R package for multi-scale network dynamics and systemic risk analysis. Provides agent-based modeling, transfer entropy networks, and macroprudential policy simulation tools for financial research.
MCPFM provides computational methods for financial network analysis
Multi-agent market simulation with heterogeneous participants including HFT, Market Makers, Institutional Investors, and Regulators.
Multi-scale network analysis using wavelet decomposition to capture information flows across different temporal horizons.
Risk assessment combining network, concentration, volatility, liquidity, and contagion factors into a unified index.
Macroprudential analysis tools for evaluating regulatory interventions and policy effectiveness.
Tools for financial analysis and risk management
Framework enabling heterogeneous financial agents to communicate through trust networks and information sharing protocols.
# Create MCP agents
hft_agent <- createMCPAgent(
"HFT_001", "HFT",
capital_base = 1e6,
risk_tolerance = 0.8
)
# Establish connections
conn <- establishMCPConnection(
hft_agent, mm_agent,
strength = 0.7
)Measure information flows between financial instruments using transfer entropy calculation with parallel processing and statistical significance testing.
Risk measurement combining multiple factors into a unified systemic risk index with real-time monitoring capabilities.
Get started with MCPFM in minutes
Ensure you have R 4.0.0 or higher and devtools package
# Check R version
R.version.string
# Install devtools if needed
install.packages("devtools")Install directly from GitHub repository
# Install MCPFM package
devtools::install_github("avishekb9/MCPFM")Load the package and verify installation
# Load the package
library(MCPFM)
# Check package info
packageVersion("MCPFM")Real-world applications and code examples
Create a realistic financial market with heterogeneous agents including high-frequency traders, market makers, institutional investors, and regulators.
# Initialize agent system
agent_system <- initializeAgentSystem(
n_hft = 20, # High-frequency traders
n_mm = 8, # Market makers
n_ii = 12, # Institutional investors
n_reg = 2 # Regulators
)
# Configure simulation parameters
agent_system$simulation_params <- list(
time_step = 0.1, # 6 seconds
total_time = 1440, # 24 hours
shock_probability = 0.001,
shock_magnitude = 0.05
)
# Run market simulation
simulation_result <- runAgentSimulation(agent_system)
# Analyze results
final_price <- tail(simulation_result$performance_metrics$market_prices, 1)
volatility <- sd(diff(simulation_result$performance_metrics$market_prices))
cat("Final Price:", final_price, "\n")
cat("Market Volatility:", volatility, "\n")Analyze information flows between financial instruments using transfer entropy calculation with statistical significance testing.
# Load market data (returns matrix)
data("market_returns") # Or load your own data
# Calculate transfer entropy network
te_result <- calculateTransferEntropy(
time_series_data = market_returns,
embedding_dimension = 3,
time_delay = 1,
parallel_cores = 4,
significance_level = 0.05
)
# Extract significant connections
significant_network <- te_result$significant_transfer_entropy
# Network metrics
network_density <- mean(significant_network > 0)
max_flow <- max(significant_network)
# Visualize network
te_heatmap <- generateRiskHeatmap(
risk_matrix = significant_network,
title = "Transfer Entropy Network",
color_palette = "Blues"
)
print(te_heatmap)Calculate systemic risk index combining network, concentration, volatility, liquidity, and contagion factors with real-time monitoring.
# Calculate comprehensive systemic risk
systemic_risk <- calculateSystemicRisk(
agent_system = simulation_result,
network_dynamics = network_dynamics,
te_network = te_result,
risk_components = c("network", "concentration",
"volatility", "liquidity", "contagion")
)
# Display risk index
cat("Systemic Risk Index:", systemic_risk$systemic_risk_index, "\n")
cat("Risk Level:", systemic_risk$risk_decomposition$risk_level, "\n")
# Component breakdown
components <- systemic_risk$risk_components
for(comp in names(components)) {
cat(comp, "Risk:", components[[comp]]$index, "\n")
}
# Set up real-time monitoring
risk_monitor <- monitorRiskMetrics(
agent_system = simulation_result,
risk_thresholds = c(
systemic_risk = 0.3,
network_risk = 0.4,
volatility_risk = 0.35
)
)
# Generate detailed report
risk_report <- generateRiskReport(simulation_result)
print(risk_report$sections$executive_summary)Evaluate the effectiveness of regulatory interventions including communication taxes, position limits, and circuit breakers.
# Simulate communication tax policy
comm_tax_result <- simulatePolicy(
agent_system = agent_system,
policy_type = "communication_tax",
policy_parameters = list(rate = 0.001),
simulation_horizon = 1440,
scenarios = 100
)
# Evaluate policy effectiveness
effectiveness <- comm_tax_result$effectiveness_metrics
cat("Risk Reduction:", effectiveness$risk_reduction * 100, "%\n")
cat("Efficiency Cost:", effectiveness$efficiency_cost * 100, "%\n")
# Compare multiple policies
macro_analysis <- analyzeMacroprudential(
agent_system = agent_system,
policy_mix = list(
communication_tax = list(rate = 0.001),
position_limits = list(limit = 0.1),
circuit_breakers = list(threshold = 0.05)
)
)
# Optimize policy parameters
optimal_params <- optimizeRegulation(
agent_system = agent_system,
policy_type = "communication_tax",
parameter_ranges = list(rate = seq(0.0001, 0.01, length.out = 20)),
objective_function = "welfare"
)
cat("Optimal tax rate:", optimal_params$optimal_parameters$rate, "\n")Everything you need to master MCPFM
Complete documentation for all 50+ functions with parameters, examples, and return values.
View Documentation →Step-by-step tutorial covering installation, basic usage, and first analysis.
Read Guide →Comprehensive test suite with realistic financial data and validation results.
View Tests →Academic paper: "Multi-Scale Network Dynamics and Systemic Risk: A Model Context Protocol Approach"
Download Paper →Connect with researchers and practitioners using MCPFM for financial analysis.
Join Discussion →