rustmetrics

rustmetrics is a Rust crate for pricing financial instruments and calculating various risk metrics. This crate aims to provide efficient and optimized calculations for a wide range of financial risk measures.

Features

• Pricing and risk calculation for plain financial instruments:

  • Bonds
  • Cross-Currency Swaps (CRS)
  • Interest Rate Swaps (IRS)
  • Futures
  • Vanilla Options
  • Korea Treasury Bond Futures (KTBF)
  • FX Swaps
  • FX Forwards
  • FX Spots

• Calculation of key risk metrics:

  • Delta
  • Gamma
  • Theta
  • Vega
  • Vega Structure
  • Vega Matrix
  • Rho
  • Rho Structure
  • Dividend Delta
  • Dividend Structure

Design Philosophy

The core design of rustmetrics focuses on optimizing calculations by reducing repetitive and redundant operations. This is achieved through two main components:

1. InstrumentCategory: Categorizes financial instruments based on their characteristics.

2. EngineGenerator: Splits instruments into groups and creates specialized engines for each group. This approach significantly reduces duplicate work, such as performing curve bumps only once per engine.

This design provides a solid foundation for future expansions, particularly for implementing numerical simulations like Finite Difference Methods (FDM) and Monte Carlo simulations, where reusing paths and decomposed matrices can lead to substantial performance improvements.

Current Status

This crate is currently in its first prototype stage. While it can handle basic plain instruments, it is actively under development with plans for expansion and optimization.

Future Plans

• Implement more complex financial instruments
• Add support for numerical simulations (FDM, Monte Carlo)
• Optimize performance for large-scale calculations
• Expand the range of supported risk metrics

Contributing

We welcome contributions to rustmetrics! If you're interested in improving financial risk calculations or optimizing performance for large-scale operations, we'd love to have your input.

License

This project is licensed under either of

• Apache License, Version 2.0, (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
• MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)

at your option.

Contact

For questions, suggestions, or contributions, please contact:

Junbeom Lee: junbeoml22@gmail.com

---

We believe this approach has significant potential and look forward to your contributions in making rustmetrics a powerful tool for financial risk analysis in Rust.

Overview

For standard examples, see ./tests/engine.rs

Crate structure

Module	Description
data	Raw market observations, which are not directly used for calculation. Data is shared by Engine object in multi-thread environment
parameters	Objects generated from data objects for actual calculation
instruments	ex) Futures, FxFutures, FxForward, FxSwap, VanillaOption, IRS, CCS, Bond, KtbFutures
time	Calendars, conventions, handling holiday
pricing_engines	Engine, EngineGenerator, and Pricer

Struct	Description
CalculationConfiguration	All information for pricing: delta bump ratio, gap days for theta calculation, etc
Pricer	Enum containing pricers for each Instrument
Engine	An Engine takes data as Arc objects and creates parameters such as ZeroCurve, DiscreteRatioDividend, etc. The parameters, as Rc<RefCell<..>> objects, are shared only inside the Engine. Then the Engine excutes Pricers repeatedly for calculating risks, e.g., delta, gamma, theta, rho, etc
CalculationResult	price, greeks, cashflows
EngineGenerator	EngineGnerator groups instruments according to InstrumentCategory, then Engines are created for each group of instruments. The purpose of separation is mmainly for compuation performance. This is especially useful for Monte-Carlo simulation (not yet developed) since the most of the computation cost in MC simulation is caused by path generation.

---
title: rustmetrics structure
---
stateDiagram-v2
    EngineGenerator --> InstrumentCategory: instruments
    InstrumentCategory --> Engine1: inst group1
    InstrumentCategory --> Engine2: inst group2
    EngineGenerator --> Engine1: data & config
    EngineGenerator --> Engine2: data & config
    Engine1 --> CalculationResult: results
    Engine2 --> CalculationResult: results
    Engine1 --> Parameters: data for inst group 1
    Parameters --> Pricer1
    Parameters --> Pricer2

Loading