The actual past volatility of an asset — your benchmark for judging whether implied volatility is cheap or expensive.
Historical Volatility (HV), also called Realized Volatility or Statistical Volatility, measures how much an asset's price actually moved over a past period. It is calculated from historical closing prices and expressed as an annualized percentage, just like implied volatility.
If Nifty's 20-day HV is 14%, it means Nifty's actual price fluctuations over the past 20 trading days, when annualized, equate to a 14% annual standard deviation. In practical terms, this means Nifty moved about 0.88% per day on average (14% / √252).
HV is entirely backward-looking. It tells you what happened, not what will happen. However, it serves as the most important benchmark for evaluating whether implied volatility (IV) is rich or cheap. If IV is 20% but HV is only 12%, options are pricing in significantly more movement than what has actually occurred — a potential selling opportunity.
Unlike IV which is derived from option prices and changes tick-by-tick, HV updates once daily when the closing price is recorded. It is a statistical measure, not a market price. This makes HV more stable and less prone to sudden spikes caused by order flow or liquidity events.
Step 1: Calculate daily log returns: ri = ln(Pi / Pi-1)
Step 2: Calculate the mean of log returns: r̄ = (1/n) × Σri
Step 3: Calculate standard deviation of log returns: σ
Step 4: Annualize by multiplying by √252 (trading days in a year)
n = Number of periods (e.g., 10, 20, or 30 trading days)
Pi = Closing price on day i
Suppose Nifty closed at these prices over 5 days: 24,400 | 24,520 | 24,480 | 24,600 | 24,550
Log returns: ln(24520/24400) = 0.0049 | ln(24480/24520) = -0.0016 | ln(24600/24480) = 0.0049 | ln(24550/24600) = -0.0020
Mean return: (0.0049 - 0.0016 + 0.0049 - 0.0020) / 4 = 0.00155
Variance: sum of squared deviations / (4-1) = 0.0000158
Daily σ = √0.0000158 = 0.00397 = 0.397%
Annualized HV = 0.397% × √252 = 6.3%
This very low HV reflects the small daily moves in our example. Real Nifty HV typically ranges from 10-20%.
The most responsive measure. Captures very recent price action — essentially the last two trading weeks. Useful for short-term traders and weekly option sellers. Prone to sharp spikes and drops. Best for comparing against near-term IV on weekly expiries.
The most popular period. Represents approximately one month of trading data. Balances responsiveness with stability. This is the standard benchmark used by most option platforms (Sensibull, Opstra) when displaying HV alongside IV. Ideal for monthly option analysis.
Slightly smoother than 20-day. Directly comparable to India VIX which measures 30-day expected volatility. When 30-day HV is 12% but India VIX is 18%, there is a 6% volatility risk premium — options are expensive relative to recent realized movement.
Longer-term measures that smooth out short-term noise. Useful for understanding the "normal" volatility regime. If 10-day HV is 20% but 90-day HV is 13%, the recent spike is likely temporary and HV will probably revert to the 13% range.
The relationship between HV and IV is the foundation of volatility trading. The difference (IV minus HV) is called the Implied Volatility Premium or Volatility Risk Premium (VRP). Understanding this premium is what separates profitable option traders from the rest.
IV almost always trades above HV. This gap (volatility premium) is what option sellers earn over time.
An HV Cone (also called Volatility Cone) plots the historical range of realized volatility across different lookback periods. It shows you the minimum, 25th percentile, median, 75th percentile, and maximum HV for each period. This creates a "cone" shape that helps you assess whether current HV is normal, high, or low relative to history.
If Nifty's current 20-day HV of 11% sits at the 20th percentile of its historical range, it means HV has been lower than 11% only 20% of the time. This is a low-volatility regime. If IV is at 16% (above the 50th percentile), the spread suggests options are expensive — a selling opportunity. Conversely, if HV is at the 80th percentile but IV is at the 40th percentile, options are cheap relative to actual movement.
Download Nifty daily closing prices for the past 3-5 years. Calculate rolling HV for periods of 10, 20, 30, 60, and 90 days. For each period, compute the min, 25th percentile, median, 75th percentile, and max. Plot these as bands. Overlay current HV values. Tools like Python with pandas make this straightforward, or use platforms like Opstra which provide pre-built cones.
Nifty is in a calm, trending market. Daily moves are under 100 points. Common during strong bull runs (late 2023, early 2024). Option premiums should be cheap. Buy protective puts and long-dated straddles at these levels — volatility will eventually return.
The most common regime for Nifty. Daily moves of 100-200 points. Weekly option premiums are fairly priced. Both buying and selling strategies can work. This is the median zone where no extreme bias is warranted.
Markets are volatile with 200-400 point daily swings. Typically seen during earnings season, global risk events, or sector rotations. Option premiums are rich. Selling strategies with defined risk (iron condors, credit spreads) work well if you time entries after initial spikes.
Crisis-level volatility. COVID crash saw 60-day HV reach 45%. Nifty moves 500+ points daily. Option premiums are extraordinarily expensive. These periods are rare (1-2% of the time) and typically short-lived. HV will revert — the question is when, not if.
HV is essential for backtesting option strategies because it tells you what actually happened — not what the market expected. When you backtest a short strangle, you need HV to determine whether Nifty's actual moves stayed within your breakevens.
You want to test: Sell 300-point OTM strangle on weekly Nifty options every Monday.
Step 1: Calculate 5-day HV for every week in the past 2 years.
Step 2: Convert weekly HV to expected range. If 5-day HV = 12%, expected weekly range = 12% / √52 × 24,500 = ±408 points.
Step 3: Check how many weeks the actual move exceeded 300 points.
Result: In 70% of weeks, Nifty moved less than 300 points. Your strangle would have been profitable 70% of the time before commissions.
By segmenting this by VIX level, you find that in weeks where VIX > 18, the strangle was profitable only 55% of the time (wider moves) but earned more premium per trade.
HV tells you what happened, not what will happen. A stock with 10% HV can spike to 30% HV overnight on a single news event. HV is a benchmark, not a forecast. Use HV as context for IV, not as a prediction of future moves.
Using 90-day HV to evaluate weekly options is inappropriate. The lookback period should match your trading timeframe. For weekly options, 10-day HV is more relevant than 90-day HV. Match HV period to your option's time to expiry.
Close-to-close HV ignores intraday moves. Parkinson (high-low) and Garman-Klass (OHLC) estimators capture more information and are typically 10-30% higher than close-to-close HV. Be consistent in your methodology and know which estimator your platform uses.
Low HV regimes are temporary. Volatility is mean-reverting. Extended periods of low HV (like Nifty in late 2023 at 8-9%) are often followed by sharp spikes. Low HV is a coiled spring. Use low HV as a signal to buy cheap options for protection or speculative plays.
Open your Demat account today and take the first step towards mastering the stock market.
Click Here to Get Started