Home / Risk Management Fundamentals / Portfolio Risk
Risk metricIntermediate

Portfolio Risk

Portfolio risk is the aggregate risk of all positions held together, which depends critically on the correlations between them, so that the combined risk is usually less than the sum of individual risks but can converge toward it in a crisis.

Quick answer: Portfolio risk is the aggregate risk of all positions held together, which depends critically on the correlations between them, so that the combined risk is usually less than the sum of individual risks but can converge toward it in a crisis.

In simple words

Portfolio risk is the total risk of everything you hold at once, not just the risk of each trade added up. What matters is how the positions move together: holding two things that rise and fall in sync is like doubling one bet, while holding things that move independently spreads the risk. The catch is that in a crisis, seemingly unrelated positions often fall together, so the diversification you counted on can vanish exactly when you need it.

Purpose

This page explains how the risk of a whole book depends on correlation, why diversification reduces but never fully removes risk, and why aggregate exposure must be managed at the portfolio level.

Visual explanation

Portfolio Risk

Capital spread across positions whose combined risk depends on how correlated their returns are.

Portfolio Allocation & Limits18%Position A16%Position B14%Position C12%Position D40%Cash reserve100% of capital — no single position dominatesPer-position limits + a cash reserve cap concentration and tail risk

Professional explanation

Why portfolio risk is not the sum of parts

The risk of a portfolio is not simply the total of each position's risk, because positions partly offset or reinforce one another depending on how their returns move together. When returns are less than perfectly correlated, some losses coincide with gains elsewhere, so the combined volatility is lower than the sum of the individual volatilities. This is the mathematical basis of diversification: the whole is less risky than its parts precisely because they do not all move in lockstep. Managing risk one trade at a time, while ignoring how the trades combine, systematically understates or misjudges the true exposure of the book.

Correlation is the key variable

Correlation, ranging from +1 to −1, measures how two positions move relative to each other. At +1 they move identically and provide no diversification, so two correlated longs are effectively one double-sized bet. At 0 they move independently and combining them reduces relative risk. At −1 they move oppositely and can hedge each other. The portfolio's risk depends heavily on the average correlation among its positions, which means a book of a dozen highly correlated stocks is far riskier than its count of positions suggests. Estimating and monitoring correlation is therefore central to portfolio risk, not an academic nicety.

The portfolio volatility formula

For two positions the combined variance is σp² = w1²σ1² + w2²σ2² + 2 w1 w2 ρ σ1 σ2, where the correlation term ρ determines how much the risks offset or add. The same logic generalises to many positions through a covariance matrix. The essential insight is the cross term: as correlation ρ rises toward 1, the combined risk approaches the simple sum, and as it falls toward −1, risk can shrink dramatically. This formula makes precise why adding a position that is uncorrelated with the book can reduce total risk even though it adds a new source of loss, and why adding a correlated one barely helps.

Correlation instability and crisis convergence

The dangerous limitation of diversification is that correlations are not stable. In calm markets a set of positions may look nicely uncorrelated, but in a crisis, when liquidity dries up and everyone de-risks at once, correlations across risk assets tend to spike toward +1. Diversification that was measured in normal conditions therefore evaporates precisely when it is most needed, and a book that looked well spread suffers a coordinated drawdown. Prudent portfolio risk management assumes correlations will worsen under stress and stress-tests the book against a scenario where the assumed diversification fails.

Aggregate exposure and portfolio heat

Beyond correlation, a portfolio must control total exposure, sometimes called heat: the sum of capital at risk across all open positions if their stops are hit. Even well-diversified positions can, together, put too much of the account at risk, so a heat limit, for example no more than 6 percent of capital at risk across all trades at once, caps the worst-case coordinated loss. Concentration limits prevent any single name, sector or theme from dominating. Managing heat and concentration at the book level is what stops a collection of individually reasonable trades from combining into an unreasonable total risk.

Hidden correlation and factor exposure

Positions can be correlated through shared factors that are not obvious from their names. Several different Indian stocks may all be exposed to the same interest-rate, currency or index factor, so a book that looks diversified by ticker is actually a concentrated bet on one factor. Option positions add another layer, since many short-volatility positions across different underlyings all lose together when volatility spikes. True portfolio risk management looks through the individual instruments to the underlying factor exposures, because that is where hidden concentration, and the coordinated losses it produces, actually lives.

