Meridian
"I trade the moment the exhausted sprinter collapses — one pattern, one direction, always."
Meridian
"I trade the moment the exhausted sprinter collapses — one pattern, one direction, always."
Meridian is the most focused agent on the desk. While Wicker trades multiple sweep patterns in both directions and Null trades gap fills in both directions, Meridian does exactly one thing: it waits for a specific price level to get briefly breached — with extreme buyer exhaustion confirmed — and then it bets on the fall.
Meridian only goes short. It never bets that a price will rise. It exclusively profits when a price that briefly pushed too high comes back down. This isn't a limitation — it's a deliberate design. Meridian found that this one specific bearish pattern is highly reliable, while the opposite bullish pattern (buying a brief dip at a prior low) has proven to be a statistical loser across two independent tests.
Because the pattern is rare and highly filtered, Meridian only trades about 8 times per year per instrument. Each trade is high-conviction, carefully gated, and executed precisely.
Imagine the price level is a famous ceiling in the sky. All morning, the bulls have been running hard, pushing the price higher and higher. It looks like a powerful breakout. But look closely at the chariot of the sun driven by Apollo. The horses are running at their maximum capacity, pulling the heavy golden chariot to the absolute zenith of the heavens.
Just as they reach the highest peak (the meridian high), they hit the wall of exhaustion. They have no energy left to continue. The breakthrough was their final, desperate burst of strength. As the horses falter at the zenith, the chariot inevitably begins its descent.
That descent is Meridian's trade. The brief spike above the ceiling is a manipulation sweep — institutions pushed the price up to fill their own short positions. Meridian waits for the sun to reach its zenith (RSI above 88), confirms the horses are spent, and enters short as the chariot begins to fall.
Meridian will not trade unless three independent conditions are met simultaneously. All three must be true. If any one is missing, there is no trade that day, no matter what.
Price must spike above the prior trading day's session high — then close back below it within the same candle. A true breakout keeps climbing; a sweep closes back down immediately.
The RSI exhaustion indicator must be above 88 at the moment the sweep candle closes. RSI above 88 is an extreme reading — buyers have pushed far too hard for too long. They're spent.
The trend-strength indicator must be between 18 and 32. Too weak (below 18): no real directional character. Too strong (above 32): full-blown trend, reversal patterns fail.
The ceiling breach alone isn't enough. Prices break through prior highs all the time in genuine uptrends. The breach only matters if it's a false breakout, and you need the other two gates to confirm it.
RSI above 88 alone isn't enough. A market can be very overbought and keep pushing higher. Extreme overbought conditions in a strong trend just mean the trend is powerful. You need the structural false breakout to confirm institutions are unloading.
ADX in the band alone isn't enough. Medium-range ADX (18–32) just means the market has some directional character. All three together create a very specific, very rare situation where the exhaustion is confirmed, the level matters structurally, and the trend conditions are exactly right for a reversal to hold.
RSI stands for Relative Strength Index. It's a number from 0 to 100 that measures how overbought or oversold a market is in recent sessions. But Meridian uses a 2-period RSI, not the standard 14-period one. The 2-period RSI reacts extremely fast — it measures the last 2 candles, not the last 14. This makes it hypersensitive.
"Overbought" territory. The market has moved up significantly faster than normal. In classic analysis, this suggests caution. But in many trend-following markets, overbought conditions persist for days.
Extreme exhaustion. This level occurs roughly once every 10 sessions on average. It means buyers have pushed relentlessly for an extended period with almost no pullback. Energy is nearly depleted.
A standard 14-period RSI is like measuring the average endurance of the chariot's horses over their entire journey — it gives you a smoothed sense of their health. The 2-period RSI is like monitoring the flare of their nostrils at this exact stride. If their nostrils are wide and fire-breathing, they have just made a frantic, desperate surge. That surge cannot be sustained, and Apollo's sentinel knows they will falter on the next step.
Meridian watches for that nostril-flare at the moment the false breakout closes. Maximum exertion + failed structural push = the sprint is over.
Volatility check. Calculate yesterday's full price range (high minus low). Compare to the 20-day average daily range. If yesterday was an unusually wild day (range > 1.25× average), today's session is flagged as elevated volatility — no trades. Wild days bleed into the next morning's structure.
ADX check. Pull the 15-minute ADX reading. It must be between 18 and 32. If it's below 18, the market is directionless — sweep patterns fail in choppy markets. If it's above 32, the trend is too powerful — reversals don't hold.
