Commit Graph

10 Commits

Author SHA1 Message Date
Harwinder Sekhon ea306df484 Overhaul news engine: fix false signals, add surprise detection + caching
Improvements from a live-verified audit of the news/calendar pipeline:

- Currency tagging: keyword matching now uses letter-boundary regexes
  and drops ambiguous "us "/"eu "/"uk " tokens — "versus" no longer
  tags headlines as USD (confirmed live false positive polluting the
  XAU/USD sentiment average).

- Fetch caching: calendar events and headlines are cached class-level
  (10 min TTL, 60s retry after empty/failed fetch, lock-guarded for
  concurrent scans). Stops per-scan refetching that was tripping the
  Forex Factory rate limit; TTL stays under the 30-min pre-event
  buffer so the danger window can't go stale.

- Surprise detection: EconomicEvent.surprise is computed from parsed
  actual vs forecast (handles %, K/M/B/T suffixes, negatives). A
  surprise print widens the post-release caution window to 45 min;
  inline releases keep the original 15 min.

- Gold-aware sentiment: headlines are scored per currency, with
  monetary-policy keywords (hawkish/rate hike/tightening vs dovish/
  rate cut/easing/stimulus) inverted for XAU — "hawkish Fed" headlines
  scored +1.0 for gold before, -1.0 now (verified on live feed).

- New sentiment sources: Google News gold-focused query and CNBC
  Economy RSS (verified live; 40 → 74 unique headlines). Added
  cross-feed title dedup so syndicated stories count once.

- events_today now buckets by UTC date (feed timestamps carry US-
  Eastern offsets); --json output includes news_description.

- Removed dead NewsConfig.min_impact and unused HIGH_IMPACT_EVENTS.

Tests: 142 passed (7 new covering boundary matching, surprise
classification, post-event windows, cache reuse, gold inversion).
2026-07-10 09:43:11 -04:00
harwinder 04328eaeab Refocus on XAU/USD: add momentum indicators, tune TP, exhaustion block
Strip EUR/USD, GBP/USD, USD/JPY pairs and strategies from entire codebase.
Add 3 new scored indicators (ROC, Keltner/TTM Squeeze, ADX slope) and
enable MACD-as-pillar + divergence detection for early trend entry.
Introduce "transitioning" regime (ADX < 25, slope > 3.0) to catch trend
births before ADX confirms. Gold-specific composite weights per regime.

Tune TP multipliers: quality-tiered scaling (A+/A get 1.5-2.0x ATR via
XAU ATR method, B/C use swing-level targeting). Backtest: +24R total,
65% WR, PF 1.12 over 365 days.

Add momentum-confirmed exhaustion block: hard-reject trades when 2+
static exhaustion signals fire AND ROC/MACD confirm fading momentum.
Prevents late trend entries without killing strong trend trades.

Telegram: simplify buttons to timeframe-only labels, add "Scan All TFs"
multi-timeframe summary, add /scan command, fix BB/KC level breach keys,
fix price formatting (.2f for gold), fix deprecated asyncio call, fix
"1 pairs" grammar and 60s/30s dashboard refresh mismatch.
2026-05-15 10:51:51 -04:00
harwinder 058ef57d8f Pro tier upgrade: quant rearchitect + WebSocket level-breach monitor
- Restore XAU/USD (Pro plan), drop USD/CAD, raise rate limit to 55/min
- Reduce signal engine to 4 uncorrelated pillars (EMA, RSI, BB, Candlestick)
  with regime-adaptive weights (trending boosts EMA, ranging boosts RSI+BB)
- Add structure-based S/R (swing highs/lows) for trade setup targeting
- Fix confidence death spiral: all adjustments now additive, not multiplicative
- Dynamic R:R trade setup: SL behind swing level, TP at next structure level,
  reject if R:R < 1.5
- Fix news engine for all 4 pairs (was hardcoded to USDJPY)
- Add WebSocket monitor: streams live ticks, triggers M15 scan on level breach
  with 5-min cooldown per pair — catches breakouts between candle boundaries
- Add correlation filter (EUR/USD+GBP/USD, XAU/USD+USD/JPY exposure warnings)
2026-05-11 16:33:23 -04:00
Harwinder 7ccd8c4b04 Strip backtesting, rebalance indicator weights toward leading signals
Remove backtest.py, backtest_gpu.py, metrics.py, local CSV data,
Dukascopy loader, and all backtest CLI modes from main.py. Rebalance
engine weights: leading indicators (divergence, RSI, stochastic, BB)
now carry 66% of signal weight vs 34% for lagging (EMA, MACD). Add
ws_twelvedata.py for live tick streaming. Drop pyarrow and tqdm deps.
2026-05-10 23:40:15 -04:00
Harwinder 2e5d66db1b Add session filter + auto-toggle, extend trade journal diagnostics
Session filter restricts trades to a UTC hour window (default 07-16,
London + NY) — opt-in via --session-filter, with --session-start /
--session-end overrides. Wired into both the backtester (cheap hour
check next to the regime gate) and live mode (downgrades signal to
NEUTRAL/F when current UTC hour is outside the window).

