The Darvas Box — Definition and Detection

In 1956, a Hungarian-born professional ballroom dancer named Nicolas Darvas was touring the world performing in 22 countries. Between performances, in hotel rooms and dressing rooms, he traded the U.S. stock market by telegram, receiving end-of-day price information from his broker and sending buy and sell orders in return. He had no economic training, no Wall Street pedigree, and no real-time market data. He had only a method he had developed by trial, error, and a great deal of expensive failure: the Darvas Box.

In the 18 months between June 1957 and December 1958, Darvas turned $36,000 of his own trading capital into $2.25 million. The story became the bestseller How I Made $2,000,000 in the Stock Market, and the Darvas Box method became one of the foundational pattern-trading systems of the 20th century. Sixty-five years later, the rules still work — because they are rooted in a market truth that has not changed: price action and volume tell the story, and structured discipline beats opinion.

What a Box Is

A Darvas Box is a defined rectangle drawn around a stock's price action, bounded by:

The Top: a recent high that is not exceeded for three consecutive trading days afterward.
The Bottom: a recent low (forming after the top is established) that is not undercut for three consecutive trading days afterward.

That's it. No moving averages, no oscillators, no fundamental analysis. The box is purely a geometric description of the range within which a stock has settled after an advance. The market itself draws the box; your job is only to recognize it and react when it resolves.

The Three-Day Confirmation Rule

The critical detail is the three-day confirmation. A new high is not a box top until three trading days pass without a higher high being printed. A new low is not a box bottom until three trading days pass without a lower low.

Why three days? Darvas found through trial and error that:

One day is noise. Any single bar can produce a spurious high or low that doesn't represent real supply/demand.
Two days is suggestive but unreliable. Many false signals.
Three days is statistically significant. When a high stands for three full trading sessions without being exceeded, the market is telling you that price has been rejected at that level — supply lives there. The same logic applies to lows.

Three is the minimum confirmation; sometimes a high stands for 5, 10, or 30 days before it's exceeded. The longer the boundary holds, the more meaningful it is.

How to Detect a Box in Practice

Step-by-step:

Find a stock that has recently advanced — preferably a stock making 52-week or all-time highs. Darvas insisted on this; boxes in downtrending stocks are not buyable.
Identify the most recent high. Mark it.
Wait three full trading days. If no higher high is printed in those three days, the most recent high becomes the Box Top.
Watch for the next low after the top. When the stock pulls back and prints a low that holds for three trading days without being undercut, that low becomes the Box Bottom.
Draw the box. You now have a rectangle bounded by the Top and Bottom. The stock is "in the box" — consolidating within this range.
Wait for the breakout. When the stock breaks above the Box Top on heavy volume, that's the Darvas buy signal. Stop loss goes just below the Box Top (now acting as new support).

The wait between drawing the box and the breakout can be days, weeks, or longer. Patience is the entire game.

Box Dimensions

Darvas didn't specify rigid dimensional rules, but his trades clustered around boxes with these properties:

Box height: typically 10-25% from top to bottom. Boxes narrower than 5% are usually noise; wider than 30% suggests the stock is too volatile to be in a clean consolidation.
Duration: most successful boxes resolved within 2-8 weeks of formation. Longer boxes can work but require more patience.
Position: the box should form at or near 52-week or all-time highs. Boxes in the middle of trading ranges are weaker; boxes near downtrend lows are not Darvas setups at all.

Stacked Boxes — The Pyramid

One of Darvas's most powerful observations: strong stocks form a series of progressively higher boxes. After a breakout from Box 1, the stock advances 10-25%, then enters a new consolidation (Box 2) at a higher level. Box 2 breaks out, and the stock advances to Box 3. And so on.

Each box is a discrete trade. You buy the breakout, set a stop below the box top, and ride the advance to the next box. When the new box forms, your stop trails up to the new box top — locking in the gain from the prior leg.

This stacked-box approach is how Darvas made his largest gains. A single stock might form 4-6 boxes over a 12-month run, each box producing a 15-25% gain. Stacked together, that's a 100-300% return on a single name.

The Volume Confirmation

Darvas was emphatic: breakouts must be confirmed by heavy volume. He had no real-time data, so he relied on end-of-day volume reports from his broker. His rule of thumb: breakout day volume should be visibly larger than any volume bar in the prior 10-15 trading sessions, ideally 2-3x the recent average.

Low-volume breakouts were the most common failure mode Darvas encountered, and he developed a strict discipline of not acting on them. A price breakout without volume confirmation is a fake-out by definition — institutions aren't participating, and the move will reverse.

What the Box Is Not

The Darvas Box is not:

A support/resistance zone you draw subjectively. The box is mechanical — three-day confirmation defines the boundaries.
A pattern you can use in downtrends. Darvas boxes form only in stocks making new highs.
A timing tool for entering inside the box. You buy only on the breakout out of the box, never inside it.
A "long-term hold" strategy. Each box is a discrete trade; the stop discipline determines when you exit.

Why It Works

The Darvas Box works because it is a structural representation of supply and demand reaching balance. The box top represents the price at which supply has been heaviest; the box bottom represents the price at which demand has held. When the stock breaks out of the box on volume, the equilibrium has been broken in favor of demand — supply at the old top has been absorbed, and the stock is free to seek the next level.

This is the same logic that powers all breakout systems. What makes Darvas unique is the mechanical simplicity of the box construction: three-day confirmation, no judgment calls, no moving averages to debate. The box either forms or it doesn't; the breakout either confirms or it doesn't. The system removes you from the decision in real time, which is exactly the kind of discipline that compounds.

← Back to Nicolas Darvas