Pair Currency Strength Meter MT5
- Indicators
- Konstantin Chechnev
- Version: 1.0
- Activations: 5
Eight currencies, up to 28 pairs, and one convenient way to be wrong: assuming the market begins and ends with the chart you opened. Pair Currency Strength Meter reconstructs the full FX network, separates currency strength from one pair's random move, and searches for the cleanest expression of a strong-versus-weak campaign.
This is not a percentage table or a multicolored RSI. The indicator builds synchronized closed-M1 snapshots, solves a graph model for USD, EUR, GBP, JPY, CHF, CAD, AUD and NZD, grades the network's trustworthiness, and only then creates a scenario for an individual pair.
Network first, signal second
- Broker-aware Universe recognizes prefixes, suffixes and symbol families, selects one instrument per currency relation and avoids mixing incompatible feeds. Auto Server seeks the complete network; Market Watch, Saved Profile and manual modes can constrain it.
- Network Grade A–D measures eight-currency coverage, direct edges, components, minimum node degree, cycle redundancy, rank, conditioning, history coverage and timestamp synchronization. A pretty strength ranking is not presented as reliable when the network underneath is weak.
- Robust Graph Solver models each pair's return as base strength minus quote strength. Up to six IRLS passes with Huber 1.345, ridge regularization and conditioning checks suppress isolated outliers and recover eight latent currency factors.
- Causal normalization uses completed M1 snapshots and its own rolling median/MAD scale. A temporal firewall prevents the next minute from rewriting an earlier market state.
- Three horizons follow the selected style. Intraday reads 15/90/360 minutes, Balanced 30/180/1440 and Swing 240/1440/7200, keeping pulse, core move and context separate.
Currency strength still does not guarantee a good entry
Campaign Engine moves a currency through Candidate, Forming, Active, Mature, Fading, Resolved or Aborted. It evaluates breadth across counterpart currencies, coherence, persistence, displacement, residual concentration and data quality, and distinguishes expanding, stable, contracting, conflicted and data-hold states.
Every pair is tested as a possible clean strong-versus-weak expression: driver conflict, excessive residual, mature extension, wide spread, weak history and structural obstacles can reject it. Attention Engine promotes the best scenario instead of requiring 28 charts to be inspected manually.
Signal Planner creates entry, pullback, target and stop zones plus RR while checking price trend, factor gap, breadth, coherence, data trust, spread and a clear path. A plan proceeds WAIT → ARMED → ACTIVE → COMPLETED/CANCELLED, becomes a committed scenario and is not rebuilt around a known outcome.
Confirmation uses the executable quote side: Ask for BUY and Bid for SELL; target and stop are checked on the opposite side, so spread cannot vanish from statistics. Completed scenarios feed session wins/losses and R results, while cooldown suppresses the same story being sold as a new idea.
A campaign terminal, not a narrow meter
The full-screen Canvas scene shows the network pulse, eight currency rails, strongest and weakest currencies, best pair, direction, breadth out of 7, coherence, data trust, network grade and plan card. A live tick layer adds activity and flow organization without rewriting locked M1 analytics.
An embedded chronicle preserves event lifecycle and the active plan across reinitialization. If storage becomes unavailable, analytics continue read-only; a presentation-layer failure likewise does not erase the analytical state.
With chart Auto Scroll enabled, Auto Follow can switch the current chart to the best pair after a stable plan is confirmed and restore that locked plan. Manual history inspection takes priority, and symbol switching is disabled in the Strategy Tester.
Test the network before trusting the ranking
Run the demo in the visual Strategy Tester on M1 with Every tick based on real ticks. Let available companion-pair history load and wait for the network calculation; in the tester, the signal remains tied to the tested symbol.
Compare Network Grade, data trust, campaign state, actual Bid/Ask and the full WAIT → ACTIVE → result sequence—not just wins. Test quality depends on how much synchronized M1 network history the broker provides.
A market map does not press Buy
Pair Currency Strength Meter is an indicator, not an Expert Advisor. It never sends or modifies trading orders. Entry, stop and target form an analytical plan; size, execution, management and risk remain the trader's decision.
Strength, scores, session statistics and historical tests do not guarantee profit or repeatable results. Symbol availability, quotes, spread, history, latency, liquidity and broker conditions can alter the network, signal and actual execution.
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Settings: seven controls instead of 28 opinions
- Universe mode — Auto Server. Keep the full broker network initially; use Market Watch only when you intentionally control its symbol set. Check Network Grade and history coverage before judging signals.
- Trading style — Balanced. Intraday reacts faster; Swing needs substantially deeper M1 history. Change style separately from signal thresholds.
- Signal selectivity — 1.00 (0.70–1.40, step 0.05). Lower admits more candidates; higher demands a stronger campaign and cleaner pair expression. First range: 0.90–1.15.
- Confirmation strength — 1.00 (0.60–1.80, step 0.10). Controls snapshot persistence and dwell time: lower is faster, higher stricter. First range: 0.8–1.3.
- Entry depth — 1.00 (0.70–1.35) moves entry from shallow to deeper pullback. Target ambition — 1.00 (0.75–1.40) changes target distance. Optimize separately.
- Stop protection — 1.00 (0.80–1.35) adjusts structural protection. Spread tolerance — 1.00 (0.60–1.80) sets acceptable execution cost.
- Repeat cooldown — 4 hours (1–12, step 1) prevents one scenario from returning as a “new” idea. Interface language is automatic or one of 11 choices.
Defaults are a baseline, not a universal preset. First obtain a stable network and synchronized history; then alter one semantic control and validate on a later period.
