MCPx (Mathema Calculation Plus) — Professional FICC Derivatives Pricing Platform
Why Choose MCPx? Production-Grade FICC Solutions
Trusted by major financial institutions worldwide, MCP delivers enterprise-level derivatives pricing with unmatched performance and reliability. From trading desks to risk management teams, MCP powers critical financial operations with professional-grade tools and transparent models.
Full FICC coverage
FX Options & Exotics • Interest Rate Swaps & Bonds • Structured Products • GPU-Accelerated Monte Carlo • Volatility Surfaces & Curves
Seamless Integration
Native Excel Add-in with PyXLL • Python API for Advanced Users • C++ Core for Maximum Performance • Real-time Market Data Integration
Transparent & Customizable
White-box Models with Full Documentation • Mathematical Formulas & Implementation Details • Custom Model Support • Parameter Adjustments
Enterprise-Grade Models
Black-Scholes, Heston, Bachelier • Local Volatility & Stochastic Vol • Hull-White, CIR Interest Rate Models • Complex Structured Products
High-Performance Computing
GPU Acceleration for Monte Carlo • Batch Processing for Large Datasets • Real-time Pricing & Risk Calculations • Scalable Architecture
Developer-Friendly
Comprehensive Documentation & Examples • One-Click Installation Scripts • Interactive Excel Templates • Professional Technical Support
Always Up-to-Date
Regular Updates for Market Changes • Latest Product Types & Regulations • Professional Support & Maintenance • Global Market Standards
mcp_marketdata is a standard pack, not a quote dump
Cleaned, organised, conventions pre-defined per type. Top-level JSON keys are MCP object definitions (curves / surfaces / credit …). The China EOD window (~90 calendar days) is the public sample. Point Manager / Store at snapshots/; after MCP conversion you can price and measure risk.
GitHub repo
github.com/MDTSH/mcp_marketdata • one MCP_MARKET_DATA per day • point at snapshots/, not the repo root
The node is the type
SwapCurve / YieldCurve / FXVolSurface / CreditCurve — calendar, day-count, interpolation, units already set
Object, then price
LiveStore / RawMD / JsonReader → MCP objects • same snapshot for pricing and risk • Excel TC31–TC46
Ask if it is usable
Four readiness gates • dirty data can still print a number
pip install mathema-mcp
Python only, no Excel. Windows / Linux, 3.9–3.13.