Portfolio construction

Modern portfolio theory

Modern portfolio theory is a framework for comparing portfolios by expected return and risk, using asset weights and correlations to identify combinations with the lowest modeled risk for a given expected return.

What it means

Modern portfolio theory (MPT), associated with Markowitz (1952), examines how assets work together rather than judging each holding in isolation. Its central idea is that portfolio risk depends not only on each asset’s volatility but also on how their returns move together. The combinations offering the highest modeled expected return at each level of modeled risk form the efficient frontier.

In the standard mean-variance framework, risk means the standard deviation of portfolio returns. Expected returns, volatilities and correlations are estimates, not known future outcomes. A portfolio can look diversified by number of stocks yet remain exposed to similar risks.

How it is calculated

For two assets, a common measure of modeled portfolio risk is:

σₚ = √(w₁²σ₁² + w₂²σ₂² + 2w₁w₂σ₁σ₂ρ₁₂)

  • σₚ is the standard deviation of portfolio returns, expressed as a percentage for the period measured.
  • w₁ and w₂ are the fractions of portfolio value in assets 1 and 2; for a fully invested, unleveraged two-asset portfolio, they sum to 1.
  • σ₁ and σ₂ are the assets’ return standard deviations for the same period, expressed as percentages.
  • ρ₁₂ is their return correlation, a unitless number from −1 to +1.

The corresponding expected return is E[Rₚ] = w₁E[R₁] + w₂E[R₂], where E[Rₚ] is the portfolio’s estimated return and E[R₁] and E[R₂] are the assets’ estimated returns, all for the same period. With more assets, the risk calculation uses the covariance of every pair of returns. An efficient frontier compares many feasible weight combinations using these inputs.

Worked example

Suppose, hypothetically, a US$10,000 portfolio holds US$5,000 in stock A and US$5,000 in stock B. Both weights are therefore 0.5. Assume annual return standard deviations of 20% and 10%, a correlation of zero, and estimated annual returns of 8% and 4%. These round numbers are assumptions, not forecasts.

  1. Calculate estimated annual return: 0.5 × 8% + 0.5 × 4% = 6%.
  2. Calculate annual variance using returns in decimal form: 0.5² × 0.20² + 0.5² × 0.10² + 2 × 0.5 × 0.5 × 0.20 × 0.10 × 0 = 0.0125.
  3. Take the square root: √0.0125 ≈ 0.1118, or 11.18% modeled annual standard deviation.

For comparison, a hypothetical 100% allocation to A would have 20% modeled annual standard deviation. The difference illustrates diversification under the stated assumptions; it does not establish that the two-stock portfolio is optimal or that losses are limited to 11.18%.

Why it matters before you trade

A proposed trade changes portfolio weights and may change exposure to risks shared across holdings. MPT provides a way to ask whether a new combination changes modeled portfolio-level risk, rather than considering only the stock being traded. Correlation matters: two individually volatile holdings can behave differently together depending on how closely their returns move.

The framework does not account for every practical issue. Expected returns and correlations are uncertain and can change; standard deviation does not describe the full size or likelihood of extreme losses. Trading costs and the ability to execute an order also matter before a theoretical allocation becomes an actual portfolio.

How PreTrAIde uses it

PreTrAIde is an AI pre-trade analysis platform for stock baskets, not a stated MPT optimizer or efficient-frontier calculator. A user specifies each stock’s buy or sell side, quantity and execution conditions. Before an order is placed, PreTrAIde returns basket- and line-level GO, NO-GO or MODIFY verdicts with model-estimated success probabilities, risk assessments, suggested execution strategies and plain-language reasoning. These verdicts are decision support, not recommendations.

Its pre-trade checks relate to portfolio-level questions through breakdowns by country, sector, region, currency and market capitalization, plus a Herfindahl-Hirschman concentration index. It also calculates volatility and execution-related measures. Those checks can highlight concentration or implementation considerations, but they should not be mistaken for MPT estimates of an optimal portfolio.

Common mistakes

  • Treating estimates as facts: Historical relationships and assumed returns may not persist.
  • Counting holdings instead of examining exposures: Several stocks can share sector or country risks.
  • Confusing volatility with a loss limit: Standard deviation is not the worst possible loss.
  • Ignoring execution: An allocation calculated without spreads and market impact may differ from what a trade can achieve.

Frequently asked questions

What is the efficient frontier in modern portfolio theory?

It is the set of portfolios with the highest modeled expected return for each level of modeled risk, given the assets, assumptions and constraints used. Changing those inputs can change the frontier.

Does modern portfolio theory guarantee diversification will prevent losses?

No. Diversification can reduce modeled risk when holdings do not move perfectly together, but correlations can change and a diversified portfolio can still lose value.

Does PreTrAIde calculate an MPT-optimal portfolio?

PreTrAIde is not described as an MPT optimizer. Its pre-trade analysis assesses a proposed stock basket, including concentration, volatility and execution-related measures, before the user places any order with their own broker.

Related terms

Efficient frontier Efficient frontier is the set of portfolios offering the highest estimated expected return for each level of volatility, or the lowest volatility for each estimated expected return, under specified assumptions. Diversification Diversification is the practice of spreading a portfolio across holdings with different sources of risk so that its outcome depends less on any single holding. Correlation Correlation is a unitless measure, from −1 to +1, of how closely two variables move together in a linear relationship. Covariance Covariance is a measure of how two variables, such as stock returns, move together relative to their respective averages. Portfolio concentration Portfolio concentration is the extent to which a portfolio’s value depends on a small number of holdings or on shared exposures such as a sector or country. Pre-trade analysis Pre-trade analysis is the assessment of a proposed stock order or basket before placement, estimating execution costs, liquidity and risk to inform the investor’s decision and execution approach.

Portfolio construction

Herfindahl-Hirschman index (HHI) Effective number of stocks Sector exposure Country exposure Market capitalization Large-cap stocks Mid-cap stocks Small-cap stocks All glossary terms →

Check your next trade before you place it

Build a basket, run a pre-trade analysis, and get a GO, NO-GO or MODIFY verdict with the cost, liquidity and risk behind it.

Start your free trial Try the free tools

Educational content, not investment advice. PreTrAIde is analysis software, not a broker: it never executes orders, and its verdicts are decision support, not recommendations to buy or sell.