The dragonfly is the deadliest hunter on Earth. Not the lion (25%). Not the great white (50%). Not the cheetah (58%).
The dragonfly catches its prey 95% of the time.
Biologists studying Plathemis lydia in the 2010s found something that should haunt every active trader: dragonflies don't chase. They intercept. Their tiny brains run a real-time calculation — where will the prey be when I arrive? — and steer to that future point, not the prey's current position.
Cheetahs chase. Wild dogs chase. Most traders chase. They see a stock breaking out, they buy. They see a stock crashing, they sell. They are reacting to where the prey was — not where it will be.
Reactive hunters lose 50-75% of the time. The dragonfly's interception math wins 19 times out of 20.
This essay is about how to stop chasing.
The interception equation
A dragonfly tracking a fly doesn't ask "where is the fly?" It asks two questions simultaneously:
1. Where is the prey heading (vector) 2. Where will I be when our paths cross (interception point)
It then flies straight to that crossing point. The prey, predictably, lands there a moment later.
For markets, the equivalent questions are:
1. What direction is this stock's relative strength accelerating? (RS Rating delta) 2. At what price level will the stock break out + my entry stop be defined? (pivot + risk)
A reactive trader watches the breakout candle print and chases the close. A predictive trader has the buy order at the pivot price two days before the breakout — and a stop loss already calibrated. When the breakout fires, they're already in. Like the dragonfly, they were never chasing — they were waiting at the interception point.
This isn't day-trading wizardry. It's the basic IBD-CANSLIM playbook William O'Neil published in 1988, dressed up in dragonfly biology.
The three things the dragonfly teaches us
1. Patience is computed, not felt
The dragonfly hovers. It does not feel boredom. Hovering is the optimal state until the interception math says go.
For traders: most days the math says no. No setup. No edge. No trade. The dragonfly does not chase a fly that's going the wrong direction; it waits for one whose path crosses its own. Most days the right answer is to do nothing.
If you traded only the days a dragonfly would have moved, you'd trade about 1-3 times per week.
2. The kill is calibrated, not improvised
When the dragonfly strikes, the entire move is pre-computed: angle, speed, jaw timing. There is no improvisation in the strike itself. All the work happened in the hover.
For traders: all the work should happen before the buy. Define the pivot. Define the stop. Define the size. Define the partial-profit point at 2R. Define the trail rule. Define the climax exit. When the breakout fires, you are not deciding — you are executing a plan you wrote during the hover.
3. Misses are accepted, not chased
Even the 95% dragonfly misses 1 in 20. When it misses, it does not pursue. It returns to the hover.
For traders: stops mean stop. When the trade hits the stop, you are out. You do not pull the order. You do not rationalize. You return to the hover. The hunter that chases its miss becomes the prey.
What this looks like in practice
Our research engine (MOEasymmetry, paper-traded since November 2025) implements interception via three layers:
1. Tracking — a daily relative-strength scan ranks ~5,000 US stocks and ~700 Thai stocks. Stocks that begin accelerating against the universe enter the watchlist days or weeks before they break out.
2. Hovering — the watchlist sits. Pre-computed pivot prices, stop levels, and position sizes are attached to each name. Some sit for 60+ days before the math fires.
3. Striking — when the pivot breaks on volume, with relative strength still accelerating and the broader market in a confirmed uptrend, the order fires automatically. The trader is not deciding in the moment — they are executing a plan written during the hover.
This is not novel. It is how O'Neil traded since 1958. It is how Mark Minervini compounds 220% a year. It is how Kristjan Qullamaggie turned $5K into $100M.
It is also the dragonfly's interception, dressed in price charts.
The honest part
We have not beaten the market with this. Not yet.
After running 20-window walk-forward across 2006-2025 (covering the 2008 GFC, 2011 Eurocrisis, 2015 China devaluation, 2018 Q4 mini-bear, 2020 COVID, and 2022 rate-hike bear), we have two validated entry-exit combos:
- failed_reentry × Minervini partial-take — Oliver Kell's "biggest winners take 2-3 attempts" paired with Minervini's 2R partial profit + 21-DMA trail. 13 of 20 walk-forward windows positive.
- first_pullback × Webster Power Trend exit — Komar/Patel's "first pullback to MA" paired with Webster's index-based exit (21-EMA below 50-SMA → exit all). +536R cumulative over 20 years on the realistic next-day-open execution model.
Both at 0.5% risk per trade, 5 concurrent positions max, ρ correlation between systems = -0.19 (genuine diversification).
The single most important finding: the system did not lose money in 2008. The regime filter (Paul Tudor Jones's 200-day veto + Weinstein's Stage 2 gate) correctly kept the system out of the worst bear market in modern history. That is what regime filters are for. They worked.
Two months ago we thought we had six validated systems. Walk-forward killed four of them. Most of what looked like edge in-sample was the bear-recovery of 2024-2025, not real signal. We delete what doesn't work, even when we built it. Especially when we built it.
We are starting a six-month paper-trade period right now. First real-capital deployment: November 1, 2026. We publish every paper trade openly — wins and losses both.
How to start hovering
You don't need our system to apply this. You need:
1. A universe (e.g., S&P 500, Nasdaq 100, SET100) — the prey field 2. A relative-strength rank refreshed weekly — to find what's accelerating 3. A pivot price + stop loss for each watchlist name — to define the interception point 4. The discipline to do nothing on most days — to hover
Most traders fail not because they pick wrong stocks but because they buy at the chase, not at the interception. The dragonfly's 95% is not skill at killing — it is skill at not flying until the math says go.
Hover well.
What's coming next
Two more issues this month:
- Issue #2 (May 18) — DCA Truth: I ran ten years of TISCO DRIP simulations on Thai stocks. The numbers will surprise you. Some "obvious" winners produced negative real returns when adjusted for inflation. Three quiet stocks 10x'd. The math is honest.
- Issue #3 (May 25) — Charts Are Risk Maps, Not Crystal Balls: I tested 9 famous chart patterns and concluded they all "failed." I was wrong. Not because patterns work — because I tested the wrong question. The unifying principle that connects Minervini, O'Neil, Webster, and Komar.
Both will be free for the first month. Subscribe at [link] to get them in your inbox.
This is research-not-advice. Past performance does not guarantee future results. We publish our paper-trade record openly, including failures. The two validated systems above are paper-only until November 1, 2026 — six full months of live observation before any real capital deploys. We honor the Thai SEC's research-not-advice framing throughout.
— MOEasymmetry · Track. Study. Wait. Strike. · ตามรอย ศึกษา รอจังหวะ จู่โจม