Mark the prior session high. Draw a horizontal line at yesterday's highest price (9:30 AM – 4:00 PM ET). This is the only level Meridian cares about. Prior lows are deliberately ignored.
Wait for the spike. Watch for price to push above the marked prior-session high — but close back below it within the same candle. That close-back is the structural signal: a real breakout doesn't close back immediately.
Confirm RSI above 88. At the exact moment the sweep candle closes back below the level, check the 5-minute RSI(2). If it's below 88, skip the trade.
Find the Fair Value Gap. Drop to the 1-minute chart. The displacement candle that formed during the reversal left a tiny price imbalance — the zone where price moved too fast. Enter short at the midpoint of that gap as price drifts up to fill it. Limit order. If price doesn't retrace into the gap within 15 minutes, the setup is void.
Two instruments, independent signals. Meridian watches QQQ and SPY separately. A signal on QQQ is independent from a signal on SPY. Both can trade the same day if both independently qualify.
Stop-loss. Placed 2× ATR above the swept wick. If price pushes back above that level, the reversal thesis is wrong — exit. The stop is wider than Wicker's because the entry is on a 5-minute chart rather than a 1-minute chart, requiring more room.
Profit target. 1.5× the stop distance, minimum. Take 50% off at 1× the stop distance (i.e., at 1:1 R:R), then move the stop to breakeven on the remaining 50% and target 1.5R. If the prior session's low is reachable before the 1.5R level, let the trade run to there.
Hard time exit. All positions closed by 12:30 PM ET. No exceptions. Afternoon sessions are unpredictable for this type of pattern.
Daily loss limit. If two trades are stopped out in the same day, no more entries across either instrument. Cap the damage.
Meridian tested the opposite direction extensively. The "mirror image" trade — buying when a price briefly spikes below the prior session low and reverses — was tested twice and failed both times:
| Pattern | Direction | Win Rate (Test 1) | Win Rate (Test 2) | Verdict |
|---|---|---|---|---|
| Prior HIGH sweep | Short ↓ | 68.8% ✓ | Consistent ✓ | Deployed |
| Prior LOW sweep | Long ↑ | 46.2% ✗ | 37.5% ✗ | Abandoned |
Two tests, same filter, opposite result. The bearish pattern works; the bullish analog doesn't. Rather than forcing symmetry where none exists, Meridian accepted the asymmetry.
The sky is asymmetric. Flying upward toward the sun is a trial of fire that melts waxen wings (buying sweeps in a crash leads to immediate destruction), whereas falling downward back to earth is the natural, inevitable pull of gravity. Icarus' flight showed that rising into the sky has a different mechanical reality than falling. Meridian respects this celestial physics: it only trades the fall (going short), leaving the ascent alone.
Meridian turned left twice. Both times: disaster (below 50% win rate — worse than a coin flip). The pattern simply doesn't work in that direction. So Meridian turns right only, forever, until new data says otherwise.
One of the strongest things about Meridian's current backtest (v6) is that it deliberately covers three very different market environments — not just the recent bull market.
The Nasdaq fell 35% peak-to-trough. Volatility was high. Meridian proved its resilience by surviving this bear market structure without performance breakdown.
Strong recovery rally. 73% combined win rate on 15 trades. Prior-high sweeps are a common occurrence when bulls push hard, making this environment highly favorable.
75% combined win rate on 12 trades. Two consecutive bull years with strong results, validating the core edge in modern trend regimes.
The strategy needed at least 40 total trades to be statistically valid, but 2 years only produced 32. The solution was to extend the test window backward to include 2022 — adding real historical regime diversity. Adding a third instrument (like IWM) would have introduced unknowns. Extending the existing instruments' history is the more conservative, honest approach.
Meridian is in Phase 3 — paper trading is active. The Backtest Agent granted clearance on strategy v6, confirmed via a v7 resubmission that corrected contract sizing (strategy parameters unchanged). All three founding agents are now running live paper trades in parallel.
42 trades over 3 years (July 2022 – June 2025). 69.0% combined win rate. Profit factor 2.17. Max drawdown 1.9%. QQQ win rate 71.4%, SPY win rate 66.7%. All 6 benchmarks pass. The SPY win rate is tracked separately in Phase 3 — it's the most fragile component and suspension triggers if it drops below 50% on a rolling 20-trade window.
Meridian has filed a concentration regime reflection — five specific questions for the Backtest Agent about how the v6 strategy performed in high-concentration vs. pre-2020 market structure. After the first Phase 3 retrospective, those answers will inform whether a v7 strategy is warranted.