Every strategy is pitched with one number — total return. It is the biggest, shiniest figure on the page, and it is the most misleading, because it never tells you how long it took, what you endured along the way, or whether it can be repeated. Six metrics break "profit" into its parts. All six come from the same backtest we took apart earlier in this series — Adaptive Trend on Bitcoin, 2016→2026.
Here is the curve every one of these numbers is really measuring:
1. Total return
Formula: final value ÷ initial value − 1. Means: how much capital grew from the first day of the test to the last — the number that headlines every strategy pitch. Hides: time and path together. 2,926% over ten years is not 2,926% over two, and the figure never mentions the drawdown that nearly forced you out before you reached it. The number: 2,926% for the strategy against 14,638% for simply buying and holding.
2. Compound annual growth (CAGR)
Formula: (final ÷ initial) ^ (1 ÷ years) − 1. Means: the constant yearly rate that, repeated unchanged, would land you on the same final result — it standardizes time so two strategies of different lengths can be compared. Hides: it flattens the path completely. A +150% year and a −40% year become a smooth rising line that never happened in reality, and the gap between that smooth line and the real one is what decides whether you survive. The number: 38.10% per year over ten years.
3. Annualized volatility
Formula: standard deviation of daily returns × √252. Means: how widely the portfolio swings around its average, expressed on a comparable annual scale. Hides: it treats upside and downside identically — but only one of them wakes you at 3am and pushes you to sell at the worst possible moment. That blind spot is exactly why the risk-adjusted metrics in the next part were invented. The number: held near a 25% target, resized as volatility changes.
4. Maximum drawdown
Formula: the minimum of (current value ÷ prior peak) − 1. Means: the deepest peak-to-trough fall the portfolio lived through, measured from its highest point. Hides: almost nothing — which is exactly why it matters. It is the only metric that measures what you will actually feel rather than what you will earn, and most people who quit, quit at the bottom of this number. The number: 28.99% — $100,000 became $71,010 at the worst point.
5. Time under water
Formula: the longest continuous stretch below the prior peak. Means: how long you wait to recover what you lost and climb back to where the fall began. Hides: it is missing from most performance reports even though it is the leading cause of quitting — people do not abandon a strategy when they lose, they abandon it when the wait to get back to even runs too long. The number: a −30% drawdown needs a +43% gain just to return to even.
6. Market exposure
Formula: days with an open position ÷ total test days. Means: how much of the time your capital was actually in the market and at risk, rather than sitting in cash. Hides: an equal return at 84% exposure beats the same return at 100% — the same result for less time under risk, and more room left to absorb a sudden shock when one arrives. The number: 84% in the market, 16% fully in cash.
The six at a glance
| Metric | What it means | What it hides | The number |
|---|---|---|---|
| Total return | Capital growth, start to finish | Time, path, and the drawdown survived | 2,926% vs 14,638% buy-and-hold |
| CAGR | The constant yearly rate | It flattens a jagged path into a smooth line | 38.10% per year |
| Annualized volatility | How widely it swings | Treats upside and downside the same | 25% target |
| Maximum drawdown | Deepest peak-to-trough fall | Almost nothing — it's the one you feel | 28.99% ($100k → $71,010) |
| Time under water | How long recovery takes | Missing from most reports | −30% needs +43% to recover |
| Market exposure | Time actually at risk | 84% for the same return beats 100% | 84% in market · 16% cash |
The recovery trap
The reason maximum drawdown and time under water matter more than the headline return is a piece of arithmetic most people never run: losses and the gains needed to undo them are not symmetric. The deeper the hole, the disproportionately larger the climb out.
| Drawdown | Gain needed to get back to even |
|---|---|
| −10% | +11% |
| −20% | +25% |
| −30% | +43% |
| −50% | +100% |
| −90% | +900% |
A strategy that never digs a deep hole is not just more comfortable — it compounds from a higher floor. Two systems with identical total return are not equal if one of them spent three years underwater to get there.
In exchange for what?
Total return answers how much? These six answer in exchange for what? — how long, how deep, how often you were exposed. The headline number is where the marketing lives; the other five are where the truth is. A quantitative investor never compares two strategies on the first question alone.
The next part goes one level deeper still: the risk-adjusted ratios that ask not just what happened, but whether it was repeatable skill or luck in a single sample.