The last post covered the structural flaw in a 25-step grid — how the risk-reward ratio quietly broke down in the later stages. This post focuses on a different problem that lived alongside it: the entry point itself. Changing a single number — how far below the recent high the grid starts buying — turned out to reshape the strategy more than I expected.
The Grid Was Sitting Out the Best Opportunities
The original design didn't start buying until the price had already fallen 4.5% below the recent high. The logic seemed reasonable at the time: "wait for a meaningful pullback before entering, so the entry has some depth behind it."
The problem showed up in the data. Stocks with real buying momentum — the kind this strategy was specifically designed for — rarely pulled back 4.5%. They dipped 1% or 2%, and then broke higher again. The 4.5% threshold wasn't filtering out weak opportunities; it was filtering out the strongest ones, leaving the strategy to pick up entries only when momentum had already stalled or reversed. That's the opposite of what a breakout-based approach should do.
What Changed When the Entry Moved to -1.0%
Shifting the starting entry to 1.0% below the recent high had an immediate and visible effect: the frequency of actual entries went up significantly. This wasn't just "more trades" for the sake of it. It meant the system was now engaging with the full range of relevant opportunities rather than waiting through the best of them.
There's a subtler reason this matters too. A shallower entry threshold isn't just about getting more chances — it also reflects a more honest read of how strong stocks actually behave. If the idea behind the strategy is to enter stocks that show real buying pressure, it doesn't make sense to require a deep pullback before acting. Deep pullbacks in strong stocks are often a sign that the buying pressure has faded, not confirmed it.
Moving the entry earlier meant accepting a different kind of noise: more instances where the shallow dip turns out to be the beginning of a larger decline. That's a real tradeoff, and managing it required strengthening the other filters in the system — trading volume, trade intensity, the quality of the overall pattern — rather than relying on entry depth alone.
The Stop-Loss Came In With It
Adjusting the entry point didn't happen in isolation. Tightening where buying starts made it necessary to also tighten where it stops. The original stop-loss boundary sat at -15.0% — a very wide band, which meant the system could keep buying all the way through a severe decline. Paired with a -4.5% entry, that 10.5% range felt intentional. But paired with a -1.0% entry, it would allow the system to accumulate all the way through a 14% drop, which was far too wide.
The stop-loss boundary moved to -8.0%, keeping the active buying zone at a tighter, more controlled width. The number of grid steps dropped from 25 to 15 as well — the narrower range simply didn't need 25 stages to cover it, and having too many stages with small capital at each one made the average-cost improvement nearly meaningless in practice.
Every Number Is a Tradeoff
The underlying lesson wasn't specific to -4.5% or -1.0%. It was that every parameter in a trading strategy involves a tradeoff, and the right number can only be validated against actual data — not intuition.
Setting the entry threshold shallower opens up more opportunities but brings in more noise. Setting it deeper reduces noise but misses the best setups. Setting the stop-loss wider gives the trade more room to breathe but increases maximum loss exposure. Setting it tighter reduces exposure but increases the chance of getting stopped out on normal volatility — a phenomenon called a whipsaw.
What makes these tradeoffs manageable is treating them systematically: run the numbers at each candidate setting, look at what actually improves and what gets worse, and make the decision based on that rather than a feeling. The shift from -4.5% to -1.0% wasn't obvious until the data made it clear.
Today's Investing Insight — Pullbacks and Where They Tend to Stall
In technical analysis, the depth of a pullback in an uptrending stock is often mapped against Fibonacci retracement levels — 23.6%, 38.2%, 50%, 61.8%, and so on. The reasoning behind using these levels isn't that they carry any inherent market magic; it's that enough traders reference them that they become self-fulfilling to some degree, with real buy and sell orders clustering around those price zones. That said, how deep a "normal" pullback runs varies significantly from stock to stock and from one market environment to another. A fixed percentage threshold applied uniformly across all stocks will be too tight for some and too loose for others — which is exactly why later in this project, I replaced fixed percentage thresholds with a volatility-adjusted measure called ATR (Average True Range). That comes up later in this series.
---
This post documents a personal journey of building an algorithmic trading system and is not a recommendation of any specific strategy. Parameter values described here are for illustrative purposes only and do not guarantee results in any market condition. All investment decisions and their outcomes are the sole responsibility of the investor.
Comments
Post a Comment