Auto-toggle (--auto-session-filter) tracks rolling mean pnl_pct over
the last N closed trades; when it falls below threshold, the session
filter activates, releasing once expectancy recovers. Default lookback
10 trades, threshold -0.3%.

Trade dataclass extended with diagnostic fields captured at entry:
entry_hour, entry_dow, ADX, ATR%, market_regime, volatility_regime,
mtf_agreement. These let --journal-csv exports surface loss-pattern
clusters when a window underperforms.

Validated on 2024 trend / 2023 chop / 2026 recent windows. Auto-toggle
preserves ~50% of trend edge while still cutting ~25% of recent-regime
losses — pure regime filter is still best on cumulative, so both
session flags ship default-off.
2026-05-07 18:38:49 -04:00
Harwinder 28a00b7da0 Compact hierarchical live signal output
Restructure print_signal_report so decision (signal · grade · score ·
confidence · MTF) sits at the top, followed by trade setup, context line,
MTF breakdown, top supporting/opposing reasons in two columns, warnings,
and an optional detail block under --detail. Adds _compact_reason helper
that strips known indicator name prefixes via _INDICATOR_PREFIXES so
reason lines stay short.
2026-05-07 17:53:50 -04:00
Harwinder ce76684a62 Add indicator ablation framework + apply ablation-driven cleanup
Framework:
- TechnicalConfig.disabled_indicators skips named indicators in engine.analyze
- Optimizer supports disable_indicator override for parallel ablation sweeps
- main.py --ablate runs leave-one-out across all 10 directional indicators

Cleanup applied (validated across 2024 trend / 2023 chop / 2026 recent):
- Drop pivot by default (cumulative +9.5% to disable)
- Drop candlestick by default (+3.4% cumulative)
- Regime-gate ichimoku: 1.3x in trending, 0.2x in ranging (was a chop villain)
- Trim RSI weight 1.0 -> 0.3 (cumulative noise across regimes)

Result: +10.60% on 2024 trend (+33.24% -> +43.84%, Sharpe 2.17 -> 2.77),
+15.71% on 2023 chop (-18.27% -> -2.56%). Recent regime still adverse.
2026-05-07 17:33:07 -04:00
Harwinder 045f4248f3 Refocus on USD/JPY and add walk-forward backtester with regime filter
- Strip multi-pair plumbing; drop synthetic-data fallback
- Add BacktestConfig + walk-forward replay through the live signal pipeline
- Add metrics module (PF, Sharpe, Sortino, max DD, trade journal)
- Add parallel grid optimizer with disk parquet cache
- Add H4 ADX regime gate (--regime-filter, --regime-min-adx) — cuts chop
  losses ~84% in 2023 H1 while preserving trend edge
2026-05-07 17:15:41 -04:00
harry e4138ee60a Fixed code 2026-05-04 14:29:32 +00:00
harry 43319d9a98 Add main.py 2026-05-04 04:10:17 +00:00