Trading Gates & Pattern Filter - Research Notes

Experiment Overview

Item	Details
Date	2025-12-18
Goal	Fix pattern filter blocking all trades; add dashboard gate visibility
Environment	Python 3.10, Alpaca API, Live Trading
Status	Success

Problem Statement

The live trader was showing "Pattern filter rejected - no high-win-rate pattern match" for all symbols, even when the dashboard showed valid BUY/SELL signals. Investigation revealed two issues:

•Missing predator/prey values: _get_predprey_validation() wasn't returning predator/prey values
•Misaligned thresholds: Pattern filter used absolute thresholds but Lotka-Volterra dynamics decay values to near-zero

Failed Attempts

Attempt	Why It Failed
Lowering absolute thresholds (e.g., 0.01 instead of 0.4)	Thresholds become meaningless when values vary by orders of magnitude
Using raw predator/prey values directly	Values decay to 0.01-0.02, making both patterns fail
Bypassing pattern filter	Defeats purpose of high-win-rate filtering

Solution: Normalized Dominance Ratios

The key insight is that while raw predator/prey values can be any magnitude, their ratio indicates regime dominance:

python

# In filter_by_patterns() - alpaca_trading/signals/pattern_filter.py

# Extract raw values
raw_prey = regime_context.get('prey_strength', 0.5)
raw_predator = regime_context.get('predator_strength', 0.5)

# Normalize to dominance ratios (0-1 range)
total_strength = raw_predator + raw_prey + 1e-9  # Avoid division by zero
prey_strength = raw_prey / total_strength  # Higher = more mean-reverting
predator_strength = raw_predator / total_strength  # Higher = more trending

Example

•Raw values: predator=0.015, prey=0.17
•Normalized: predator_dominance=8%, prey_dominance=92%
•Interpretation: Strong mean-reverting regime (VWAP reversion pattern applies)

Trading Gates Architecture

Orders must pass ALL gates before execution:

Gate	Threshold	Implementation
Confidence	Adaptive (0.50-0.65)	`get_adaptive_entry_threshold()` based on GARCH regime
Pattern Filter	65% min win rate	VWAP reversion (68%) OR momentum continuation (71%)
Crypto Short	Block shorts	`detect_asset_type() == CRYPTO and signal < 0`
Portfolio Limit	<80% exposure	`portfolio_metrics.total_exposure`
Capital Manager	30% safety buffer	`capital_mgr.check_trade_allowed()`
Portfolio Risk	VaR <2%	`portfolio_risk_mgr.check_risk()`

Pattern Filter Requirements

VWAP Reversion (68% win rate):

•prey_dominance >= 0.4 (mean-reverting regime)
•Price 2+ standard deviations from VWAP
•Signal direction matches expected reversion

Momentum Continuation (71% win rate):

•predator_dominance >= 0.6 (trending regime)
•trend_prob_up > 0.7 OR trend_prob_down > 0.7
•Pullback in trend direction
•Signal aligns with trend

Dashboard Gate Display

Added gate status indicators to plot_symbol_signals():

python

# In dashboard.py - alpaca_trading/visualization/dashboard.py

gate_statuses = symbol_signals.get('gate_statuses', [])
for i, sym in enumerate(symbols):
    gate = gate_statuses[i]
    status = gate.get('final_status', 'UNKNOWN')

    if status == 'READY':
        status_color = 'green'
        status_text = 'READY'
    elif status == 'BLOCKED':
        status_color = 'red'
        status_text = gate.get('block_reason', 'blocked')[:12]
    elif status == 'HOLD':
        status_color = 'gray'
        status_text = 'HOLD'

Key Files Modified

File	Change
`alpaca_trading/prediction/multi_tf_predictor.py`	Added predator/prey to `_get_predprey_validation()` return
`alpaca_trading/signals/pattern_filter.py`	Normalized predator/prey to dominance ratios
`alpaca_trading/executor.py`	Changed `qty: int` to `qty: float` for fractional shares
`scripts/live_trader.py`	Added crypto short block, `--dashboard` flag
`scripts/monitor_dashboard.py`	Added `check_trading_gates()` function
`alpaca_trading/visualization/dashboard.py`	Added gate status display

Verification

After fix:

•AAPL: PASS (vwap_reversion) - prey_dominance=92%
•AMD: PASS (vwap_reversion) - prey_dominance=91.9%
•AVGO: FAIL (no_match) - SELL signal in prey-dominant regime (expects BUY)
•BTCUSD: FAIL (crypto_no_short) - Alpaca doesn't support crypto shorts

Key Learnings

•Lotka-Volterra dynamics decay: Both predator and prey values decay to near-zero over time; use ratios, not absolutes
•Crypto limitations: Alpaca doesn't support short selling for crypto; must block SELL signals for crypto entries
•Fractional shares: Executor must use float qty, not int, for proper fractional share support
•Dashboard visibility: Showing gate status helps debug why signals don't result in orders

Related Skills

•position-reconciliation: Broker state sync
•markov-regime-features: Debugging constant Markov features
•drawdown-guardrails-pattern: Drawdown control across systems