What Is Modern Portfolio Theory (MPT)? Markowitz Explained

Search Fintrev

Search our guides, reviews and glossary.

Modern portfolio theory (MPT) is a mathematical framework for building a portfolio that delivers the highest expected return for a given level of risk, or the lowest risk for a given expected return. Its central idea is that no investment should be judged on its own: what matters is how each asset changes the risk and return of the whole portfolio. American economist Harry Markowitz introduced it in his 1952 paper “Portfolio Selection” in the Journal of Finance. That is why the approach is also called Markowitz portfolio theory, the Markowitz model or mean-variance analysis.

MPT measures return as the expected (average) return and risk as variance, or its square root, standard deviation. Because assets rarely move in perfect lockstep, combining them produces a portfolio less volatile than the weighted average of the individual volatilities, so diversification lowers risk more than it lowers return.

Who Created Modern Portfolio Theory?

In his 1990 Nobel lecture, Markowitz said the basic concepts of portfolio theory came to him one afternoon in a library while he was reading John Burr Williams’s Theory of Investment Value, which argued that a stock is worth the present value of its future dividends. Taken literally, that implies an investor should put everything into the single security with the highest expected return. Real investors diversify, so they must care about risk as well as return; what they lacked was a method for deciding how much to diversify and in what combination. Markowitz’s key realization was that a portfolio’s volatility depends not only on the volatility of its components but on the extent to which they move up and down together.

The idea was unusual enough at the time that when Markowitz defended his dissertation at the University of Chicago in 1955, his former teacher Milton Friedman remarked that it was not economics. Recognition came later. By the 1960s the work was gaining wide acceptance, and in 1990 Markowitz shared the Nobel Memorial Prize in Economic Sciences with William F. Sharpe and Merton Miller. Paul Samuelson summed up its influence by saying that Wall Street stands on the shoulders of Harry Markowitz.

How Modern Portfolio Theory Works

MPT rests on a few assumptions. Investors are risk-averse: offered two portfolios with the same expected return, they choose the less volatile one. They evaluate portfolios over a single period using two numbers, expected return and variance. They will accept more risk only in exchange for a higher expected return.

Expected return of a portfolio

Portfolio expected return is the weighted sum of each asset’s expected return:

E(Rp) = w₁E(R₁) + w₂E(R₂) + … + wₙE(Rₙ)

where w is the share of the portfolio in each asset and the weights add up to 100% in a fully invested portfolio. With four equally weighted assets expected to return 4%, 6%, 10% and 14%, the portfolio’s expected return is (4% + 6% + 10% + 14%) × 25% = 8.5%. Expected return is an estimate of the average outcome, not a forecast for any single year.

Risk of a portfolio

Risk is where MPT departs from simple averaging. For two assets, portfolio variance is:

σp² = w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρ₁₂σ₁σ₂

Here σ is each asset’s standard deviation and ρ₁₂ is the correlation between them, which ranges from +1 (perfectly in step) to −1 (perfectly opposite). The last term carries the diversification effect. When correlation is +1, portfolio standard deviation equals the weighted average of the two standard deviations. Below +1, it is lower. With a correlation of −1, an appropriate mix can in theory cancel volatility entirely.

With more assets, every pair needs its own correlation (or covariance). A portfolio of n assets needs n variances plus n(n−1)/2 pairwise correlations: 6 for four assets, 45 for ten, and 124,750 for 500. That count explains why Markowitz’s method needed computers to become practical, and why later models tried to simplify it.

Example of Modern Portfolio Theory

Hypothetical example, assumptions chosen only to show the mechanics. Asset A, a stock fund, has an expected return of 10% and a standard deviation of 20%. Asset B, a bond fund, has an expected return of 4% and a standard deviation of 6%. The portfolio holds 60% A and 40% B.

Expected return: (0.6 × 10%) + (0.4 × 4%) = 7.6%, whatever the correlation.

Standard deviation depends on the assumed correlation:

At ρ = +1 the result is exactly the weighted average (0.6 × 20% + 0.4 × 6% = 14.4%), so diversification adds nothing. At lower correlations, the same 7.6% expected return comes with noticeably less volatility. That is the “free lunch” of diversification: risk falls without any change to expected return. The investor can keep the lower risk or shift weight back toward Asset A to raise expected return at the original risk level.

The same logic explains the classic case of adding government bonds to a stock portfolio: when stocks and Treasuries have moved in opposite directions, a modest bond allocation cut portfolio volatility far more than it reduced expected return. The reverse also works on paper. A small slice of small-cap value stocks in a Treasury portfolio can raise expected return and, if those stocks tend to do well when bonds do poorly (for example in some inflationary periods), lower overall volatility too. Correlations shift over time, though, which matters for the limitations below.

