
| 9 | 20% | Above current |
| 10 | 21% | Above current |
| 11 | 22% | Above current |
| 12 | 24% | Above current |
IV Rank = (18 − 14) / (24 − 14) × 100 = 40.0000%
IV Percentile = 5 / 12 × 100 = 41.6667%
Today is excluded. Ties remain in the denominator. Scores are rounded for display; the underlying calculation keeps full precision.
So about 41.67% of this sample had a lower IV. That is not a 41.67% chance that the index rises, that IV falls, or that a trade profits. The total is the twelve prior readings. It does not include today, days without readings, or an assumed full year.
Why can the two scores disagree?
They answer two different questions about the same history.
They keep different information. Rank keeps how far the extremes reach. Percentile keeps how many readings sit on each side. To see this, take a separate twelve-reading history where the oldest reading is 24% and the other eleven are 14%, 15%, 16%, 16%, 17%, 18%, 18%, 19%, 20%, 21% and 22%. Now swap only that oldest 24% for 50%. Today stays at 18%.
The high jumps from 24% to 50%, so the ruler stretches. IV Rank falls from 40 to 11.11. That reading was above 18% before and after, so the strict-below count is still five out of twelve. IV Percentile stays at 41.67. Neither is wrong. They answer two questions.
One extreme stretches the ruler
Replace the oldest 24% with 50%. Current 18% and the other eleven observations stay fixed.
Separate hypothetical case · Oldest reading 24% · current IV 18%
(18 − 14) ÷ (24 − 14) × 100
5 of 12 readings strictly below · 2 equal · 5 above
Inspect the prior observations and calculation
Current IV 18% is separate from these 12 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 24% | Above current |
| 2 | 14% | Below current |
| 3 | 15% | Below current |
| 4 | 16% | Below current |
| 5 | 16% | Below current |
| 6 | 17% | Below current |
| 7 | 18% | Equal to current |
| 8 | 18% | Equal to current |
| 9 | 19% | Above current |
| 10 | 20% | Above current |
| 11 | 21% | Above current |
| 12 | 22% | Above current |
IV Rank = (18 − 14) / (24 − 14) × 100 = 40.0000%
IV Percentile = 5 / 12 × 100 = 41.6667%
Both changed readings are above current IV. Rank changes; the strict-below count does not. Neither score identifies mispricing.
Percentile is not immune to odd data. If a reading moves from above today's IV to below it, the count changes. In a twelve-reading sample, one such change moves the percentile by about 8.33 points. In a longer sample the effect is smaller. Which score helps depends on whether you care about how far the extreme reaches or how often readings were lower.
Rank moves, percentile stays
Current IV is 18% in both cases. Only the oldest reading changes, from 24% to 50%.
Oldest reading 24%
Oldest reading 50%
Hypothetical twelve-reading history, prior readings only, strictly below. Scores are positions in a history, not probabilities.
Why can IV Rank move when IV does not?
The range itself can change as old readings drop out.
Because the window rolls. Old readings drop out and new ones come in. In the high-extreme case, remove the oldest 50% and add a new prior reading of 24%. The twelve readings become the same set as our main example. With today still at 18%, Rank goes from 11.11 back to 40, and Percentile stays at 41.67. Today's level did not cause that. The reference changed.
This is a controlled comparison, not two real trading days. In a live market, today's IV changes as readings enter and leave the window. To read a score change, look at both the current level and the history. An extreme stays in a rolling window until it leaves the window or a new extreme replaces it. There is no fixed twelve-month rule.
The window changes, current IV stays 18%
Remove the oldest 50% and append a prior 24% reading to the twelve-observation window.
Controlled window comparison · Oldest reading 50% · current IV 18%
(18 − 14) ÷ (50 − 14) × 100
5 of 12 readings strictly below · 2 equal · 5 above
Leaving the window: 50%. Entering: 24%. Today’s IV stays 18%.
