Lesson 13
Implied Volatility (IV) and India VIX
Read the IV of one option, understand the wider India VIX reading, and tell priced volatility from actual movement.
One annual reading. Three horizons.
Calendar days ahead
Move across the chart, or drag on a phone, to read each line.
Hypothetical NIFTY 25,000 · 20% annual volatility · calendar days · a scale for typical movement, not a forecast
Read the volatility behind the quote
I’m Mr. Chartist. Let’s work from a NIFTY option quote to its implied volatility, then see what India VIX adds, and what neither number can promise.
By the end you will be able to answer
- What question does IV answer?
- What does 18% actually mean?
- How does a quote turn into IV?
You are watching a NIFTY 25,000 Call before a policy announcement. The index has barely moved, yet the option now costs more. Why? The index candle cannot tell you. The answer sits in the option quote itself, and in how much movement that quote is pricing in.
Volatility is the size of price swings, not their direction. Implied volatility, or IV, is the yearly volatility number that makes a pricing model return exactly the premium you see. It is read backwards from the quote. It is not a fee added on top of the premium.
India VIX asks a related question across many NIFTY option quotes at once. It gives one 30-day volatility reading for the whole index. A single Call can show an IV of 18% on a day India VIX reads 20. Those are hypothetical numbers from two different calculations, so they need not match. We keep the two apart all through this lesson.
Updated
What question does IV answer?
IV tells you how much movement a premium assumes.
The index barely moved, yet the NIFTY Call costs more. Why?
The candle cannot tell you. The answer sits in the quote itself, and in how much movement it is pricing in.
So IV is a fee added on top of the premium?
No. IV is the yearly volatility number that makes a pricing model return exactly the premium you see. It is read backwards from the quote.
And India VIX is the same thing?
A related question across many NIFTY quotes at once. The two numbers come from different calculations, so we keep them apart.
In short: IV tells you how much movement a premium assumes. This lesson reads it from one worked contract and then adds India VIX on top.
Our worked contract is a hypothetical NIFTY 25,000 Call. Spot and strike are both 25,000, with 30 calendar days to expiry. The model uses a 6% yearly interest rate and a 1% yearly dividend yield, both continuously compounded. The model is European Black-Scholes-Merton (BSM): a plain Call or Put, exercised only at expiry. Every price is in rupees per option unit. None of them is a live quote.
The contract starts at an annual IV of 18%. A few separate examples use a 20% reading, or short price paths, to answer one question each. Each one is labelled as its own teaching case, so do not read them as one live market snapshot. The calculations keep full precision, and the text rounds them.
The next lesson places a reading against its own history with IV Rank and IV Percentile. The one after that compares strikes and expiries (skew and term structure). Then comes a lesson on volatility around a scheduled event. Each adds context. No single IV number replaces them.
What does 18% actually mean?
IV is a yearly scale, not a promised move.
An IV of 18% is a yearly volatility figure. It is not a promise that the premium moves 18%, or that NIFTY rises 18%, or that next month has an 18% range. "Yearly" is only a common scale, so readings stay comparable even when the contract has a few days left. You still need the time to expiry to use it.
To feel the scale, assume the volatility rate stays steady through time. Volatility grows with the square root of time. So the movement scale over a period is the annual figure times the square root of the fraction of a year, and times the index level to get points. The next figure uses a separate 20% reading, NIFTY at 25,000 and a 365-day year.
One annual reading, three horizons
A hypothetical 20% annual volatility scale at NIFTY 25,000, using calendar days.
Approximate movement scale · annual volatility 20% · NIFTY 25,000
index points · 1.05% of spot
index points · 2.77% of spot
index points · 5.73% of spot
Thirty times the time gives only about 5.5 times the movement scale.
Spot × annual volatility × √(days/365). Constant variance-rate illustration; not a guaranteed interval or a direction forecast.
Over 30 days the scale is about 1,433.46 points. Over one day it is 261.71 points. Thirty times the time gives only about 5.5 times the scale. A scheduled event can pack the uncertainty into one day, so a steady rate can describe a real expiry badly.
