Home / Risk Metrics / Portfolio Volatility
Portfolio metricIntermediate

Portfolio Volatility

Portfolio volatility is the standard deviation of a portfolio's returns, and unlike the volatility of a single asset it depends not only on each position's own volatility and weight but critically on the correlations between them.

Quick answer: Portfolio volatility is the standard deviation of a portfolio's returns, and unlike the volatility of a single asset it depends not only on each position's own volatility and weight but critically on the correlations between them.

In simple words

Portfolio volatility measures how much the whole account's value swings, not just one position. The surprise is that it is usually less than the average of the individual volatilities, because when holdings are not perfectly correlated their moves partly cancel out. Correlation is the lever: two assets that move together add up to a jumpy portfolio, while two that move independently smooth each other out. This is the mathematical heart of why diversification reduces risk.

Purpose

This page shows how the volatility of a portfolio is built from the weights, individual volatilities and correlations of its holdings, so a trader can see why diversification lowers risk and why correlation, not just count of positions, decides the result.

Visual explanation

Portfolio Volatility

A correlation matrix of holdings: the pairwise correlations, more than the individual volatilities, determine how much they diversify each other.

Correlation MatrixNiftyNiftyBankNiftyBankNiftyITITGoldGold1.000.820.55-0.200.821.000.48-0.150.550.481.00-0.05-0.20-0.15-0.051.00Low or negative correlation is what diversification needs

Professional explanation

Portfolio volatility is not a simple average

A beginner assumes that combining a 20 percent-volatility asset with a 20 percent-volatility asset gives a 20 percent-volatility portfolio, but that is only true if the two are perfectly correlated. In every other case the portfolio volatility is lower, because the two return streams do not move in lockstep and their deviations partly offset. The exact result depends on the weights, the individual volatilities and the correlation between the assets, combined through the variance formula rather than by simple averaging. This gap between the naive average and the true figure is precisely the diversification benefit, and it grows as correlation falls.

The two-asset formula and what each term does

For two assets the portfolio variance is w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρ₁₂σ₁σ₂, and the portfolio volatility is the square root of that. The first two terms are each position's own contribution, scaled by the square of its weight. The third, cross, term carries the correlation ρ₁₂, and it is the only place correlation enters. When ρ is +1 the whole expression collapses to a weighted average of the two volatilities; when ρ is 0 the cross term vanishes and only the squared terms remain; when ρ is negative the cross term subtracts, pulling portfolio volatility below either asset's own. This single term is why correlation is the dominant driver.

Generalising to many assets

With more than two holdings the same logic scales: portfolio variance is the double sum over every pair, Σᵢ Σⱼ wᵢ wⱼ ρᵢⱼ σᵢ σⱼ, which includes each asset's own variance when i equals j and every pairwise covariance otherwise. As the number of positions grows, the number of correlation terms grows far faster than the number of variance terms, so in a large portfolio the average correlation between holdings matters more than any single position's own volatility. This is why a portfolio of many highly correlated positions is barely more diversified than one position, and why measuring correlation across the book is essential rather than optional.

Correlation is the driver, and it is unstable

Because correlation sits in every cross term, it is the variable that most changes portfolio volatility, and it is also the least stable. Correlations estimated in calm markets routinely converge toward +1 in a crisis, when assets that normally move independently all fall together as investors sell everything for cash. A portfolio that looked well diversified on historical correlations can therefore become far more volatile exactly when it matters most. Any portfolio-volatility figure is only as reliable as the correlation estimates behind it, and those estimates are the first thing to break under stress.

From volatility to a rupee risk figure

Portfolio volatility is a percentage of the portfolio's value per unit time, so to make it concrete you multiply by the portfolio value and, if needed, scale by the square root of the horizon in days. A portfolio worth ₹5,00,000 with a daily volatility of 1.2 percent has a one-day standard deviation of about ₹6,000, and over a five-day week roughly ₹6,000 times the square root of five, about ₹13,400, assuming returns are independent day to day. This scaling assumes returns are roughly normal and serially uncorrelated, assumptions that understate real tail moves, so the figure is a typical swing, not a worst case.

Why it underpins Sharpe, VaR and sizing

Portfolio volatility is the denominator of the Sharpe ratio, the σ in parametric Value at Risk, and the quantity volatility-based position sizing tries to hold constant. Get it wrong, usually by assuming correlations are lower than they turn out to be, and every one of those downstream measures understates risk. This is why portfolio volatility is a foundational metric: it is not just a number to report but an input that propagates into sizing, capital allocation and tail-risk estimates across the whole risk framework.