The Efficient Frontier

Every possible combination of a set of assets can be plotted on a chart with risk (standard deviation) on the horizontal axis and expected return on the vertical axis. The combinations form a region, and its upper-left edge is the efficient frontier: an upward-sloping curve of portfolios that offer the highest expected return for each level of risk.

For example, Portfolio A has an expected return of 8.5% and a standard deviation of 8%; Portfolio B has the same 8.5% return with a 9.5% standard deviation. A is more efficient because it delivers the same return for less risk. B lies below or to the right of the frontier, a suboptimal zone where some frontier portfolio either earns more at the same risk or carries less risk for the same return.

No single frontier portfolio is best for everyone: the frontier narrows the choice to efficient options, and the investor’s risk tolerance picks a point on the curve. The frontier also depends on which assets are included and on the return, volatility and correlation estimates used to draw it, so different inputs produce different frontiers.

A later extension added a risk-free asset such as Treasury bills. Combining it with one particular risky portfolio on the frontier (the tangency portfolio) produces a straight line that beats the rest of the curve. The tangency portfolio is the one with the highest return per unit of risk above the risk-free rate, which is what the Sharpe ratio measures.

How Modern Portfolio Theory Is Used

Asset allocation. Institutions, advisors and individual investors use mean-variance analysis to set the mix of stocks, bonds, real estate and other asset classes. For an individual, applying MPT means assessing risk tolerance first, then spreading money across asset classes with low correlations to each other, instead of picking whatever looks likely to rise most.

Funds that do part of the work. Broad index funds, diversified exchange-traded funds (ETFs) and target-date mutual funds package diversification into a single holding, so asset classes that once needed separate accounts can be bought cheaply in one brokerage account.

Robo-advisors. Many automated platforms build suggested portfolios from MPT principles. Markowitz himself described them as a way to bring advice to the masses, adding that they can give good or bad advice.

The basis for later models. MPT laid the groundwork for the Capital Asset Pricing Model (CAPM) and for risk-adjusted performance measures such as the Sharpe ratio and alpha.

Limitations of Modern Portfolio Theory

Variance treats all volatility as bad. The most cited criticism is that MPT measures risk as variance, which counts upside surprises the same as losses. Two portfolios with equal return and variance look identical, yet one might get there through frequent small losses and the other through rare, severe crashes. Most investors would find the first far easier to live with.

Returns are not neatly distributed. Mean and variance fully describe a normal distribution. Real returns often show fat tails and skew, so extreme losses can occur more often than the model implies.

Inputs are estimates. Expected returns, volatilities and correlations are usually drawn from historical data, and small changes in them can swing the “optimal” weights sharply. Correlations are also unstable: assets that looked independent can fall together in a market-wide selloff, exactly when diversification is needed most.

Diversification cannot remove systematic risk. Combining assets reduces idiosyncratic risk, which belongs to a single company, sector or asset class, but does little against risks that hit the entire market or financial system. In their 2021 book Moving Beyond Modern Portfolio Theory: It’s About Time!, Jon Lukomnik and James Hawley call this “the MPT paradox” and argue that systemic risks such as climate change, antimicrobial resistance and resource scarcity matter more to returns than the security-level risks MPT handles.

No fixed answer on how much to hold. The theory gives no absolute number of holdings needed for adequate diversification. Critics have also argued that following MPT mechanically can push risk-averse investors into more risk than they can tolerate.

Behavior falls outside the model. MPT assumes rational investors. Markowitz himself said the chief mistake of small investors is buying after the market rises on the assumption it will keep rising, and selling after it falls on the assumption it will keep falling.

Modern Portfolio Theory vs. Related Concepts

MPT vs. “don’t put all your eggs in one basket.” Investors understood diversification long before 1952. MPT made it measurable: it shows how much a given asset lowers risk, based on its correlation with everything else, and which combinations are efficient.

MPT vs. post-modern portfolio theory (PMPT). PMPT accepts MPT’s core premise that a mix of assets can beat a single low-risk holding. It changes the risk measure, replacing variance with downside deviation (volatility below a target return) so that only losses count as risk. The Sortino ratio applies the same idea to performance measurement.

MPT vs. CAPM. MPT is a normative tool for building portfolios. CAPM, developed afterward, is a model of how assets are priced in equilibrium. It assumes every investor follows MPT and concludes that only market (systematic) risk, measured by beta, earns a premium.

MPT vs. the efficient market hypothesis. The two are often confused because both use “efficient.” The efficient frontier describes portfolios with the best risk-return trade-off. Market efficiency describes how fully prices reflect available information. Each idea stands without the other.

The Bottom Line

Modern portfolio theory matters to anyone setting a mix of stocks, bonds and other assets, from institutions to individuals using index funds or robo-advisors. Its practical lesson is to assess risk tolerance first, then combine assets with low correlations to cut volatility without giving up expected return. Treat the results as estimates, since correlations shift, variance counts gains as risk, and no mix removes market-wide risk.