Treat these as a scale of movement, not a range the index must stay inside. They are not forecasts of the high and low, and they are not a real-world two-in-three probability. That needs more assumptions, and real returns jump and have fat tails. IV gives no direction. It is also not an unbiased promise of the volatility that follows.
How does a quote turn into IV?
IV is read backwards from the premium you see.
A pricing model normally runs forwards: give it a volatility and the other inputs, and it returns a premium. IV runs it backwards. You take a premium, hold the other inputs fixed and solve for the volatility that reproduces it. Software does this by searching. For a plain European Call or Put with time left, a higher volatility always gives a higher model value. So a premium inside the model bounds gives exactly one positive IV.
Try it on the 30-day NIFTY 25,000 Call. At 14%, 18% and 22% the model gives about ₹452.44, ₹565.99 and ₹679.70. A premium near ₹566 therefore implies about 18%. That premium need not be the last traded price. A platform may use the bid, the ask or the midpoint, and a stale trade gives a stale IV.
Say the bid is ₹550 and the ask is ₹582. The bid is what buyers offer. The ask is what sellers want. Their ₹566 midpoint is arithmetic, not a price you are sure to get. With the fixed inputs, the three prices imply about 17.44%, 18.00% and 18.56%. The midpoint gives 18.00% to two decimals. Price and IV are not in a straight line, so midpoint IV need not equal the average of the other two.
The quote selects the inferred IV
Trial model values and a hypothetical bid–midpoint–ask range for the same NIFTY 25,000 Call.
Call · C / CE · European BSM inversion
Frozen: spot 25,000 · strike 25,000 · 30 calendar days · annual rate 6% · yield 1% · European BSM
Annual model volatility
Workings: trial values and fitted quotes
| Annual volatility | Model premium / unit |
|---|---|
| 14% | ₹452.44 |
| 18% | ₹565.99 |
| 22% | ₹679.70 |
| Quote | Premium / unit | Fitted annual IV |
|---|---|---|
| Bid | ₹550.00 | 17.4370% |
| Midpoint | ₹566.00 | 18.0002% |
| Ask | ₹582.00 | 18.5633% |
Spot/strike 25,000; 30/365 years; annual continuous rate 6%, yield 1%. ₹ per unit. A midpoint is not a promised fill.
Conventions matter. A spot-based model carries a dividend yield. A forward-based model puts interest and dividends into the forward price. Mixing the two distorts the answer. Even correct platforms differ by quote used, timestamp, rate, dividend assumption and exact expiry time. NSE shows its own reference IV in the option chain, and its notes describe a 10% interest-rate assumption. Our 6% rate is a teaching input. It is not an attempt to copy that display.
Some prices fit no volatility. Here the Call has a floor of about ₹102.44 even at zero volatility, because spot equals strike and interest and dividends still tilt the forward. A ₹100 target has no non-negative IV under these inputs. When the solve fails, check the quote and the inputs first. It does not mean you found an arbitrage. Where the model barely reacts to volatility, a tiny price change can also swing IV a lot.
How is IV different from realised volatility?
Realised volatility measures how much prices actually swung over a past window. It is also called historical volatility. No option quote is inverted. You pick a sampling interval and a method. Here we take daily closes, work out each close-to-close log return (the natural log of today's close over yesterday's), and take the sample standard deviation. Then we scale to a year using 252 trading sessions. That is a different convention from the 365-day one used for pricing above.
Take two hypothetical runs of six NIFTY closes. Both start and finish at 25,000. In the first, the middle closes sit only 50 points above or below 25,000. In the second they sit 500 points away. The start-to-finish change is zero in both. Yet the second run swung far more every day. The figure shows the full sequences, so you can see why the endpoint alone says nothing.
Same finish, different volatility
Two hypothetical six-close NIFTY paths return to 25,000 after very different daily swings.