Formula

σ_p = √(w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρ₁₂σ₁σ₂); general: σ_p = √(Σᵢ Σⱼ wᵢ wⱼ ρᵢⱼ σᵢ σⱼ)

σ_p = portfolio volatility, the standard deviation of portfolio returns. w₁, w₂ = the weights of assets 1 and 2 as fractions of the portfolio, which sum to 1 across all holdings. σ₁, σ₂ = the individual volatilities (standard deviations of returns) of the two assets. ρ₁₂ = the correlation between the two assets' returns, ranging −1 to +1. In the general form the double sum runs over every pair i, j; the term with i = j contributes wᵢ²σᵢ², and ρᵢⱼ is the correlation between assets i and j.

How correlation changes a two-asset portfolio's volatility

Correlation ρEffect on portfolioDiversification benefit
+1 (perfect positive)Volatility equals the weighted average of the twoNone; moves reinforce each other
0 (independent)Cross term vanishes; volatility falls below the averageMeaningful; moves partly offset
−1 (perfect negative)Volatility can fall toward zero at the right weightsMaximum; moves cancel
Rises in a crisisVolatility jumps above the estimateDiversification fails when needed most

Practical example

Illustrative example (Indian market)

A trader holds ₹2,50,000 in a Nifty position and ₹2,50,000 in a Bank Nifty position, so each has weight 0.5 of a ₹5,00,000 portfolio. Suppose Nifty's daily volatility is 1.0 percent and Bank Nifty's is 1.6 percent, and their correlation is 0.8, high because both are Indian equity indices. Portfolio variance is 0.5²×0.010² + 0.5²×0.016² + 2×0.5×0.5×0.8×0.010×0.016, which is 0.000025 + 0.000064 + 0.000064 = 0.000153, so σ_p is √0.000153 ≈ 1.24 percent, about ₹6,180 a day on ₹5,00,000. Note this sits below the 1.3 percent simple average of the two volatilities, but only slightly, because the 0.8 correlation leaves little room to diversify. Had the correlation been 0.2, portfolio volatility would fall to about 1.03 percent, showing that correlation, not the number of positions, drives the result.

Most retail F&O books are stacked in Nifty, Bank Nifty and index-heavy stocks, which are all highly correlated Indian equity exposures. Holding five such positions feels diversified but behaves almost like one large position, because the pairwise correlations sit near 0.8 to 0.9, so the portfolio volatility is close to that of a single concentrated bet.

Advantages

  • Captures the true swing of the whole account, not one position in isolation
  • Quantifies the diversification benefit that correlation creates
  • Feeds directly into Sharpe ratio, Value at Risk and volatility-based sizing
  • Reveals hidden concentration when many positions are secretly correlated
  • Converts to a concrete rupee daily swing by scaling by portfolio value

Limitations

  • Blind spot: as a symmetric standard deviation it treats upside and downside swings alike and says nothing about the fat tail beyond a normal move
  • Depends entirely on correlation estimates, which are unstable and rise in crises
  • Assumes returns are roughly normal, understating extreme moves
  • Backward-looking: computed from history that may not describe the next regime
  • Time-scaling by the square root of horizon assumes independent daily returns

Why it matters in practice

  • Decides how much diversification actually reduces account risk
  • A wrong (too low) estimate makes every downstream risk figure understate danger

Common mistakes

  • Averaging the individual volatilities instead of using the variance formula
  • Assuming positions are diversified when their correlation is near +1
  • Using calm-period correlations that collapse toward 1 in a sell-off
  • Treating a portfolio of many correlated index trades as low risk
  • Confusing portfolio volatility with the worst case rather than a typical swing
  • Scaling to longer horizons while ignoring that tails fatten with time

Professional usage

Risk desks compute portfolio volatility continuously from a full covariance matrix rather than trusting position counts, and they stress it by shocking correlations toward one to see how the book behaves when diversification fails. They treat the reported figure as a lower bound that assumes a benign regime, size positions to hold portfolio volatility within a target band, and monitor how much each new trade adds at the margin. Above all they distrust any low volatility that rests on historically low correlations, because those are exactly the estimates that break in a crisis.

