Live CFTC Commitment of Traders positioning, 30 TradingView technical ratings, and our own candlestick & chart-pattern detection, all in one place. Accounts are created by the administrator.
| XAU/USD | BUY XAU | High conviction |
| GBP/JPY | SELL GBP | Moderate conviction |
Ranked by a composite score: 60% net non‑commercial position (as % of open interest) + 40% week‑over‑week change in net position (as % of open interest). Positive = speculative crowd net long that currency; negative = net short.
| Pair | Bias | Conviction | Strength Gap | Net Pos (Base / Quote) |
|---|---|---|---|---|
| Loading pairs… | ||||
Three independent signals per card: the top badge is CFTC weekly positioning, the pattern badge is our own daily-bar scan, and the gauge is TradingView's own rating at whatever interval is selected above it. They won't always agree — different timeframes, different methods. Agreement across all three is a stronger signal than any one alone; disagreement usually just means lower conviction right now.
Same three independent signals as above, for all 28 pairs.
Independent of COT data. Each verdict comes from our own technical-rating engine — 10 moving averages (SMA/EMA at 10/20/50/100/200) plus 9 oscillators (RSI, Stochastic, CCI, ADX/DI, MACD, Momentum, Williams %R, Ultimate Oscillator, Bull/Bear Power), computed on a year of daily bars and combined into a single score. This is a transparent, from-scratch implementation of the same general methodology technical-rating gauges use — not a copy of any vendor's proprietary formula, and not investment advice. This will often disagree with the TradingView gauges on the Dashboard tab — those follow whatever short-term interval is selected there, while this table is a longer-horizon daily view. A pair showing "Strong Buy" here while its 1-hour gauge shows "Sell" usually just means a short-term pullback inside a longer uptrend, not a contradiction. In flat/choppy markets, moving averages can also cluster together and swing the verdict on very little real signal — check the Buy/Sell/Neutral vote breakdown before trusting an extreme reading.
| Pair | Verdict | Confidence | Score (Daily) | Votes (Buy / Sell / Neutral) |
|---|---|---|---|---|
| Loading predictions… | ||||
Independent of COT data and the daily Trade Prediction tab. Computed on 15-minute bars, refreshed automatically every 10 minutes (near-real-time, not tick-level -- see the chat for why). Covers Fair Value Gaps, Break of Structure / Change of Character, Liquidity Sweeps, Order Blocks, Premium/Discount zones, and harmonic patterns (Gartley, Bat, Butterfly, Crab). These are transparent, from-scratch implementations of published, mechanical definitions -- not a copy of any paid tool's proprietary formula, and not investment advice. Order Blocks are the least standardized of these concepts across different educators; weight that one a little lighter than the others.
| Pair | Zone | Signals |
|---|---|---|
| Loading… | ||
Get pushed to Telegram the moment something changes -- not spammed every 10 minutes for a signal that's still just sitting there. Alerts only fire on genuine transitions: a new decisive Institution Verdict, a Trade Prediction reaching Strong Buy/Strong Sell, or a brand-new Intraday SMC/Harmonic signal appearing.
1. Click the button below to open Telegram.
2. Press "Start" in the chat that opens.
3. Come back here -- this page will update automatically once it's linked.
If your plan includes the Private Signal Channel, linking Telegram also activates that access automatically.
Open Telegram to LinkReal outcomes, determined by actual price action hitting TP or SL -- not by whether the pattern is still detectable. Most recent first.
| Pair | Pattern | Outcome | Entry | SL | Resolved |
|---|---|---|---|---|---|
| Loading… | |||||
Computed directly from real, resolved signal outcomes -- not a backtest or an estimate. A "win" means TP1, TP2, or TP3 was hit; a "loss" means stop-loss was hit. Signals still active or superseded by a newer one aren't counted either way, since they haven't actually resolved. Sample size shown alongside every rate -- a handful of signals isn't a statistically meaningful claim, however the percentage looks.