Pyramiding Strategy: Did Chasing Wins Bring Max Profit or Risk?
## What's the idea?
A beginner-friendly summary of the verification: “Pyramiding Strategy: Did Chasing Wins Bring Max Profit or Risk?”.
Breakout entry example (XAUUSD daily, real data): buy when price breaks above the recent high.
What’s the idea?
We’ve all heard the trading adage, “Let your winners run,” or “add to a winning position.” It sounds great in theory, right? You jump on a trend, and as it keeps going your way, you add more fuel to the fire, riding that “winning horse” for maximum profit. This is called pyramiding, and it’s what we wanted to explore in this research.
Specifically, we looked at a classic Turtle-style pyramiding strategy. Imagine you enter a trade when the price breaks out of a certain range. If the trend continues positively, you add another “unit” (another portion of your trade) for every X amount of price movement (often based on ATR, or Average True Range, which measures market volatility). You keep doing this up to a maximum number of units, with each unit having its own ATR stop-loss to protect profits. The whole position is then exited if the price breaks out of the trend in the opposite direction.
How I tested it
To even test pyramiding, we first had to beef up our trading engine. Our existing system was mostly designed for single positions, so we essentially gave it an upgrade to handle multiple, simultaneous positions going in the same direction. This was a significant technical expansion, allowing for things like partial profit-taking or different rules in the future. The good news? This core engine change was fully backward compatible, meaning our existing, proven strategies continued to work exactly as before – a crucial regression check passed with flying colors!
With the engine ready, we applied the PyramidBreakout strategy to our standard robust5 H1 setup (a robust set of parameters on a 1-hour chart). We wanted to see if “riding the winning horse” could genuinely improve our results.
What happened?
At first glance, the results were enticing! When looking at the ret/DD ratio (Return divided by Drawdown, a measure of how much return you get for the risk you take), pyramiding seemed to improve things. Our baseline system had a ret/DD of 11.7, but with pyramiding, this jumped to 14.5 and even up to 20.2! This suggested that yes, pyramiding can make more money when a strong trend develops. However, the devil, as always, was in the details. While the overall return might have looked better, the drawdown (DD) – the peak-to-trough decline in our capital – absolutely exploded! We saw drawdowns increase by a staggering 31% to 53%. Imagine your account value taking a much wilder rollercoaster ride, with deeper valleys than before. The Sharpe Ratio, which tells us how much return we get per unit of risk, also suffered, dropping from 1.02 to a range of 0.84-0.96. In simpler terms, we were taking on significantly more risk for each unit of return. And as for trailing stops – a common way to lock in profits as a trade moves in your favor – they completely fell apart under the pyramiding strategy.
The Problem with “Prop” Rules
This is where the rubber really met the road. Our primary goal is to run strategies suitable for “prop” accounts – proprietary trading accounts that often have very strict rules designed to protect capital and allow for continuous withdrawals. These rules typically include a hard limit on daily maximum loss (e.g., no more than -5% in a single day). When we evaluated the pyramiding strategy against these prop rules, the results were catastrophic:
- Drawdown: Even after trying to minimize risk per unit, the overall drawdown was still a massive -18% to -28%.
- Intraday Worst Loss: On a single day, our capital could drop by an unacceptable 11.6% to 19.9%.
- Intraday Disqualification: We faced 9 to 19 days where the strategy would have hit the daily loss limit and been disqualified!
- Total Capital Loss (Margin Call): The strategy put us at risk of losing 55% to 70% of our total capital.
- Maximum Loss Disqualification: There was a 17% to 25% chance of hitting a total maximum loss limit and being completely disqualified from the account. In contrast, our existing, non-pyramiding confirmed system (which is our baseline) showed excellent stability: a maximum DD of -7.1%, an intraday worst loss of 4.06%, 0 disqualification days, 78% capital retention, and 0.6% chance of maximum loss disqualification. While pyramiding did increase the potential monthly profit (from 0.58% to 0.90%), this came at the cost of completely destroying our safety margins.
What I learned
The big takeaway is clear: Pyramiding is fundamentally unsuitable for prop-style trading accounts with strict daily loss limits. Here’s why: Pyramiding concentrates your exposure onto strong trends. This means that if the trend suddenly reverses – which will happen eventually – all those accumulated winning positions can turn into significant unrealized losses very quickly. Even if the system might recover later, these deep, sudden reversals can easily trigger daily maximum loss limits, causing immediate disqualification. It’s a structural contradiction: a strategy designed to capitalize on extended trends clashes directly with rules designed to limit single-day volatility. While the overall ret/DD might look appealing on paper (based on maximum drawdown), the intraday and daily loss metrics paint a brutally honest picture of unmanageable risk. This reaffirms a crucial lesson from previous research: intraday performance and daily loss limits are the true constraints in a prop trading environment.
Lasting Achievements
Despite the ultimate rejection of pyramiding for our prop strategy, this research wasn’t a failure! The multi-position engine we developed is a permanent and valuable upgrade to our core infrastructure. This robust foundation will be incredibly useful for future explorations, such as implementing partial profit-taking strategies or exploring different trading rules that might benefit from managing multiple positions. For now, our existing, non-pyramiding system remains the best and safest option. It consistently delivers results without exposing us to the unacceptable risks of pyramiding under strict prop rules.
The Ongoing Challenge
This research also reconfirmed a challenging reality: achieving a consistent 2% monthly return with a safe drawdown using only price data and simple leverage (as explored in Research 52, which showed a ceiling around 0.5%) combined with the unsuitability of pyramiding, remains incredibly difficult. It pushes us to continue seeking innovative, yet safe, approaches to algorithmic trading.
How this connects
This verification builds on earlier ones (what failed before and what I tried this time, comparisons between approaches).