About three months into running the signal system in live conditions, I reached a point where I felt the core logic was reasonably stable but the results weren't as consistent as I wanted. The system was generating signals, the signals were often followed by meaningful moves, but there was too much variability in outcomes that I couldn't fully explain. On a good day, four out of five signals led to worthwhile price action. On a bad day, the same signal quality metrics produced the opposite results.
My instinct was that something was missing from the design — a filter or condition I hadn't thought of. Rather than guessing what it might be, I spent about two weeks studying how professional scalpers in Western markets describe their entry frameworks. The comparison revealed three specific gaps in my approach that I hadn't recognized until I saw them side by side with how practitioners in other markets were solving the same problems.
How I Approached the Comparison
I didn't have access to proprietary trading systems or institutional methods. What I studied was publicly available material: detailed write-ups from professional day traders who had documented their specific entry rules, recorded trading sessions, and the reasoning behind their decisions. The goal wasn't to copy their approaches — Korean market microstructure is different enough from US markets that direct transplantation rarely works — but to understand the underlying principles and see which ones applied regardless of market.
Three things emerged consistently in their frameworks that were absent from mine.
The First Gap — Volume Behavior During the Pullback
The most consistent element across multiple Western scalping frameworks was attention to what volume does during the pullback phase, not just at the breakout. The standard formulation I kept encountering was something like: look for volume to decrease as the price pulls back, then look for a specific reversal candle pattern before the breakout attempt.
The logic is straightforward. If volume decreases during the pullback, it means sellers aren't particularly aggressive — they're not driving the price down, the price is just drifting lower because buyers have temporarily stepped back. That's a very different situation from a pullback driven by active selling, where volume stays elevated or increases as the price falls. The first scenario leaves the path clear for buyers to step back in. The second means you're walking into active selling pressure when you try to buy the breakout.
My system had a condition requiring the second pullback to be deeper than the first — which addressed part of this — but it had no explicit check on volume behavior during the pullback itself. I wasn't measuring whether the pullback was happening on decreasing or increasing volume. After this comparison, I added volume behavior during the pullback to my manual checklist for evaluating signals, even though I haven't yet fully automated it.
The Second Gap — Reversal Candle Confirmation
The second consistent element in Western scalping frameworks was the concept of a reversal candle — a specific candlestick formation at or near the low of the pullback that provides visual evidence the selling pressure is exhausted before the breakout attempt begins. The most common patterns cited were candles with long lower wicks relative to their bodies, where the price had reached a low during the period but then recovered most of that decline before the period closed — leaving behind a tail that shows buyers absorbed the selling.
My system was measuring the depth of the pullback and its relationship to ATR, but it wasn't looking at the structure of the candle at the low. A pullback that ends with a long-tailed candle suggests exhaustion more clearly than one that ends with a candle whose body reaches all the way to the low, which suggests selling was still dominant at the period's close.
I started reviewing the candle structure at the pullback low as part of my manual evaluation. The practical effect was modest but consistent: signals where the pullback ended with a long-tailed reversal candle tended to produce cleaner breakouts than signals where the pullback candle showed little evidence of buying absorption at the low.
The Third Gap — Risk-Reward Calculation Before Entry
The third and most operationally significant gap was something I had thought about in the abstract — risk-reward ratio — but hadn't been applying at the moment of each specific signal.
Western scalping frameworks consistently describe pre-trade risk-reward calculation as a mandatory filter: before entering any trade, calculate the distance to the stop-loss (risk) and the distance to the first meaningful target (reward), and only enter if the ratio meets a minimum threshold — typically 1:2 or better, sometimes 1:3 or 1:4. If the ratio is only 1:1 or worse at the moment of entry, skip the trade regardless of how clean the setup looks.
I was not doing this. My system generated a grade based on trading value, trade intensity, and tick acceleration at the breakout moment — but it wasn't calculating, for each specific signal, the actual distance from the breakout price to the stop-loss level and comparing that to a realistic target. The grade told me how strong the breakout was. It didn't tell me whether the trade, at that specific entry price with that specific stop-loss location, had a favorable risk-reward profile.
This turned out to explain some of the variability I had been seeing. On days when the pullback had been deep and my stop-loss was therefore relatively far away, even a grade A signal could produce a poor risk-reward entry if the realistic target was only marginally above the breakout point. I was entering based on signal quality without checking whether the entry geometry made mathematical sense.
After recognizing this, I started calculating — for every signal before acting on it — the exact distance to the stop-loss (one tick below the valley low) and a realistic target (the recent high plus a percentage), and checking whether the ratio was at least 1:2. This single addition reduced the number of trades I took but improved the average outcome on the ones I did take.
What Carried Over From Korean Market Observation
The comparison also confirmed that several elements of my framework were sound and aligned with what experienced scalpers in other markets had independently arrived at. The concept of waiting for a pullback rather than chasing the initial breakout — present in my system from the beginning — appeared consistently as a core principle. The use of volume and buying pressure confirmation at the breakout was also a shared element. And the time-based restriction on new entries — the 09:00 to 10:30 window in my system — mapped closely to the consistent observation in Western frameworks that the highest-quality setups tend to appear in the first 60 to 90 minutes after the market opens, when price discovery is most active.
The framework wasn't wrong. It was missing specific pieces that the comparison made visible.
Today's Investing Insight — VWAP as a Reference Level in Western Scalping
One indicator that appears constantly in US-focused scalping frameworks but has no direct equivalent in the approach I built for Korean markets is VWAP — the Volume-Weighted Average Price. VWAP is calculated continuously throughout the trading day and represents the average price at which every share has traded, weighted by volume. Many institutional participants use VWAP as a benchmark — they measure their execution quality against whether they bought below or sold above the day's VWAP. This institutional use makes VWAP a meaningful reference level for price behavior: pullbacks to VWAP often find support because institutional buyers, aware of the benchmark, tend to defend it. In US markets, scalpers frequently use the VWAP as a dynamic support and resistance reference for intraday entries. In Korean markets, where institutional VWAP benchmarking is less prevalent as a disclosed execution standard, this specific dynamic is weaker — which is one reason I chose tick-based moving averages rather than VWAP as the reference level in my secondary signal layer.
---
This post documents a personal journey of building an algorithmic trading system and is not a recommendation of any specific stock or strategy. Trading approaches described here are based on personal observation and experience and may not be suitable for all market conditions. All investment decisions and their outcomes are the sole responsibility of the investor.
Comments
Post a Comment