WealthJot.ai

Modern Portfolio Theory, Plainly

advanced8 min read

The big idea: diversification lets you earn more return per unit of risk. What that really means.

Modern Portfolio Theory (MPT), the Nobel-winning framework from Harry Markowitz, gave investing its mathematical foundation. Stripped of the equations, its central insight is profound and practical: a portfolio’s risk is not just the average of its parts — it depends on how the parts move together.

MPT’s revolutionary idea is that **risk should be measured at the portfolio level, not the holding level — and combining imperfectly-correlatedHow closely two assets move together. assets can lower total risk without lowering return.* Two stocks each volatile on their own can, together*, produce a smoother portfolio if they don’t move in lockstep — because when one zigs, the other zags, and the swings partly cancel. This is the mathematical backbone of the “diversificationSpreading money across assets that don’t move together to cut risk. is the only free lunch” idea from the investing track: *correlationHow closely two assets move together., not just the individual risks, determines portfolio risk. The payoff is that for any target return there’s a mix that minimises risk, and for any risk level a mix that maximises return — you can engineer a better risk-return trade-off than any single asset offers. MPT shifted investing from “pick good assets” to “engineer a good combination*,” and everything in this module (the efficient frontier, risk parityAllocating so each asset contributes equal risk.) flows from it.
ExampleHold only Stock A (volatile) and you bear all its swings. Add Stock B, equally volatile but moving differently — on days A drops, B often rises. The combined portfolio’s ups and downs partly cancel, so it’s less volatile than either alone, with similar expected return. The magic wasn’t the stocks; it was their low correlationHow closely two assets move together., exactly as MPT predicts.
Key takeawayModern Portfolio Theory’s core idea: portfolio risk depends on how holdings co-move (correlationHow closely two assets move together.), not just their individual risks — so combining imperfectly-correlatedHow closely two assets move together. assets lowers risk without lowering return. It’s the math behind diversificationSpreading money across assets that don’t move together to cut risk. and shifts investing from picking assets to engineering combinations.
FAQs
Is MPT still useful given its known flaws?

Its *insight* (correlation-driven diversification) is foundational and timeless; its *precise optimisation* is fragile in practice because it depends on hard-to-estimate inputs (covered in the pitfalls lesson). So use MPT’s intuition — diversify across imperfectly-correlated assets — while being deeply cautious about trusting an optimiser’s exact “optimal” weights.