Reading the scoreboard

A strategy spits out an equity curve. These are the numbers that turn that squiggle into a verdict — how fast it grows, how rough the ride is, and whether the wins really outweigh the losses.

CAGR: how fast did it grow?

CAGR (compound annual growth rate) is the single yearly growth rate that would take you from your starting balance to your ending balance, compounding all the way. Turn $10,000 into $16,100 over four years and that's a CAGR of about 12.6% a year.

It's more honest than "total return" because it bakes in time: +60% sounds great until you learn it took ten years. CAGR is the great equaliser for comparing strategies of different lengths — but on its own it says nothing about how scary the journey was. For that, we need risk.

The Sharpe ratio: return per unit of risk

This is the headline quality number in all of quant. The Sharpe ratio divides your return by your risk — specifically by the volatility (σ) you met in Module 1:

Sharpe ≈ average return ÷ volatility of return  (annualised)

Why divide? Because return alone is gameable — just take more risk. Sharpe asks the better question: how much return did you earn for each unit of stomach-churning you endured? Two strategies can finish at the same place with completely different Sharpes:

Rough rule of thumb for a backtest: under 1 is weak, around 1 is decent, 2 is strong, and anything above 3 should make you suspicious rather than excited — it usually means the result is overfit (Module 5) rather than brilliant. There's a whole cottage industry of strategies that post a beautiful Sharpe in a backtest and a miserable one live.

Maximum drawdown: the pain metric

Sharpe describes the average ride. Maximum drawdown describes the worst moment — the largest drop from a peak to the following trough before a new peak. It's the number that decides whether you can actually live with a strategy.

A backtest can show a lovely CAGR and still have a −60% drawdown buried inside it. Almost nobody keeps following a system through a loss that deep — they quit at the bottom, locking it in. A strategy you can't hold isn't a strategy you own. Drawdown is where backtests meet human psychology, and it's why position sizing (Module 6) matters as much as the signal.

Win rate, payoff & expectancy

Beginners fixate on win rate — the share of trades that make money — and assume more is better. It isn't, on its own. What matters is win rate together with payoff: how big the average win is versus the average loss.

Tie them together and you get expectancy — the average profit you expect per trade:

expectancy = (win% × average win) − (loss% × average loss)

This is why a strategy can win only 35% of the time and still be excellent, if its winners are far bigger than its losers. Trend-following and momentum live here: many small losses, a few huge wins. As long as expectancy is positive, the math works over many trades.

The 35% strategy loses nearly two times out of three and is the better business — because the payoff does the heavy lifting. Traders often express each result as an R-multiple: profit measured in units of the risk taken. Risk $100, make $300, that's +3R. We watched exactly this fat-tailed, low-win-rate shape play out in a real momentum teardown.

Where to go next

You can now judge a strategy's scoreboard. But every number above can be faked by a careless or dishonest backtest — and learning to catch that is the whole point of this site. Modules 3–7 (coming soon) cover the data traps, your first backtest, and the validation gauntlet. In the meantime, the tools below let you pressure-test these very metrics today.

Put these metrics to work — free, in your browser