Formula

σp = √(w1²σ1² + w2²σ2² + 2 w1 w2 ρ σ1 σ2)

σp = portfolio standard deviation (risk); w1, w2 = weights of each position as a fraction of capital; σ1, σ2 = standard deviations (volatility) of each position; ρ = correlation between the two positions, from −1 to +1. The cross term 2 w1 w2 ρ σ1 σ2 is the diversification lever: as ρ approaches +1 combined risk approaches the simple sum, and as ρ falls toward −1 risk can shrink sharply. Generalises to many positions via a covariance matrix.

Naive risk view vs portfolio risk view

AspectNaive viewPortfolio view
Total riskSum of each trade's riskDepends on correlations between trades
DiversificationMore positions is saferOnly uncorrelated positions diversify
Hidden linkIgnoredShared factors create concealed concentration
Crisis behaviourAssumed stableCorrelations spike toward one, diversification fails
ControlPer-trade stops onlyPortfolio heat and concentration limits

Practical example

Illustrative example (Indian market)

A trader with Rs 5,00,000 holds equal Rs 2,50,000 exposures in two positions, each with 20 percent volatility. If they are uncorrelated (ρ = 0), portfolio volatility is √(0.5²×0.2² + 0.5²×0.2²) = √(0.02) ≈ 14 percent, meaningfully below the 20 percent of either alone. If they are perfectly correlated (ρ = 1), portfolio volatility is the full 20 percent, no benefit at all. Now suppose both are Nifty and Bank Nifty longs, which usually move together with correlation near 0.9; the realised diversification is small, and in a sharp sell-off the correlation approaches 1, so the book behaves like a single double-sized index bet. The lesson is that the count of positions overstated the diversification the trader actually had.

A retail book of five Nifty-heavy large-cap stocks plus long Nifty futures looks like six positions but is close to one leveraged index bet, because all six share the same market factor. In a broad NSE sell-off they fall together, and any per-trade stops trigger at once, producing a coordinated drawdown the position count never suggested.

Limitations

  • Correlations are estimated from history and are unstable, especially in crises
  • Diversification reduces but never eliminates risk, and fails when correlations spike
  • The covariance matrix grows complex and noisy as positions multiply
  • Hidden factor exposures make a book look more diversified than it is
  • Volatility-based risk understates fat-tailed, coordinated crash losses

Common mistakes

  • Counting positions as diversification while ignoring their correlation
  • Assuming calm-market correlations will hold during a crisis
  • Managing risk only per trade and never at the portfolio level
  • Holding several instruments that share one hidden factor as if independent
  • Ignoring total portfolio heat, so many small trades add to a large aggregate risk
  • Treating short-volatility positions across underlyings as unrelated bets

Professional usage

Institutional risk managers monitor the whole book, not just individual trades. They estimate a covariance matrix, decompose exposure into underlying factors to find hidden concentration, cap aggregate heat and single-name or single-factor exposure, and stress-test the portfolio under scenarios where correlations spike toward one. They treat diversification as a fragile benefit that must be verified under stress rather than assumed, and they size the book so that a correlated crisis remains survivable.

Key takeaways

  • Portfolio risk depends on correlation, not just the sum of individual risks
  • Only genuinely uncorrelated positions provide diversification
  • Correlations spike toward one in a crisis, so diversification fails when needed most
  • Manage aggregate heat and hidden factor exposure at the book level

