Portfolio Theory Basics — Mean-Variance Magic (with a wink)

You already met alternative investments in the last chapter. Now we’re going to put them in a two-person relationship with stocks and bonds, see if sparks fly, and figure out how to stop your portfolio from swiping left on diversification.

Hook: Why this actually matters (besides passing CFA Level I)

Imagine your investments are roommates. One plays loud music (equities), another hums politely (bonds), and that weird plant in the corner (real estate, hedge funds, private equity) sometimes wilts, sometimes thrives. Portfolio theory is the roommate mediation protocol: how to combine noisy returns so the rent gets paid and you don’t cry into your monthly statements.

You’ve already learned about alternative investments and how their risk factors and diversification strategies can change portfolio behavior. Here we formalize the math and intuition: expected returns, variance, covariance, the efficient frontier, and why low correlation is your portfolio’s bestie.

Core ideas (quick map)

• Expected return = weighted average of asset returns.
• Risk = variance (or standard deviation) of portfolio returns; not additive.
• Covariance/correlation = how assets move together — the secret sauce for diversification.
• Efficient frontier = portfolios that give you the highest return for a given risk.
• Systematic vs unsystematic risk = the stuff you can’t diversify away (market) vs the stuff you can (firm-specific).

Two-asset intuition (the classroom demo you’ll love)

Let assets A and B have weights w_A and w_B (w_A + w_B = 1). Then:

Expected return: E(R_p) = w_A * E(R_A) + w_B * E(R_B)

Portfolio variance:
Var(R_p) = w_A^2 * σ_A^2 + w_B^2 * σ_B^2 + 2 * w_A * w_B * Cov(A,B)

Where Cov(A,B) = ρ_AB * σ_A * σ_B

Key takeaway: variance includes a covariance term. That term can shrink your risk if correlation ρ_AB < 1. That’s why holding two imperfectly correlated assets is better than holding one.

Quick numeric toy example:

• Equities: expected return 10%, σ = 18%
• Bonds: expected return 4%, σ = 6%
• Correlation = 0.2

If you split 50/50, the portfolio standard deviation is noticeably less than 0.518% + 0.56% — because of the diversification term. Try the math in your head like it’s a sudoku puzzle you actually want to solve.

Diversification: the practical art

Why do alternatives matter? From previous topics you know alternatives often have unique risk factors, and sometimes low correlation to stocks. That lowers Cov terms and pulls your portfolio variance down.

Ask yourself:

• Does this alternative move with equities when the market panics? (If yes, correlation high → less diversification benefit.)
• Does the alternative introduce new risks we can’t diversify away (liquidity, manager risk)? Those are real — you learned about them — and they alter the risk-return tradeoff.

Remember: diversification reduces unsystematic risk. It does not remove systematic risk — the market-wide storm.

Diversification is like a spice rack; too many similar spices doesn’t help. You want contrasting flavors.

The efficient frontier & minimum-variance portfolio (elevate your portfolio game)

Plot expected return (y-axis) vs risk (x-axis). All possible portfolios trace a cloud; the upper boundary of that cloud is the efficient frontier: the set of portfolios that maximize return for a given risk.

• The minimum-variance portfolio (MVP) is the left-most point on that frontier — the least risk possible.
• The tangency (or market) portfolio is the one with the highest Sharpe ratio when combined with a risk-free asset.

Why this matters: If you’re rational and risk-averse, you pick a point on the efficient frontier consistent with your risk tolerance. The math (mean-variance optimization) finds those portfolios. But watch out — real-world estimates (returns, covariances) are noisy; optimization can overfit. Which is why human judgment and sensible constraints matter — and where the CFA brain flexes.

Systematic vs Unsystematic — simple, but lethal if misunderstood

• Systematic risk (market risk): measured by beta in CAPM. Cannot be diversified away.
• Unsystematic risk: firm-specific stuff — can be diversified away by holding many uncorrelated assets.

Practical rule: Past a certain number of stocks (often ~20–30), adding more stocks yields diminishing reductions in portfolio variance. But add an alternative with low correlation, and you may reduce risk more efficiently than by adding more equities.

Common traps and counterintuitive facts

• Adding a high-volatility asset can still reduce portfolio volatility if correlation is low enough. High σ is not always bad.
• Perfectly uncorrelated assets are rare. Mis-estimating correlation kills diversification plans.
• Alternatives may look low-correlation historically but can spike toward 1 in crises (liquidity crunches, flight-to-quality). You read about these risk factors earlier — bring that knowledge here.

Practical checklist for constructing a rational portfolio (CFA-style action list)

1. Estimate expected returns, volatilities, and correlations (use historical data, but stress-test).
2. Identify which risks are systematic vs unsystematic.
3. Use mean-variance tools to find the efficient frontier, but add realistic constraints (no short-selling? weight limits?).
4. Consider risk-adjusted returns (Sharpe ratio) and investor utility (risk aversion).
5. Stress-test portfolio performance in crisis scenarios (what happens to alternatives when equities crash?).
6. Revisit often: correlations change.

Closing — the punchline you’ll remember

Portfolio theory is both mathematical and a bit of social engineering: it teaches you how to blend humans’ appetites for return with the cold math of risk. Alternatives are not magic pixie dust — they are spices that can transform your dish if you understand their flavor profile (correlation, liquidity, manager risk). The goal is simple: get the highest return for the risk you’re willing to stomach.

If diversification is the therapy, portfolio theory is the treatment plan.

Key takeaways:

• Diversification works because of covariance, not because you simply own many things.
• Alternatives can add real value through low correlation, but beware hidden risk factors and crisis behavior.
• Use mean-variance optimization wisely: it's a tool, not an oracle.

Go forth, optimize — and maybe bring snacks for your portfolio’s roommates.