Inspect the prior observations and calculation
Current IV 18% is separate from these 12 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 50% | Above current |
| 2 | 14% | Below current |
| 3 | 15% | Below current |
| 4 | 16% | Below current |
| 5 | 16% | Below current |
| 6 | 17% | Below current |
| 7 | 18% | Equal to current |
| 8 | 18% | Equal to current |
| 9 | 19% | Above current |
| 10 | 20% | Above current |
| 11 | 21% | Above current |
| 12 | 22% | Above current |
IV Rank = (18 − 14) / (50 − 14) × 100 = 11.1111%
IV Percentile = 5 / 12 × 100 = 41.6667%
The same current level is compared with two different histories. A real daily change can combine this effect with a change in current IV.
Check a platform's formula, not its label. IBKR documents range-based IV Rank and below-current IV Percentile as separate things. Schwab's article of 15 June 2026 shows a range-position example under the name "Current IV Percentile", and also calls it IVR. So names are not standard. The method behind the number is what lets you compare.
To compare across time or providers, match the underlying, the IV series, the strike choice, the target maturity, the quote side, the sampling time, the model and the window. Ask whether today is in the reference, how ties and gaps are handled, and how many valid readings remain. Two correct scores can still describe different data. A constant 30-day series and one fixed option running down to expiry are not the same history.
What should you ask before using a score?
A score starts an enquiry, it does not end one.
Use a score to start an enquiry. If IV is near its prior high or above most prior readings, ask what the market is pricing now. It may be a scheduled announcement, changed conditions, or a different quote. That alone does not prove the option is overpriced or that selling it has an edge. A low score is not a bargain for a buyer either.
Then go back to the contract. Your Call or Put may have a different IV from the at-the-money series behind the score. Look at its bid, ask, displayed quantity and timestamp. A midpoint is arithmetic, not a fill. Past volume and open interest do not promise you can get in or out at the displayed price.
Buying an option can lose the premium as the underlying, IV and time change. Selling one can lose far more than the premium received. A high score gives no loss limit and no steady income. IV falling after entry does not by itself mean profit, because the underlying move, time, entry and exit prices, quantity and costs also count. Judge the position's exposures and funding needs. Do not use a score to pick a direction.
Settlement rules also matter. NIFTY index options settle in cash. Stock options in India can settle by delivery. A bought in-the-money stock Call held to settlement can need full payment for the shares. A bought in-the-money stock Put can need share delivery. The premium-limited loss does not remove those funding or delivery needs. Sellers must also meet margin and settlement duties. No volatility score changes these rules.
Where do these scores break?
Short or poorly built histories give weak scores.
A precise score cannot fix a poor history. A short sample can be dominated by one event. A shift in volatility regime can make an old window a weak guide to now. Corporate actions or a provider changing its method can break comparability. A missing quote must never quietly become a zero reading. Check the valid count and how the series was built.
Our prior-only rule has two clear edges. If all prior readings are 18%, high minus low is zero and IV Rank is unavailable. The strict Percentile at 18% is zero because nothing is lower. That does not mean IV or risk is zero. Against the main 14% to 24% history, a new reading of 27% gives an unclipped Rank of 130 and a Percentile of 100. A provider that includes today in the range or clips scores will show different numbers.
A score has a data boundary
Compare a flat reference with a current reading above a non-flat prior range.
Hypothetical prior-only history · current IV 18%
Prior high equals prior low
0 of 4 readings strictly below · 4 equal · 0 above
High minus low is zero, so IV Rank is unavailable. Percentile still counts strictly-below readings.
Inspect the prior observations and calculation
Current IV 18% is separate from these 4 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 18% | Equal to current |
| 2 | 18% | Equal to current |
| 3 | 18% | Equal to current |
| 4 | 18% | Equal to current |
IV Rank unavailable: high IV equals low IV.
IV Percentile = 0 / 4 × 100 = 0.0000%
Hypothetical prior-only history · current IV 27%
(27 − 14) ÷ (24 − 14) × 100
12 of 12 readings strictly below · 0 equal · 0 above
Rank is shown without clamping: the current reading sits outside the prior range, so Rank passes 100 while Percentile caps at 100.
