Darvas box breakout
BreakoutIntermediate
Nicolas Darvas boxed a stock between a recent high and a recent low after a surge in volume, then bought a breakout through the top of the box. The broken top became the stop as new boxes formed higher. It is a discretionary-looking trend method with a mechanical core: trade only rising boxes, and trail the stop to the last box top.
Not investment advice. Nothing here is a recommendation to buy or sell anything. Backtests and historical studies are not guarantees, and a rule that worked in one sample can fail in the next. Trading can lose money. Links are for education, not endorsements. Do your own research.
- Rules
- Define a box after a volume expansion: the top is a high that price fails to exceed for several sessions, the bottom is the corresponding low. Buy a close through the top. Place the stop at the top of the prior box, or at the current box bottom if no prior box exists. Add only when a new box breaks, and exit when the stop is hit. Ignore boxes that are not stepping up.
- Where AI fits
- AI can turn the narrative into a volume-threshold and a 'failed to exceed' window so two people code the same box. It should freeze that window before testing. It can also reject names whose volume spike is a one-print error. A language model that draws boxes on a screenshot is not the strategy.
- Markets
- Liquid stocks, which is the market Darvas traded. The logic can be tested on futures, but the volume rule has to be rewritten.
- Timeframes
- Daily.
- Risks
- The box is easy to draw after a winner and hard to define in real time. Volume spikes from index adds and news create false boxes. Trailing at the prior top gives back a large open profit. The original story is not a performance record you can rely on.