Quant Lab

Quant analysis, on sliders.

The metrics every hedge-fund letter throws at you — Sharpe, Sortino, α, β, drawdown, VaR, correlation, information ratio, Kelly — each with the simulated returns behind them. Move a slider and watch the ratio, distribution and equity curve react in lockstep.

Risk-adjusted return

Sharpe Ratio

Excess return per unit of total volatility. Rewards you for beating the risk-free rate, punishes you for wobbling on the way.

Sharpe = (Rₐ − R_f) / σ
All annualised: multiply daily mean by 252, daily σ by √252.
Sharpe
1.52
Excess ret
8.0%
Vol
18.0%
252 periods
Good, hedge-fund territory.
Downside-adjusted

Sortino Ratio

Same idea as Sharpe, but only penalises the volatility that hurts — the downside. Push the skew slider negative and watch Sortino diverge.

Sortino
1.71
Sharpe
1.12
Downside σ
10.7%
Total σ
16.4%
μ
When Sortino ≫ Sharpe, upside is doing most of the volatility work — a good sign. When they're close, downside dominates.
CAPM

Alpha & Beta

Beta = sensitivity to the market. Alpha = the return you can't explain by that sensitivity — pure skill (or noise). Regress the asset on the market and read the slope and intercept.

Est. β
1.18
true 1.20
Est. α (ann.)
9.38%
true 3.0%
0.87
market explains…
market →asset ↑
The line's slope is β. Where it crosses market = 0 is α (per period). Crank idiosyncratic vol and R² tanks even though β is unchanged.
Path risk

Max Drawdown & Calmar

Two portfolios can have the same average return and radically different pain. Drawdown measures peak-to-trough loss. Calmar = CAGR / |MaxDD|.

Max DD
-34.8%
165 trading days
CAGR
11.3%
Calmar
0.32
1260 periods
Shaded band = the worst peak-to-trough episode. A −50% drawdown needs a +100% recovery just to break even.
Tail risk

Value at Risk (VaR) & CVaR

VaR: on the worst 5% of days, you'll lose at least this much. CVaR (a.k.a. Expected Shortfall) is the average loss on those bad days — VaR's more honest cousin.

1-day VaR (95%)
£1,967
-1.97% move
CVaR (avg loss beyond)
£2,384
-2.38%
μVaR
Red bars = returns worse than VaR. VaR tells you the threshold; CVaR tells you how bad the average tail day actually is.
Diversification

Correlation & Portfolio Vol

Combining assets with ρ < 1 cuts risk more than return. This is the only free lunch in finance — see it appear as you drag ρ from +1 down to −1.

Port μ
11.0%
Port σ
18.2%
Diversif.
4.30%
vol saved vs naïve avg
Asset AAsset BPortfolio
Active management

Information Ratio

Excess return over a benchmark, divided by the volatility of that excess (tracking error). Rewards steady outperformance, punishes lucky bets.

IR = active return / tracking error
Grinold's law: IR = IC × √breadth. More independent bets = higher IR at the same skill level.
IR
0.40
Verdict
Weak
Marginal — often not worth the fees.
Position sizing

Kelly Criterion

How much of your bankroll should you bet? Kelly maximises long-run log growth. Bet more and you go bust; bet less and you grow slower.

Full Kelly
10.0%
You bet
10.0%
Edge
10.0%
Bankroll after 500 bets. Push fraction above 1× and watch volatility explode; below 0 (negative edge) you can't win no matter the sizing.