Key takeaways

  • Portfolio volatility is the standard deviation of the whole portfolio's returns
  • It is built from weights, individual volatilities and, crucially, correlations
  • Correlation is the dominant driver: low correlation means real diversification
  • Correlations rise toward one in crises, so the calm-market figure understates risk

Frequently asked questions

What is portfolio volatility?
Portfolio volatility is the standard deviation of the returns of an entire portfolio. It measures how much the total account value swings, and unlike a single asset's volatility it depends on the weights, individual volatilities and correlations of all the holdings combined.
Why isn't portfolio volatility just the average of the parts?
Because the holdings do not move in perfect lockstep. Unless every pair is perfectly correlated, their deviations partly offset, so the portfolio's volatility falls below the weighted average of the individual volatilities. That gap is the diversification benefit, and it grows as correlation falls.
How does correlation affect portfolio volatility?
Correlation enters through the cross term of the variance formula. At a correlation of +1 the portfolio volatility equals the weighted average; at 0 the cross term vanishes and it falls; at −1 it can fall toward zero. Correlation is the single most powerful driver of the result.
What is the formula for two-asset portfolio volatility?
The portfolio variance is w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρ₁₂σ₁σ₂, and the volatility is its square root. Here w are the weights, σ the individual volatilities and ρ₁₂ the correlation between the two assets' returns.
Does adding more positions always reduce volatility?
Only if the new positions are not highly correlated with the existing ones. Adding several positions that all move together barely reduces volatility, because the pairwise correlations keep the cross terms large. Diversification depends on low correlation, not on the number of holdings.
Why does diversification fail in a crisis?
Because correlations that look low in calm markets converge toward +1 when investors sell everything for cash at once. The cross terms then grow, portfolio volatility jumps, and a book that appeared diversified behaves like a single concentrated position exactly when it hurts most.
How do I turn portfolio volatility into a rupee figure?
Multiply the volatility by the portfolio value for a one-period swing. A 1.2 percent daily volatility on ₹5,00,000 is about ₹6,000 a day. To scale to a longer horizon you multiply by the square root of the number of periods, assuming returns are independent.
Is portfolio volatility the same as risk?
It is one measure of risk, capturing typical two-sided swings, but it is not the whole of risk. As a symmetric standard deviation it treats gains and losses alike and ignores the fat tail of extreme losses, so it should be read alongside drawdown, Value at Risk and tail measures.
What correlation is typical between Nifty and Bank Nifty?
As two large Indian equity indices with overlapping constituents and shared macro drivers, Nifty and Bank Nifty are usually highly correlated, often around 0.8 or higher. That leaves limited room to diversify one against the other, so a book holding both is less diversified than it looks.
How does portfolio volatility relate to the Sharpe ratio?
Portfolio volatility is the denominator of the Sharpe ratio, which divides excess return by volatility. If you underestimate portfolio volatility, usually by assuming correlations are too low, you overstate the Sharpe ratio and understate the risk taken to earn the return.
Can portfolio volatility be lower than every individual asset's volatility?
Yes. With low or negative correlations, the offsetting moves can push portfolio volatility below the volatility of any single holding. This is the strongest form of the diversification benefit and is most pronounced when assets are negatively correlated.
Is portfolio volatility backward or forward looking?
As computed, it is backward looking: it uses historical volatilities and correlations. It is used to forecast future risk, but that forecast is only valid if the future regime resembles the past, which is often untrue around shocks, so it should be stress-tested rather than trusted blindly.
How is portfolio volatility used in position sizing?
Volatility-based sizing adjusts position sizes so that each contributes a similar amount of risk and the total portfolio volatility stays within a target. When an asset's volatility or its correlation with the book rises, its prudent size falls, keeping the portfolio's overall swing controlled.
Why square the weights in the formula?
Because variance, not volatility, adds cleanly, and variance scales with the square of the weight. The formula is built in variance space, using squared weights and squared volatilities plus the covariance cross terms, and only at the end is the square root taken to return to volatility units.

Voice search & related questions

Natural-language questions people ask about Portfolio Volatility.

What is portfolio volatility in simple terms?
It is how much your whole account value swings up and down, taking all your positions together, not just one trade on its own.
Why is my portfolio less jumpy than my individual trades?
Because your positions do not all move the same way at the same time. When they are not perfectly linked, their ups and downs partly cancel out, smoothing the whole account.
What matters most for portfolio risk?
Correlation. How closely your positions move together matters more than how many you hold. Low correlation gives real diversification; high correlation does not.
Does holding many index positions make me diversified?
Not really, if they are all Indian index trades like Nifty and Bank Nifty. They move together, so five of them can behave almost like one big position.
Why does diversification stop working when markets crash?
Because in a panic almost everything falls together as people rush to cash. Correlations shoot up, and positions that used to offset each other now drop at the same time.
How do I turn volatility into rupees?
Multiply the percentage by your account size. A one point two percent daily swing on five lakh rupees is about six thousand rupees a day.

Sources & references

    Last reviewed 12 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.

    ← Previous
    Ulcer Index
    Educational content only — not investment advice. Examples use illustrative numbers and simplified models. Risk-management techniques reduce but never remove risk, and trading derivatives involves substantial risk of loss. See our Risk Disclosure and SEBI Disclaimer.