// Documentation
Docs & User Support
Everything you need to set up, configure, and operate MirrorQuant.
Core Principles
How It Works
MirrorQuant is built on a three-tier quantitative framework: Dynamic Probability Calibration, Microstructure Alpha Capture, and Risk Budget Optimization. In binary prediction markets like Polymarket, edge comes not from predicting the future, but from capturing deviations between market-implied probability and true event probability—while maximizing expected value under risk constraints.
Bayesian Calibration Engine
We aggregate heterogeneous data streams (on-chain capital flows, cross-platform odds differentials, and real-time event signals) to build a prior distribution π₀(θ), then apply sequential Bayesian updating: πₜ(θ) = πₜ₋₁(θ) · ℒ(Eₜ|θ) / ∫πₜ₋₁(θ)·ℒ(Eₜ|θ)dθ. The Hidden Markov Model (HMM) likelihood function filters noise and identifies event evolution phases.
Microstructure Alpha
Our Rust-based low-latency event bus parses real-time order book metrics: VPIN (order flow toxicity estimation), Roll spread decomposition, and Hawkes process modeling for order clustering. When book imbalance Iₜ = (Qᵇ−Qᵃ)/(Qᵇ+Qᵃ) aligns with toxicity thresholds, we execute limit orders in liquidity vacuum zones to capture statistical arbitrage returns.
Risk-Adjusted Sizing
We employ fractional Kelly optimization under volatility targeting: fₜ* = argmax 𝔼ₜ[ln(1+f·rₜ)] − λᵣᵢₛₖ/2 · Varₜ(rₜ). Position sizing respects liquidity depth constraints and correlation penalties via shrinkage covariance matrices.
Parameter Guide
Risk Controls
per_trade_usd: Base position size per round (USDC). | take_profit_pct / stop_loss_pct: Hard stop thresholds as percentage of entry price (10% / 5% default). | max_loss_8h_usd: Rolling loss limit over risk_loss_window_sec (28800s=8h). Auto-pauses trading when exceeded.
Phase A: Early Round Entry
ev_threshold_A: Minimum expected value threshold for Phase A entry (3%). | phase_a_min/max_token_price: Valid token price range for Phase A (0.30-0.70). Filters extreme mispricing. | max_A_trades: Maximum Phase A entries per round.
Phase B: Mid-Round Entry
phase_b_start_sec: Seconds after round open when Phase B activates (240s=4min). | phase_b_max_token_price: Token price ceiling for Phase B entry (0.90). | phase_b_momentum_threshold_usd: Minimum BTC price delta required for entry ($10).
Dynamic Hedging
dynamic_hedge_enabled: Master switch for layered hedge positions. | dyn_hedge_max_entry_price: Maximum token price to enter hedge (0.30). | dyn_hedge_zone_*_retreat: Retreat thresholds triggering different hedge zone sizes (safe/warn/neutral/caution/danger). | dyn_hedge_arb_enabled: Enable arbitrage-style hedge when token price exceeds threshold (0.96).
Smart Stop-Loss
phase_b_smart_sl_enabled: Dual-confirmation stop loss requiring both market thesis invalidation and token price decline. | phase_b_smart_sl_token_confirm_pct: Minimum token price drop to confirm stop (12%). | phase_b_sl_active_window_sec: Time window before round close when stops are active (5s).
Execution Settings
signal_eval_interval_sec: Signal computation frequency (0.3s). | fok_retry_enabled/max: Fill-or-Kill retry for insufficient liquidity. | spread_tolerance: Maximum acceptable spread as fraction (3%).
Price Sources
btc_start_price_source: Reference price source for round-open baseline (external data feed / on-chain oracle / platform API). | binance_ptb_dual_confirm_enabled: Dual-confirmation mode using independent price sources for signal validation.
Installation & Usage
Step 1: Download
Download MirrorQuant TradingFrame Setup.exe (≈1.5GB) from your account dashboard. The package includes the Electron desktop interface and bundled API server—no Python or Node.js installation required.
Step 2: Install
Run the installer and follow the wizard. The application installs to %LocalAppData%MirrorQuant TradingFrame. Desktop and Start Menu shortcuts are created automatically.
Step 3: Wallet Setup
Launch the app and connect your Metamask wallet. The tool requires scoped API keys (read + trade permissions, no withdrawal access) from your Polymarket account. Your private keys never leave your device.
Step 4: Sandbox Testing
Start with the built-in sandbox mode to validate configuration. The simulator replays historical market conditions without placing real orders. Verify your parameter settings before live deployment.
Step 5: Live Trading
Toggle to live mode when ready. The dashboard displays real-time PnL, win rate, and round history. All trades are recorded locally with full audit trails.
Step 6: Parameter Tuning
Edit config/pde_runtime_config.json while the engine runs—changes auto-reload within 5 seconds. Hot-swappable parameters include sizing, thresholds, and risk limits. Restart required only for market selection changes.
Getting Started
What is MirrorQuant?
A quantitative execution engine for prediction markets. We mirror institutional flow patterns and execute against Polymarket's CLOB at sub-second latency.
Account Setup
Create a Polymarket account, generate scoped API keys (read + trade, no withdrawal), and paste them into your MirrorQuant dashboard. We never see your seed phrase.
First Strategy: BTC 5M
The default Alpha-1 strategy targets BTC 5-minute UP/DOWN markets. Default sizing is 250 USDC per trade with a 0.5% slippage cap.
Strategy Reference
Signal Generation
We compute a real-time edge using a Bayesian model fed by BTC tape, order book imbalance, and historical 5-minute outcome distributions.
Position Sizing
Volume-weighted by conviction. Higher edge → larger size, capped at your configured maximum.
Risk Controls
Hard stops on session drawdown, max concurrent positions, and per-trade loss caps. All configurable.
API Reference
Authentication
Bearer-token auth. POST your API key to /v1/sessions to obtain a JWT for the WebSocket feed.
WebSocket Signal Feed
Connect to wss://stream.mirrorquant.io/v1/signals. You'll receive JSON events: { ts, market, side, edge, suggested_size }.
Trade History
GET /v1/trades?since=… returns the last N executed trades with full audit detail.
Support
Discord
Join the community Discord for setup help, strategy discussion, and roadmap previews. Pro users get a dedicated channel.
support@mirrorquant.io · response within 24h on weekdays. Pro tier: 4h SLA.
Status Page
Subscribe to status.mirrorquant.io for execution-node uptime and incident reports.