Observed return paths
Observed session · synthetic daily closes
Quieter daily path
- Starting / ending close
- 25,000 / 25,000
- Endpoint return
- 0.00%
- Daily log-return observations
- 5
Wider daily path
- Starting / ending close
- 25,000 / 25,000
- Endpoint return
- 0.00%
- Daily log-return observations
- 5
Sample standard deviation of daily log returns × √252 trading sessions per year. This is observed-path arithmetic, not future option IV.
Inspect every synthetic close and daily log return
| Session | Close | Daily log return |
|---|---|---|
| 0 | 25,000 | Starting close |
| 1 | 25,050 | 0.1998% |
| 2 | 24,950 | −0.4000% |
| 3 | 25,050 | 0.4000% |
| 4 | 24,950 | −0.4000% |
| 5 | 25,000 | 0.2002% |
| Session | Close | Daily log return |
|---|---|---|
| 0 | 25,000 | Starting close |
| 1 | 25,500 | 1.9803% |
| 2 | 24,500 | −4.0005% |
| 3 | 25,500 | 4.0005% |
| 4 | 24,500 | −4.0005% |
| 5 | 25,000 | 2.0203% |
Five daily log returns; sample standard deviation × √252. A short teaching window, not a forecast. Endpoint change alone cannot measure realised volatility.
The two annualised readings are about 5.94% and 59.41%. Six closes give only five returns. These short paths show the arithmetic. They are not a reliable estimate. A different window or sampling step changes the answer, and daily closes miss intraday swings.
You can compare IV with realised volatility, but they answer different questions. IV comes from today's prices and carries expectations plus a price for bearing risk. Realised volatility describes one past path. Comparing them fairly needs matching periods and methods. A gap can inform your analysis. One outcome does not prove the option was mispriced. A small endpoint move also does not show that realised volatility was low, or that a straddle seller made money.
Same index, two different questions
Implied volatility looks forward from a quote. Realised volatility looks back along a path.
Hypothetical readings from this lesson. 18% is the Call IV; the other two come from the six-close paths. Different measures, not a spread to trade.
What does India VIX measure?
One 30-day reading across many NIFTY option quotes.
India VIX is NSE's volatility index, built from NIFTY option quotes. It looks 30 calendar days ahead and is quoted as an annual percentage. It is not the IV of one Call or Put, and it says nothing about a single stock. A reading of 20 means a 20% annualised volatility measure. It does not mean NIFTY will move 20% in 30 days.
The build starts with bid-ask midpoints across a strip of strikes in the monthly expiries. A weighted sum of those quotes gives a variance figure for each expiry. The variances are blended by time to reach 30 days, and the square root turns variance back into volatility. NSE uses the matching NIFTY futures price as the reference. This is a different calculation from solving BSM contract by contract and averaging the IVs.
The next figure starts after the quote-strip step. It supplies synthetic readings of 18% for 20 days and 22% for 40 days, and a 30-day target between them. We blend total variance, which is variance times time, then scale to 30 days. The result is about 20.75%, not the plain average of 20%. The 20 and 40 day points are chosen for convenience. They are not the NSE expiry calendar, and this is not a full India VIX calculator.
Combine variance over time
A simplified bracketed 30-day illustration using two synthetic maturity volatility measures.
One construction step only
Variance each expiry contributes to the 30-day target
Combining variance over time gives 20.7525%, not the simple 20% average of the two readings.
Interpolate total variance, normalise to 30 days, then take its square root. Not live data, the NSE expiry calendar or a complete India VIX calculator.
The published method uses the near and next monthly expiries, and rolls to next and far when three trading days remain. Time is counted in calendar minutes. The chosen expiries do not always sit on either side of 30 days. The method also drops unsuitable quotes, fills some missing ones from neighbouring strikes within limits, and carries the prior value when a monthly variance cannot be found. These rules show why quote quality matters, even though the method is called model-free.
Model-free means it avoids inverting one contract through BSM. It does not mean assumption-free, and it does not know future realised volatility. "Fear gauge" is a catchy nickname with no direction rule behind it. India VIX can rise without NIFTY falling. A calm VIX does not stop a stock gapping on its own news. Use it as background next to the contract you actually hold.
