# Pyramiding

> Pyramiding is a position-management rule that adds to a trade only after it has moved in its favour, with each addition smaller than the last, so the average entry stays close to the original price and the new size is funded by open profit. Adding 4, 3, 2 and 1 units at 100, 104, 108 and 112 gives an average entry of 104; the same units added in reverse give 108, and a pullback of under 4% erases the whole gain. Pyramiding is the anti-martingale: size rises after wins and never after losses, and every addition is re-sized against the current stop so total open risk stays inside the per-trade budget.

Canonical HTML page: https://qanterion.com/glossary/pyramiding
Updated: 2026-09-02
Also known as: scaling into a position, adding to winners, anti-martingale sizing

## How it is calculated

```
Upright pyramid: units_k = units_0 × r^k with r < 1 (each add smaller than the last)
average entry = Σ(units_k × price_k) ÷ Σ units_k
open risk after each add = Σ units × (average entry − stop) ≤ per-trade risk budget

Worked example — 4 / 3 / 2 / 1 units at 100 / 104 / 108 / 112:
  average entry = (400 + 312 + 216 + 112) ÷ 10 = 104
  stop trailed to 106 → 10 × (104 − 106) = −20, i.e. 20 of profit locked in
Inverted 1 / 2 / 3 / 4 at the same prices: average entry = 108, same stop → 10 × (108 − 106) = 20 at risk
```

## Worked example

Four units at 100, three at 104, two at 108 and one at 112 average 104; with the stop trailed to 106 the position cannot lose, and a fall to 106 gives back 60 of the 80 open profit shown at 112.

## The shape of the pyramid decides the average entry

An upright pyramid — 4, 3, 2, 1 units at 100, 104, 108, 112 — carries ten units at an average of 104, so a trailing stop at 106 locks in 20 of profit even after the last add. Inverting the same sequence puts four units at 112 and one at 100: the average becomes 108, and the same stop turns into a 20 loss. Each add therefore has to be sized against the current stop, and the stop has to move up to the invalidation level of the latest add; a pyramid whose stop stays at the original level is simply a large position bought near the top.

## Adding to winners only pays in the regime that has winners

Time-series momentum returns come from a minority of long, persistent trends; most signals end in small losses and the year is made by the few that keep going. Pyramiding amplifies exactly that skew, adding size where the trend has already proven itself and leaving the failed entries at their original, smaller size. In a choppy market the arithmetic reverses: every add is made at a local high and every trailing stop is hit on the next swing, so the pyramid loses more often than a single fixed position would. QANTERION's momentum strategies do not pyramid; they hold minute-level positions under a fixed allocation and a drawdown cap.

## How it gets misread

The common mistake is to add to a winner because "it is working" without moving the stop, which raises exposure at the moment reversal risk is highest and turns a small position with a small stop into a large position with a large one. The other is to confuse pyramiding with averaging down: adding to a losing trade to lower the average entry is a martingale, and the size it demands grows fastest precisely when the trade is proving the thesis wrong.

## Calculate it

- Position Size Calculator: https://qanterion.com/tools/position-size-calculator
- Risk of Ruin Calculator: https://qanterion.com/tools/risk-of-ruin-calculator

## See also

- Position sizing: https://qanterion.com/glossary/position-sizing
- Trend following: https://qanterion.com/glossary/trend-following
- Momentum trading: https://qanterion.com/glossary/momentum

## Sources

- [Time Series Momentum](https://www.aqr.com/Insights/Research/Journal-Article/Time-Series-Momentum) — Journal of Financial Economics (Moskowitz, Ooi & Pedersen), hosted by AQR
- [A New Interpretation of Information Rate](https://www.princeton.edu/~wbialek/rome/refs/kelly_56.pdf) — J. L. Kelly, Jr., Bell System Technical Journal (hosted by Princeton University)
- [Good and bad properties of the Kelly criterion](https://www.stat.berkeley.edu/~aldous/157/Papers/Good_Bad_Kelly.pdf) — MacLean, Thorp & Ziemba, hosted by the UC Berkeley Department of Statistics

## Other languages

- 简体中文: https://qanterion.com/zh/glossary/pyramiding
- 繁體中文（台灣）: https://qanterion.com/zh-TW/glossary/pyramiding
- 繁體中文（香港）: https://qanterion.com/zh-HK/glossary/pyramiding
- 日本語: https://qanterion.com/ja/glossary/pyramiding
- 한국어: https://qanterion.com/ko/glossary/pyramiding
- Português (Brasil): https://qanterion.com/pt-BR/glossary/pyramiding
- Türkçe: https://qanterion.com/tr/glossary/pyramiding
- Bahasa Indonesia: https://qanterion.com/id/glossary/pyramiding
- العربية: https://qanterion.com/ar/glossary/pyramiding

## Risk notice

Definitions are educational. Nothing here is investment advice, and no metric described on this page predicts future results.