Momentum Alchemist
- Göstergeler
-
Antonello Belgrano
MQL5 Programmer specialized in creating professional Expert Advisors and Indicators.
I develop high-quality strategies as well as custom solutions, optimizations, and personalized Expert Advisors tailored to your specific needs.
Feel free to contact me for any requests or custom projects. - Sürüm: 2.0
- Etkinleştirmeler: 5
Two independent detection systems. One confirmed signal.
Momentum Alchemist fuses two complementary engines into a single, non-repainting confluence tool for MetaTrader 5:
- Force Trend System — a pressure/trend-state engine built on a volume-weighted moving structure and an ATR-based Supertrend filter. It reacts fast to shifts in buying and selling pressure.
- Pattern Recognition Engine — a historical pattern-matching module that scans thousands of past bars, finds the price structures most similar to the current one, and projects a weighted bias from how those situations resolved.
Instead of trading two indicators separately, Momentum Alchemist only upgrades a signal to Strong Buy / Strong Sell when both engines agree — filtering out low-conviction noise and highlighting the moments where trend pressure and historical precedent point the same way.
Key Features
- Dual confirmation logic — Buy/Sell arrows from the Force Trend System, escalated to Strong Buy/Strong Sell only when the Pattern Recognition Engine's bias confirms direction.
- Visual Pattern Line & Cloud — a smoothed projection line, color-coded bullish/bearish/neutral, with an ATR-based envelope showing the engine's expected range.
- Historical match visualization — optional boxes and labels on the chart marking the most similar past patterns, with similarity scores, so you can see why the engine reached its conclusion.
- Configurable EMA overlay — optional 20/50/200 EMAs for quick trend-context reference.
- Built-in alerts on new Buy/Sell events.
- Fully parametrized — sensitivity, scan depth, pattern length, smoothing, and thresholds are all adjustable, so the tool can be tuned per instrument and timeframe.
How It Works
- The Force Trend System continuously tracks buying/selling pressure against a dynamic threshold band; a raw signal fires when pressure crosses that band with sufficient conviction.
- In parallel, the Pattern Recognition Engine measures the current bar's short-term momentum, RSI, volatility, and relative volume, and compares this "fingerprint" against historical bars to rank the closest matches.
- If the projected bias from those matches agrees with the prevailing trend direction, the current signal is confirmed as Strong — otherwise it remains a standard Buy/Sell.
Usage Configurations
The indicator supports several distinct trading approaches, not just one "confirmation" workflow. Pick the one that matches your style:
1. Trend-following baseline (works well on its own) Take Buy arrows only when price is above the EMA 200 (magenta line); take Sell arrows only when price is below it. This alone already cuts out most counter-trend noise and is a solid, self-sufficient setup on H1–H4 — no need to wait for anything else.
2. Strong-only swing trading Ignore the plain Buy/Sell arrows entirely and trade only Strong Buy/Strong Sell. Far fewer trades, but each one has both engines pointing the same way — best suited to H4/D1 swing positions where you want quality over frequency.
3. Early-entry / scalping setup On M5–M15, take the plain Buy/Sell arrows as-is with a tight stop just beyond the recent swing low/high. Use the Pattern Cloud edge as a first partial-profit target. If a Strong signal shows up afterward in the same direction, it's a natural point to add to the position.
4. Pattern Line as a standalone bias tool Forget the arrows completely: just trade in the direction the Pattern Line is colored (green = bullish bias, red = bearish bias), using EMA 20/50 pullbacks as your entry trigger. This turns the indicator into a pure directional-bias filter for your own entry technique.
5. Mean-reversion / fade setup Watch for Strong signals that fire when price is stretched far from the EMA 200. Instead of following them, treat them as exhaustion points and fade back toward the average, using the Pattern Cloud as a rough target zone.
6. Multi-timeframe read Check the Pattern Line color on a higher timeframe (e.g. H4) for directional bias, then only take plain Buy/Sell arrows in that same direction on a lower timeframe (e.g. M15) for entries.
Choppiness warning When the red/green Pattern Cloud bands run tight and parallel, hugging close to the Pattern Line instead of spreading out, that's a sign of a choppy, low-conviction market. Signals firing in that condition are more prone to being false — treat them with extra caution or sit them out, regardless of which configuration above you're using.
Tuning Notes
- Timeframes: works on any timeframe, but H1–H4 give the Pattern Recognition Engine enough historical depth for meaningful matches without excessive noise. On M1–M15, raise Historical Scan Depth to keep the sample relevant.
- Scan depth vs. speed: Historical Scan Depth and Top Similar Matches drive computation cost. Start around 300/5 and increase gradually — very high values slow chart loading, especially on lower timeframes.
- Sensitivity: lower Similarity Sensitivity widens what counts as a "match" (more matches, noisier bias); higher values narrow it to near-identical historical setups (fewer, more precise matches).
- Volume source: if your broker/symbol has unreliable tick volume (common on some CFDs/indices), enable Use Real Volume where real volume data is available, for more accurate pressure calculations.
- Match boxes are diagnostic — use them to see why the Pattern Line is biased a certain way, not as entry/exit zones by themselves.
- Forward-test any configuration on a demo account for a few weeks before going live — optimal Sensitivity , Pattern Length , and Smooth Length vary by instrument and volatility regime.
Recommended Starting Setup
| Parameter | Suggested Start | Notes |
|---|---|---|
| Historical Scan Depth | 300–500 | Increase for lower timeframes |
| Top Similar Matches | 5 | Raise to 8–10 for smoother (less reactive) bias |
| Pattern Length | 10 | Shorter = more reactive, longer = more stable |
| Similarity Sensitivity | 1.5 | Lower = broader matches, higher = stricter matches |
| Smooth Length | 50 | Controls Pattern Line responsiveness |
Disclaimer: Always test on a demo account before live use.
