The Darvas box

When a stock goes on a long, historic run, it almost never moves in a straight line. It surges, pauses, dips, and surges again. Most traders make one of two mistakes: they sell too early on the first scary pause, or they cling on far too long after the trend has quietly died and give back everything.

In the 1950s a professional dancer named Nicolas Darvas - touring the world, reading stock prices by telegram, with no screen and no news feed - found a way through. He reduced the whole messy market to a stack of simple rectangles. How can drawing a chart as a set of boxes strip almost all the emotion out of riding a trend?

The Darvas box - a rectangle drawn around a stock's sideways pause, with the recent high as the roof and the recent low as the floor - exists as a mechanical framework to enforce discipline and trail risk.

Darvas accepted that predicting the market was impossible. So instead of predicting, he observed. He noticed that strong stocks move in steps: a surge to a new high, a stall at resistance, a pullback, and then a few weeks of chopping sideways. He drew a box around that sideways chop. The roof was the resistance; the floor was the support.

His rules were brutally simple. While the stock stayed inside the box, he did nothing. If it broke the roof, he bought - momentum had resumed. If it broke the floor, he sold - the move was over. By defining the box in advance, he removed the news, the tips, the stories, and his own feelings from the decision entirely. All that was left was the price and two lines.

The context is worth dwelling on, because it explains why the method is so unemotional. Darvas was often on the other side of the world, weeks from any market gossip, receiving only a short cable of prices. He could not react to a rumour because he never heard it. That forced isolation turned out to be an advantage: with nothing to read but the numbers, he was left with pure supply and demand, and he built a system that a person sitting in front of a screen of scrolling news would find almost impossible to follow - not because it is complicated, but because it asks you to ignore everything except two prices. The method only works because trends, once genuinely underway, tend to keep going in steps - each higher box is the trend pausing to gather itself before the next leg.

The system has four moving parts: find momentum, draw the box, buy the breakout, and trail the stop by stacking boxes.

1. Require an uptrend. The stock must already be in a strong advance - what earlier modules called a Stage 2 uptrend, the phase where a stock is steadily making higher highs. Darvas only traded stocks pushing to new highs.
2. Draw the roof. The stock hits a new high, then fails to exceed it for three straight days. That peak becomes the roof.
3. Draw the floor. It pulls back to a low, then fails to break below it for three straight days. That low becomes the floor.
4. Buy the breakout, then stack. You buy when price clears the roof on heavy volume. As the stock rises it forms a new, higher box; you raise your stop-loss - the pre-set price at which you will automatically sell to cap a loss - up to the floor of that new box. Repeat, stacking boxes higher and higher, until the stock finally breaks the floor of the highest box and you are out.

The genius is the simplicity. It forces you to stay in winners by ignoring the noise inside the box, and it forces you out of losers the moment a floor breaks.

Every price and chart in this module is a made-up, illustrative example, never a real quote. [illustrative]

Walk one composite case. illustrative

A trader buys the breakout of Box A at ₹100. The stock runs beautifully to ₹150 and forms a higher box (Box B) between ₹135 and ₹150. The mechanical move is obvious: raise the stop to ₹135, locking in most of the gain.

But the trader has fallen in love. They have decided this is a "forever hold." When the market turns hostile and the stock breaks the floor of Box B at ₹135, they do not sell - they tell themselves it is a temporary dip. It slides to ₹120. Panic flickers, but they hold, now hoping it will bounce off the roof of the old Box A at ₹100. It slices straight through ₹100 and keeps going to ₹70.

By ignoring the one rule - exit when the floor of the current box breaks - the trader turned a large, locked-in gain into a painful loss.

The disciplined version sold mechanically at ₹135 the moment the floor gave way, kept the gain, and moved on. This is the entire method in one line: let the stacked boxes carry the profit up, and step off the instant the current floor breaks - the trailing stop is what turns a good trend into kept money.

A Darvas box cannot stop a breakout from being a fake. In choppy, volatile markets a stock will frequently pop above the roof, trigger the buyers, and then reverse straight back down through the floor. Every box breakout is a probability, not a promise; some meaningful share of them fail immediately, and the method survives only because the losses are kept small.

So the floor of the current box is your unbreakable line. If price loses it, you cannot invent excuses - you cannot decide to "wait for the old box to hold." The current structure has failed; you execute the stop, mechanically. That is not a flaw in the system; it is the system.

It also cannot tell you why the stock is moving, or whether it deserves to. The boxes are pure price and volume. A stock can stack perfect boxes on a story that later collapses. And the method is not magic: it will hand you a string of small losing breakouts in a sideways, whippy market, and no amount of neat boxes changes that. There is no setup that wins without losses.

Almost every Darvas failure is a failure of discipline, not of the boxes.

1. Buying inside the box. Bored of waiting, the trader buys in the middle, guessing the breakout. Often the stock breaks down instead.

2. Ignoring the floor. The floor breaks and the trader holds anyway, "sure" it is a shakeout. Sometimes it is; often it is the start of a real decline, and the small planned loss becomes a large one.

3. Applying it in a downtrend. Drawing boxes on a falling stock, where there is no underlying demand for the method to ride.

4. Trading the chop. Buying near the floor and selling near the roof inside a wide box, getting sliced to pieces by the very noise the method tells you to ignore.