How Much of Each Stock Should I Own? 4 Weighting Methods
Equal weight, inverse volatility, mean-variance optimization and risk budgeting compared on one sample portfolio, with the backtest record of each. Education, not a recommendation.
The biggest position is often not the riskiest. Here is how risk contribution works, how to estimate it with public data, and a worked example where a 15% holding drives 30% of the swings.
What is your own portfolio exposed to? Check it with Exent →
The one with the largest risk contribution: the share of the portfolio’s total volatility that comes from that holding. It depends on three things at once: its weight, its own volatility, and how much it moves with everything else you own. A small position in a volatile stock that moves with the rest of the portfolio can outweigh a large position in a calm index fund.
Dollar weights answer “where is my money?”. Risk contribution answers “where are my swings coming from?”. The two lists are often in a different order, and the gap is where to look first.
The full formula needs every pair’s covariance. A good shortcut uses only each holding’s volatility and its beta to the market: the shared market part plus each holding’s own risk. The definition is in the risk contribution glossary entry.
Sample portfolio, Exent data as of Sep 30, 2026 (5-year monthly volatility and beta), shares estimated with the one-factor shortcut:
| Holding | Weight | 5-year volatility | Beta | Share of portfolio risk | Risk ÷ weight |
|---|---|---|---|---|---|
| SPY (S&P 500 fund) | 40% | 15.7% | 1.00 | 29% | 0.7× |
| QQQ (Nasdaq-100 fund) | 25% | 20.9% | 1.22 | 23% | 0.9× |
| AAPL (Apple) | 20% | 24.7% | 1.07 | 18% | 0.9× |
| NVDA (Nvidia) | 15% | 50.1% | 2.21 | 30% | 2.0× |
| Portfolio | 100% | 20.8% (estimated) | 1.25 | 100% |
Nvidia is the smallest position and the largest source of risk, at about twice its weight. The shortcut ignores links between holdings beyond the market, so it likely understates how much the three tech-heavy holdings reinforce each other.
Because volatility and market sensitivity multiply. Five stocks at 20% each, same method and data:
| Stock (20% each) | 5-year volatility | Beta | Share of risk |
|---|---|---|---|
| NVDA | 50.1% | 2.21 | 46% |
| AAPL | 24.7% | 1.07 | 19% |
| JPM | 24.1% | 1.02 | 18% |
| XOM | 27.0% | 0.22 | 11% |
| JNJ | 17.6% | 0.21 | 6% |
Exxon Mobil was more volatile than Apple on its own, yet carried less of the portfolio’s risk, because it moved largely independently of the market and of the others. Low market sensitivity is why a volatile holding can still be a diversifier. Different ways to set weights from this are compared in how much of each stock to own.
No. Each measure answers a different question, and only risk contribution and crash share are about the portfolio rather than the holding alone:
| Measure | Question it answers | Blind spot |
|---|---|---|
| Weight | Where is my money? | Ignores how risky each holding is |
| Volatility | How much does this holding swing on its own? | Ignores how it moves with the rest |
| Beta | How much does it move with the S&P 500? | Ignores its own company-specific risk |
| Risk contribution | How much of my portfolio’s swings come from it? | Averages over calm and bad months |
| Crash liability | How much of my worst months’ loss came from it? | Few data points; the next crash may differ |
Using two of them together, one for everyday swings and one for crashes, gives a steadier picture than any single number. See volatility and tail risk.
Often, but not always. Volatility-based risk spreads across all months; a crash view looks only at the worst ones. Replaying each holding’s peak-to-trough drop in past crises on the four-holding sample:
| Share of loss | SPY (40%) | QQQ (25%) | AAPL (20%) | NVDA (15%) | Sample loss |
|---|---|---|---|---|---|
| 2008 financial crisis | 37% | 22% | 20% | 21% | −60.9% |
| 2018 trade war | 27% | 19% | 25% | 28% | −29.6% |
| COVID crash, 2020 | 42% | 22% | 19% | 17% | −32.6% |
| 2022 bear market | 30% | 26% | 17% | 27% | −33.8% |
| 2025 tariff selloff | 31% | 24% | 25% | 20% | −24.3% |
Holdings rarely hit bottom on the same day, so these sums are conservative. A related measure is crash liability: each holding’s share of the portfolio’s loss in its worst 5% of months, over 3, 5 and 8 years. The crash liability tool calculates it.
These estimates come from the past five years and correlations change, especially in crashes (see do correlations go to 1 in a crash). More under portfolio optimization.
It is how much a portfolio’s volatility changes for a small increase in one holding’s weight. Multiplied by the weight, it gives that holding’s risk contribution.
Not necessarily. In a sample portfolio, a 40% S&P 500 fund carried an estimated 29% of the risk while a 15% Nvidia position carried 30%.
Build the covariance matrix of your holdings’ returns, multiply it by the weight vector to get each holding’s covariance with the portfolio, then multiply by each weight and divide by the portfolio variance. The results are shares that add up to 100%.
Yes. A holding that tends to rise when the rest of the portfolio falls lowers total volatility, so its contribution is negative.
Risk contribution spreads volatility across all months. Crash liability, as Exent defines it, splits the loss in the portfolio’s worst 5% of months by holding.
Upload a screenshot of your holdings and get concentration, overlap, crash exposure and a risk-matched benchmark in about a minute.
Educational content, not investment advice. Examples are illustrative; past performance does not guarantee future results.