Inspect the prior observations and calculation
Current IV 27% is separate from these 12 prior observations. Equal readings remain in the denominator, but are not counted below.
| Observation | Annual IV | Compared with current |
|---|---|---|
| 1 | 14% | Below current |
| 2 | 15% | Below current |
| 3 | 16% | Below current |
| 4 | 16% | Below current |
| 5 | 17% | Below current |
| 6 | 18% | Below current |
| 7 | 18% | Below current |
| 8 | 19% | Below current |
| 9 | 20% | Below current |
| 10 | 21% | Below current |
| 11 | 22% | Below current |
| 12 | 24% | Below current |
IV Rank = (27 − 14) / (24 − 14) × 100 = 130.0000%
IV Percentile = 12 / 12 × 100 = 100.0000%
These are prior-only, strict-below calculations. Unavailable Rank, Rank above 100 or Percentile zero do not describe zero risk or a trade signal.
With no valid history, neither score exists. With a valid history, neither is a forecast probability, and neither guarantees the past pattern will continue. Keep the original IV level, the reference definition and the actual option quote beside the score.
We have compared one clearly defined IV measure with its own past. The next lesson changes the question. Why can different strikes or expiries show different IV at the same moment? That is volatility skew and the wider shape of IV across contracts. One Rank or Percentile cannot describe it.
Before you go
Check yourself
5 questions from this lesson. Pick an answer to see why, then slide on to the next.
Question 1 of 5
Questions readers ask
Common questions
What is the difference between IV Rank and IV Percentile?
IV Rank measures current IV’s distance from a historical low relative to the full low-to-high range. IV Percentile measures the share of reference observations below current IV under the chosen tie rule. Range midpoint and sample median are different ideas. Neither score replaces the annual IV level or predicts direction.
How do tied readings and today’s IV affect the percentile?
This lesson excludes today and counts only strictly lower prior readings. At current 18%, five of twelve prior readings are below, two equal and five above, giving 41.67. Ties remain in the denominator. Counting lower-or-equal instead gives 58.33; a half-tie convention gives 50. Providers can choose different rules, so compare the actual calculation and denominator.
Why can one unusual IV reading change Rank without changing Percentile?
An extreme can change the range endpoint substantially. In our separate example, replacing one 24% reading with 50% stretches the range and lowers Rank at current 18% from 40 to 11.11. Both readings are above 18%, so the strict-below count stays five of twelve and Percentile remains 41.67. If an observation crosses current IV, its contribution to the count changes too; Percentile is not immune to data changes.
Can IV Rank change when current IV stays the same?
Yes. Its reference endpoints can change as a rolling window loses old observations and gains new ones. Removing the oldest 50% and adding a prior 24% in our example raises Rank from 11.11 to 40 while current IV stays 18%. This isolates a reference-window effect. In actual daily data, the current level and the history may both change.
Why do two platforms show different IV Rank or Percentile values?
Their names, formulas or input series may differ. Check the underlying, maturity and strike selection, quote side, sampling time, model assumptions, lookback, valid count, tie rule and whether today enters the reference. Some documented interfaces use a percentile label for range-position arithmetic. Matching a label alone does not establish that two scores answer the same question.
Can IV Rank exceed 100, or be unavailable?
With this lesson’s prior-only, unclipped formula, current IV above the historical maximum gives Rank above 100; 27% against a 14%–24% prior range gives 130. A reading below the historical low gives a negative result. If high equals low, Rank is unavailable because division by zero is undefined. Platforms that include today or clip output can behave differently. An empty valid history provides neither score.
Does a high IV Rank or Percentile mean I should sell options?
No. A high score gives historical context for its selected IV measure; it does not establish fair value, a seller’s trading edge or a win probability. IV can rise further and an adverse underlying move can dominate a fall in IV. Actual contract quotes, exposures, loss potential, margin, settlement and costs still matter. A low score likewise does not guarantee profitable option buying.
Which lookback and IV history should I compare?
Use a clearly defined history suited to the question and disclose its window and actual observation count. 52 weeks is one documented choice, not a requirement of the formulas. Keep the underlying, target maturity, strike selection and calculation method comparable through time. A shrinking-expiry contract is different from a constant-30-day IV series. A short or structurally changed history needs careful interpretation rather than a full-year label.
Sources for further reading
8 sources- investors.interactivebrokers.comIBKR — IV Rank and Percentile data definitions (checked 8 October 2026)
- interactivebrokers.comOIC / OCC — Volatility Concepts for Options Investors (16 November 2021; checked 8 October 2026)
- docs.scipy.orgSciPy — Strict, weak and tied-score percentile conventions (checked 8 October 2026)