Risk, Return, and Diversification

The Risk-Return Tradeoff

Markets don't hand out free returns. If something offers a higher expected payoff, it comes bundled with more risk — that trade is basically the whole discipline of finance in one sentence. Three-month U.S. Treasury bills pay around 5% and are about as close to risk-free as money gets. The S&P 500 has averaged closer to 10% historically, but "historically" is doing a lot of work in that sentence — it includes 2008 (−37%) and 2022 (−18%). Put money into early-stage startup equity and you might see 50%+ returns, or you might see zero. All of it, forever.

The standard way to measure risk is volatility — the standard deviation of returns. A stock with 25% annual volatility will typically wander within about ±25% of its expected return in roughly two years out of three. That's not a guarantee, just what "one standard deviation" means in practice.

Measuring Volatility

Given a series of daily returns r₁, r₂, …, rₙ, here's the actual recipe:

Mean return: μ = (1/n) Σ rᵢ
Variance: σ² = (1/(n−1)) Σ (rᵢ − μ)²
Daily volatility: σ_daily = √(variance)
Annualized volatility: σ_annual = σ_daily × √252

That 252 is the number of trading days in a typical year, and the square-root scaling is a convention you'll see used everywhere in finance, not something specific to this lesson.

Diversification

Here's one of the only genuinely free lunches finance has to offer: spreading money across multiple assets lowers your risk without lowering your expected return. It works because each asset's random ups and downs partially cancel each other out, as long as they're not moving in perfect lockstep.

The correlation between two assets, ρ, runs from −1 (perfectly opposite) to +1 (perfectly identical). For a two-asset portfolio:

σ²_portfolio = w₁²σ₁² + w₂²σ₂² + 2·w₁·w₂·σ₁·σ₂·ρ

Any time ρ is less than 1, the combined portfolio ends up less volatile than a simple weighted average of its two pieces would suggest. Push ρ all the way to −1 and you could, in theory, build a portfolio with zero variance — a perfect hedge. Real stocks don't cooperate that well; correlations usually sit somewhere between 0.3 and 0.7. Still positive, but far from 1, which is exactly why diversifying always helps at least a little.

Harry Markowitz turned this into a Nobel Prize in 1952 by formalizing it as Modern Portfolio Theory. His core claim: risk you can diversify away shouldn't earn you anything extra for holding it. Only the risk you're stuck with no matter how many stocks you own — systematic, market-wide risk — actually deserves to be compensated.

The Sharpe Ratio

Raw return numbers alone are a trap. A fund that returned 20% last year by taking on huge risk isn't obviously better than one that returned a steadier 12%. The Sharpe ratio fixes that by dividing out the risk:

Sharpe = (Return_portfolio − Return_risk_free) / σ_portfolio

A Sharpe of 1.0 is generally considered solid — one unit of extra return per unit of risk taken. Above 2.0 is rare and excellent. Below 0.5 is weak. Warren Buffett's Berkshire Hathaway has run around a 0.7 Sharpe ratio across decades — which sounds unremarkable until you remember it's been sustained at enormous scale for longer than most investors have been alive.

Systematic vs. Idiosyncratic Risk

All of a stock's risk splits into two buckets:

• Systematic risk — recessions, rate hikes, pandemics. Economy-wide forces that touch every stock at once. You can't diversify your way out of this one; it's measured by beta, which the Quant track covers.
• Idiosyncratic risk — a bad product launch, a CEO scandal, one company's specific accounting fraud. This is the free-lunch part: own enough different companies and this risk mostly cancels itself out.