In 1952, Harry Markowitz made an argument that sounds obvious now but was not obvious then: what matters in a portfolio is not how each asset performs on its own but how they perform together. That shift in thinking, from individual securities to the portfolio as a whole, is what Modern Portfolio Theory is built on.
The core idea
MPT uses statistics to find combinations of assets that reduce overall volatility without proportionally reducing returns. The mechanism is correlation. Two assets that tend to move independently of each other, when held together, produce a smoother ride than either one alone.
The mathematics
Expected return: E = Σ(wᵢ × Eᵢ)
Portfolio variance: σ² = ΣΣ(wᵢ × wⱼ × σᵢ × σⱼ × ρᵢⱼ)
Where wᵢ is each asset's weight, Eᵢ its expected return, σᵢ its standard deviation, and ρᵢⱼ the correlation between any two assets. That last term is the one that does the work.
A quick example
60% in Stock A (12% expected return, 10% standard deviation) and 40% in Stock B (8% expected return, 6% standard deviation), with a correlation of 0.3.
Expected return: (0.6 × 12) + (0.4 × 8) = 10.4%
Portfolio variance: (0.6² × 10²) + (0.4² × 6²) + (2 × 0.6 × 0.4 × 10 × 6 × 0.3) = 46.56
Portfolio standard deviation: √46.56 = 6.82%
Stock A alone carries 10% volatility. The combined portfolio carries 6.82%. Nothing changed except combining two assets that do not move in lockstep.
The efficient frontier
Plot every possible combination of a set of assets on a risk-return chart. The upper edge of that cloud of portfolios is the efficient frontier. Every point on it represents a portfolio where you cannot get more return without taking more risk, or less risk without sacrificing return.
| Portfolio type | Expected return (%) | Standard deviation (%) |
| Conservative | 8 | 7 |
| Balanced | 10 | 9 |
| Growth | 12 | 11 |
| Aggressive | 14 | 14 |
Portfolios that fall below the frontier are inefficient. Same risk, lower return. No reason to be there.
The Capital Market Line
Add a risk-free asset, a Government of India bond for instance, and the efficient frontier becomes a straight line called the Capital Market Line. Its slope is the Sharpe Ratio.
E = Rf + [(Em - Rf) / σm] × σ
Indian practitioners typically use Treasury Bill yields as the risk-free rate proxy.
Key concepts worth knowing
- Diversification reduces unsystematic risk, the risk specific to individual companies or sectors
- The risk-return trade-off means higher expected returns come with higher variance; no way around it
- Correlation is the input that most determines whether diversification actually works
- Optimisation finds the weighting that sits on the frontier for a given risk tolerance
Where the theory runs into reality
MPT assumes returns follow a normal distribution. They do not, not reliably. Market returns have fat tails, meaning extreme events happen more often than the model expects.
It also assumes correlations are stable. They are not. During the 2008 financial crisis, assets that had historically moved independently started falling together. Indian markets showed similar behaviour during the 2020 pandemic selloff. Diversification tends to fail when it is most needed.
The model treats upside and downside volatility as equivalent. A 5% daily gain and a 5% daily loss register identically in the variance calculation. Most investors do not experience them identically.
And the inputs, expected returns and correlations, are estimated from historical data. Past correlations do not always predict future ones.
How it is used today
Despite those gaps, MPT remains the foundation most institutional asset allocation is built on. Mutual fund houses use it to balance equity, debt, and money market exposure. The NPS applies similar logic to manage long-term capital growth alongside income needs. Robo-advisory platforms in India have built MPT-based allocation engines into their core product.
Post-Modern Portfolio Theory addresses the upside-downside asymmetry problem by focusing on downside risk specifically. The Black-Litterman model allows investor views to be blended with market equilibrium rather than relying solely on historical data. These are refinements, not replacements.
Conclusion
The diversification argument at the heart of MPT holds up. Combining assets that do not move together produces better risk-adjusted outcomes than picking individual securities, however good they are. Where the theory asks for caution is in its assumptions. Normal distributions, stable correlations, rational markets. These simplify the mathematics but do not always describe how markets actually behave.
For Indian investors, the framework is worth understanding. The outputs are worth questioning.
