Tempo submerso: a métrica de drawdown que quase ninguém mede
Quick answer: drawdown has two dimensions and you only measure one. Depth tells you how much you lost; time underwater tells you how many days or months you spent below your previous peak. Almost nobody quits over a 12% fall. People quit after eleven months without a new high, even when the fall was small.
Pontos principais
- Two systems with identical drawdowns can be completely different to live with.
- What wears you down is not the loss; it is the prolonged absence of new highs.
- It is measured as two numbers: the longest stretch and the total percentage of time.
- A very high percentage usually points at costs or frequency, not the strategy.
Two curves with the same drawdown
| System A | System B | |
|---|---|---|
| Result for the year | +18% | +18% |
| Largest drawdown | 12% | 12% |
| Longest stretch underwater | 6 weeks | 9 months |
| Months setting a new high | 8 of 12 | 2 of 12 |
Same return, same depth, opposite experiences. System B is the one people abandon in month seven, right before it starts working again.
How to measure it
- Sort your history by date and compute cumulative equity after each trade.
- Track the running peak: at each point, the highest value reached so far.
- Mark each day as “underwater” if equity sits below that peak.
- Record the longest consecutive stretch and the total percentage of days underwater.
How to read the percentage
| Time underwater | Reading | What to check |
|---|---|---|
| Under 30% | Very healthy, frequent new highs | Nothing |
| 30–50% | Normal for most systems | Nothing |
| 50–70% | Slow grind, draining to trade | Costs and frequency |
| Over 70% | The edge is marginal or gone | Expectancy over 100 trades |
Indicative ranges: they depend on style and frequency. A low-frequency swing system naturally spends more time underwater than an intraday one.
Why this metric predicts quitting
Depth hurts once; time hurts every day. A trader can accept losing 12% in two weeks if the third week brings a new high. The same 12% spread across nine months of stagnation produces exactly the behaviours that ruin a system that was working: raising size to “speed things up”, switching strategy, adding new instruments, breaking rules out of boredom.
Which is why it is worth measuring before you need it. If you know your system has historically gone up to four months without a new high, the fourth month stops being an alarm and becomes an expected data point.
How to reduce it
- Cut costs. They are what turns slow progress into stagnation.
- Reduce frequency if your trades cancel each other out.
- Add an uncorrelated second source: more chances to set a new high.
- Revisit the target: overly ambitious targets lengthen the flat periods.
- Do not raise size to get out faster. That trades duration for depth.
What you need to measure it
A log with a date and a result per trade; with that, a spreadsheet is enough. The exact format is in the free ebook The Trading Journal Blueprint. If your diagnosis points at costs, there is The True Cost of Trading; if it points at adding a second source, the Cortex Signal Kit and Cortex Automation operate at different moments and therefore tend not to stall at the same time.
What is time underwater?
The period your account spends below its previous peak, until it sets a new one.
How much is too much?
It depends on style, but above 70% of the time it is worth reviewing expectancy across at least 100 trades.
Is it worse than a deep drawdown?
It is different. Depth threatens your capital; duration threatens your participation, and without participation the capital does not help either.
How often should I calculate it?
Once a month is enough. It is a trend metric, not a daily one.
Does an automated system reduce it?
Not by itself, but it stops you making it worse: most long stretches get longer because of improvised changes midway through.
Measure the duration, not just the fall. See the Cortex Signal Kit and Cortex Automation.
Disclaimer: educational content. Examples and ranges are illustrative. Cortex Next is not a financial service and guarantees no results. Trading involves risk of loss.
