Portfolio risk analysis, anomaly detection, market regime classification, and regulatory compliance
High-performance financial risk analysis combining sparse inference for market signal processing with GNN for transaction network analysis. Real-time anomaly detection, portfolio risk scoring, and automated compliance reporting for SEC, FINRA, and Basel III.
Provide institutional-grade financial risk analytics: calculate VaR/CVaR, detect fraud and AML violations, classify market regimes, verify regulatory compliance, and run stress tests -- all with the explainability required by financial regulators.
Calculate comprehensive portfolio risk metrics including VaR, CVaR, Sharpe ratio, Sortino ratio, and max drawdown.
mcp.invoke('finance/portfolio-risk', {
holdings: [
{ symbol: 'SPY', quantity: 1000, assetClass: 'equity' },
{ symbol: 'BND', quantity: 500, assetClass: 'bond' },
{ symbol: 'GLD', quantity: 200, assetClass: 'commodity' }
],
riskMetrics: ['var', 'cvar', 'sharpe', 'max_drawdown'],
confidenceLevel: 0.99,
horizon: '1w'
});
Detect anomalies in financial transactions using GNN and sparse inference. Supports fraud, AML, and market manipulation contexts.
mcp.invoke('finance/anomaly-detect', {
transactions: [{ id: 'tx-12345', amount: 250000, timestamp: '...', parties: ['CORP-A', 'CORP-B'] }],
sensitivity: 0.9,
context: 'aml' // fraud, aml, market_manipulation, all
});
Classify current market regime (bull, bear, sideways, high_vol, crisis, recovery) with transition probability matrices.
mcp.invoke('finance/market-regime', {
marketData: { prices: [...], volumes: [...], volatility: [...] },
lookbackPeriod: 252,
regimeTypes: ['bull', 'bear', 'high_vol', 'crisis']
});
Automated regulatory compliance verification against Basel III, MiFID II, Dodd-Frank, AML, and KYC.
mcp.invoke('finance/compliance-check', {
entity: 'BANK-001',
regulations: ['basel3', 'dodd_frank'],
scope: 'capital',
asOfDate: '2024-01-15'
});
Run historical and hypothetical stress scenarios on portfolios.
mcp.invoke('finance/stress-test', {
portfolio: { holdings: [...] },
scenarios: [
{ name: '2008 Crisis', type: 'historical', shocks: {} },
{ name: 'Rate +300bp', type: 'hypothetical', shocks: { rateShock: 0.03 } }
],
metrics: ['pnl', 'var_change', 'margin_call_risk']
});
| Role | Allowed Tools |
|---|---|
| TRADER | portfolio-risk, market-regime |
| RISK_MANAGER | portfolio-risk, anomaly-detect, stress-test, market-regime |
| COMPLIANCE_OFFICER | compliance-check, anomaly-detect |
| AUDITOR | compliance-check (read-only with full audit access) |
| QUANT | portfolio-risk, market-regime, stress-test |
ruvector-economy-wasm) -- Token economics and market microstructure modeling, risk metric calculationruvector-sparse-inference-wasm) -- Efficient sparse financial feature processing, anomaly detection| Metric | Target |
|---|---|
| Portfolio VaR (10K positions) | <100ms |
| Transaction anomaly scoring | <5ms per transaction |
| Market regime classification | <50ms for 1-year history |
| Compliance check | <1s for full entity scan |