Frequently asked questions

What is portfolio risk?
Portfolio risk is the combined risk of all positions held together, which depends on how correlated the positions are rather than the simple sum of their individual risks. When positions are less than perfectly correlated the combined risk is lower than the sum, but the benefit shrinks as correlation rises.
Why isn't portfolio risk just the sum of each trade's risk?
Because positions partly offset or reinforce each other depending on how their returns move together. When returns are imperfectly correlated, some losses coincide with gains elsewhere, so combined volatility is less than the total of individual volatilities. Adding risks directly ignores this and misjudges the true exposure.
What is correlation in a portfolio?
Correlation measures how two positions move relative to each other, from +1 (identical moves, no diversification) through 0 (independent) to −1 (opposite moves, hedging). The average correlation among positions largely determines portfolio risk, so a book of highly correlated positions is far riskier than its count suggests.
How does diversification reduce risk?
Diversification reduces risk by combining positions whose returns are not perfectly correlated, so their losses do not all occur together. Mathematically the cross term in the portfolio variance formula shrinks combined risk when correlation is below one, which is why uncorrelated positions lower total volatility.
Why does diversification fail in a crisis?
Because correlations are unstable and tend to spike toward +1 in a crisis, when liquidity dries up and participants de-risk simultaneously. Positions that looked uncorrelated in calm markets then fall together, so the diversification measured in normal conditions evaporates exactly when it is most needed.
What is portfolio heat?
Portfolio heat is the total capital at risk across all open positions if their stops are hit. A heat limit, such as no more than about 6 percent of capital at risk at once, caps the worst-case coordinated loss, preventing a collection of individually reasonable trades from combining into an unreasonable total risk.
What is hidden correlation?
Hidden correlation is when positions that look different are linked through a shared factor, such as several stocks all exposed to the same index, interest-rate or currency factor. A book diversified by name can be a concentrated factor bet, so true risk management looks through instruments to their underlying factor exposures.
How do I calculate portfolio risk for two positions?
Use σp = √(w1²σ1² + w2²σ2² + 2 w1 w2 ρ σ1 σ2), where w are the weights, σ the volatilities and ρ the correlation. The cross term determines the diversification benefit: as ρ approaches one the risk approaches the simple sum, and as it falls the combined risk shrinks.
Is holding more positions always safer?
No. More positions only reduce risk if they are genuinely uncorrelated. Holding a dozen stocks that all track the same index provides little diversification, because they share one factor and fall together, so the count of positions overstates the real risk reduction.
How are options positions correlated in a portfolio?
Beyond price correlation, option positions share exposure to volatility, so many short-volatility positions across different underlyings all lose together when volatility spikes. A book of short options on various names can therefore be a concentrated short-volatility bet despite looking spread across instruments.
What is concentration risk?
Concentration risk is having too much exposure to a single name, sector, theme or factor, so that one adverse event causes an outsized loss. Concentration limits at the portfolio level cap how much any single exposure can dominate, complementing correlation management in controlling aggregate risk.
How do professionals manage portfolio risk?
They monitor the whole book with a covariance matrix, decompose exposure into factors to reveal hidden concentration, cap aggregate heat and single-name or single-factor exposure, and stress-test under scenarios where correlations spike toward one. They treat diversification as a fragile benefit to be verified under stress, not assumed.
Does volatility fully capture portfolio risk?
No. Volatility-based measures assume roughly normal behaviour and understate fat-tailed, coordinated crash losses where correlations converge. They are useful for everyday risk but must be supplemented by stress tests and scenario analysis that account for the tail events volatility misses.
How many uncorrelated positions do I need to diversify?
The biggest reduction in relative risk comes from the first several genuinely uncorrelated positions, after which each extra one helps less. Because true independence is rare and shared factors lurk, a modest number of genuinely uncorrelated exposures diversifies more than a long list of positions that secretly move together.

Voice search & related questions

Natural-language questions people ask about Portfolio Risk.

What is portfolio risk?
It is the total risk of everything you hold at once. It is not just each trade's risk added up; it depends on how much your positions move together.
Does holding more positions make me safer?
Only if they move independently. If you hold ten stocks that all follow the same index, it is close to one big bet, not ten separate ones.
What is correlation in simple terms?
It is how much two positions move together. If they rise and fall in sync they add up like one bigger bet; if they move independently they spread your risk.
Why does diversification fail in a crash?
Because in a panic almost everything falls together. The independence you counted on disappears just when you need it, so a spread-out book can still drop hard.
What is portfolio heat?
It is the total amount of your capital at risk across all your open trades at once. Capping it stops many small trades from adding up to one huge risk.
Can two different stocks be secretly linked?
Yes. If they share a driver, like the same index or interest rates, they move together even though the names look different, so your book is less diversified than it seems.

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
    Trading Psychology and Risk
    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.