Pyramid Trading

Pyramiding adds size to a winner. Each add is smaller than the last in the classic version, and each add has its own stop, so the combined position can still be exited for a gain or a small loss if the trend breaks. This is the opposite of averaging down. You are paying a worse price for additional size because the market has confirmed the direction. The trade-off is real. The average entry gets worse, and a sharp reversal can take back the open profit of the whole pyramid. In exchange, you have more size on during the part of the trend that is actually trending, and you were smallest when the idea was only a guess.

Write the adds before the first entry. For example, add only after a new higher low, add half the prior size, and raise the stop on the whole position to the latest higher low. If that stop would lock a loss on the combined size, the add is too big or too early. Do not add because the bar is green. Add because the structure advanced and the stop can move. Cap the number of adds. A pyramid of seven pieces is a position you cannot mentally track, and the last add is usually the one closest to the end of the trend. Two or three planned adds will capture the idea without turning the exit into a puzzle.

Use pyramids on trend trades with open-ended targets, where a single small entry leaves too much of the move uncaptured. Do not pyramid a range fade. Adding to a fade as it goes your way means you are largest at the mean, where the edge is gone. After each add, recompute the cash loss if the trail is hit. The number that matters is the worst case of the combined position, not the fact that each piece once looked small. If the worst case exceeds the risk budget, you cannot add, even if the chart is perfect.

Adding equal or larger size at each step, which makes the last, worst price the dominant one. Also adding without moving the stop, so a return to the original entry wipes every add and the original risk. And calling an average-down a pyramid because you added. The direction of the add relative to the entry is the definition. Adds below a long entry are averaging. They are a different, riskier decision.

Pyramids concentrate you in the names that have already moved, which is what you want in a trend and what hurts when correlations jump and every winner reverses together. They also complicate the journal. Record the position as one trade with adds, and record the planned risk after each add, or you will remember the pyramid as a brilliant add and forget the one that gave the whole swing back. Gaps through the trail exit the entire size at once. The last add made that gap more expensive. Size as if you know that.