What should you check before trusting an IV number?
Look at the quote behind the number first.
First, name the contract, the price used, the timestamp and the model. Look at the bid and the ask, not only the last trade. Wide or stale quotes make IV a weak clue, especially when one price tick swings it a lot. A midpoint is something you can analyse. It is not evidence that anyone will fill there.
Second, ask which question you are on. One contract's IV is not the IV of the whole underlying. The surface, meaning the readings across strikes and expiries, moves in parts. India VIX summarises selected NIFTY quotes by its own rules. A high reading or a high historical rank does not prove an option is expensive for the risk ahead. Compare your own movement view, the event calendar, the right part of the surface and the price you can actually get.
Risk stays with the position. An uncovered short Call has unlimited theoretical loss as the underlying rises. An uncovered short Put loses a finite amount if the underlying falls to zero, but that can be large. Shorts can need extra margin before expiry, and a stop order cannot promise an exit through a gap. A bought option can lose at most its premium plus costs. Settlement still matters. NIFTY index options settle in cash. Stock options in India settle by delivery. An in-the-money long stock Call can need full payment for the shares, and an in-the-money long stock Put can need share delivery. A premium-limited loss does not remove those funding or delivery needs. For margin and pledging basics, see the Start lesson on margin and pledging.
You can now read IV without treating it as a promise. Check its units, check the quote behind it, keep it apart from realised volatility, and know what India VIX combines. The next lesson asks where a reading sits within its own past, through IV Rank and IV Percentile. Today's fair value stays a separate question.
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 implied volatility in an option chain?
It is the yearly volatility input that makes a chosen pricing model return a chosen option premium, given stated spot or forward, strike, time, interest and dividend inputs. It shows the movement scale priced into that contract. It is not a separate charge, a direction forecast or a promise of future realised volatility.
How is IV calculated from an option price?
Software solves the pricing model backwards for volatility, usually by numerical search. A valid positive answer needs a price inside the model bounds. The same quote can give a different IV if the model, timestamp, dividend or interest inputs change. Matching a quote is calibration. It is not independent proof of fair value.
Why are bid IV, midpoint IV and ask IV different?
Each inverts a different price while the other inputs stay fixed. In our hypothetical 30-day NIFTY Call, a ₹550 bid, ₹566 midpoint and ₹582 ask imply about 17.44%, 18.00% and 18.56%. A midpoint is not an assured fill, and its IV need not be the average of the other two.
What is the difference between IV and realised volatility?
IV is inferred from current option prices and reflects expectations and the price of risk. Realised volatility measures how much an observed price path swung, using a chosen window and method. Two paths can start and end at the same level with very different realised volatility. Daily closes also miss intraday moves.
What does India VIX at 20 mean?
It means NSE's quote-strip method gives a 20% annualised volatility measure for a constant 30-calendar-day horizon. It does not mean NIFTY will move 20% in a month. At a hypothetical NIFTY 25,000, a steady-rate illustration gives a rough 30-day scale of 1,433.46 points. That is not a guaranteed range or a direction call.
Is India VIX the same as my NIFTY option’s IV?
No. A contract IV comes from one selected quote and a pricing model. India VIX combines selected NIFTY Call and Put quotes across strikes and monthly expiries into one constant-horizon variance measure. It does not set any contract's IV, and a change in it does not by itself fix your option's premium change.
Does higher IV make both Calls and Puts more valuable?
For European BSM options with time left, higher own-contract volatility raises both values when the other inputs are fixed. That does not guarantee a higher market quote or a profit. Spot, time, the IV surface, entry premium, fills and costs also matter. A rise in India VIX alone cannot fix one contract's change.
Does high IV or IV Rank mean an option is overpriced?
A high reading only says where the measure sits against its own past. It does not set fair value for the risk ahead. Scheduled events, movement assumptions, strike and expiry differences, liquidity and executable prices still matter. A quiet outcome also cannot prove that an earlier volatility price was wrong.