bond-spreads

sujitn's avatarfrom sujitn

Fixed income bond spread calculations with Bloomberg YAS parity. Implements Z-Spread, I-Spread, G-Spread, ASW (Asset Swap Spread), and OAS (Option-Adjusted Spread) for USD/EUR/GBP markets. Use when: (1) Calculating any bond spread measure, (2) Building yield curve bootstrapping, (3) Implementing spread solvers, (4) Working with post-LIBOR curves (SOFR/SONIA/€STR), (5) Bond pricing analytics, (6) Comparing spreads across methodologies, (7) Implementing Bloomberg YAS-equivalent functionality.

⭐2stars🔀0forks📁View on GitHub🕐Updated Jan 7, 2026⬇️2downloads

When & Why to Use This Skill

This Claude skill provides comprehensive fixed income bond spread calculations with Bloomberg YAS parity, supporting Z-Spread, I-Spread, G-Spread, ASW, and OAS across USD, EUR, and GBP markets. It facilitates advanced financial modeling using post-LIBOR curves like SOFR and SONIA, offering a robust framework for yield curve bootstrapping, bond pricing analytics, and spread solver implementation.

Use Cases

  • Financial Analysis: Calculate precise bond spreads (Z-Spread, OAS) to evaluate fixed income investment opportunities and credit risk.
  • Model Development: Build and validate yield curve bootstrapping models for USD, EUR, and GBP markets using modern RFRs like SOFR, SONIA, and €STR.
  • System Integration: Implement Bloomberg-equivalent YAS functionality and spread solvers within custom fintech applications or trading platforms using Rust-based data structures.
  • Market Comparison: Compare various spread methodologies (G-Spread vs. I-Spread) to assess relative value and market liquidity across different currency benchmarks.
namebond-spreads
description"Fixed income bond spread calculations with Bloomberg YAS parity. Implements Z-Spread, I-Spread, G-Spread, ASW (Asset Swap Spread), and OAS (Option-Adjusted Spread) for USD/EUR/GBP markets. Use when: (1) Calculating any bond spread measure, (2) Building yield curve bootstrapping, (3) Implementing spread solvers, (4) Working with post-LIBOR curves (SOFR/SONIA/€STR), (5) Bond pricing analytics, (6) Comparing spreads across methodologies, (7) Implementing Bloomberg YAS-equivalent functionality."

Bond Spread Calculations

Implements five spread methodologies for fixed income analytics with Bloomberg YAS parity and data-source neutral architecture.

Spread Hierarchy (Simplest → Most Complex)

Spread What It Measures Complexity Curve Required
G-Spread Yield over interpolated government bond Low Gov't par curve
I-Spread Yield over interpolated swap rate Low Swap par curve
Z-Spread Constant spread over zero curve matching dirty price Medium Zero/spot curve
ASW Spread in swap market terms Medium Zero + forward curves
OAS Spread after removing embedded option value High Zero + vol surface

Typical ordering: OAS ≤ Z-Spread ≈ ASW < I-Spread < G-Spread for investment-grade bonds.

Quick Reference

Z-Spread (Primary Spread)

Solve for constant Z where:

P_dirty = Σ [CF_i / (1 + (s_i + Z)/2)^(2×t_i)]  // Semi-annual
P_dirty = Σ [CF_i × exp(-(s_i + Z) × t_i)]       // Continuous

Solver: Brent's method (recommended) or Newton-Raphson

  • Tolerance: 1e-8 (~0.001 bp)
  • Bounds: [-0.10, 1.00]
  • Initial guess: YTM - benchmark_YTM

I-Spread / G-Spread

I-Spread = Bond_YTM - Interpolated_Swap_Rate(maturity)
G-Spread = Bond_YTM - Interpolated_Govt_Yield(maturity)

ISDA interpolation: Linear on calendar days between bracketing tenors.

ASW (Par-Par Convention - Bloomberg Default)

ASW = (PV_bond_coupons_at_swap_rates - P_dirty) / PV01_annuity

OAS

Requires Hull-White model with trinomial tree or Monte Carlo. See references/spreads.md.

Currency Conventions (Post-LIBOR)

Currency RFR Day Count Spot Fixed Freq
USD SOFR ACT/360 T+2 Annual
EUR €STR ACT/360 T+2 Annual
GBP SONIA ACT/365F T+0 Annual

Government benchmarks: USD=Treasuries, EUR=Bunds, GBP=Gilts

Implementation Checklist

  1. Curve infrastructure - Zero rates, discount factors, forward rates, par rates with interpolation
  2. Day count functions - ACT/360, ACT/365F, ACT/ACT ICMA, 30/360
  3. Root-finding solver - Brent's method with fallback to bisection
  4. Cash flow generation - Handle settlement, ex-dividend, accrued interest

Detailed References

Rust Data Structures

pub trait YieldCurve {
    fn zero_rate(&self, t: f64) -> f64;
    fn discount_factor(&self, t: f64) -> f64;
    fn forward_rate(&self, t1: f64, t2: f64) -> f64;
    fn par_rate(&self, tenor: f64) -> f64;
}

pub struct BondSpreadResult {
    pub z_spread: Option<f64>,
    pub i_spread: Option<f64>,
    pub g_spread: Option<f64>,
    pub asw_par: Option<f64>,
    pub asw_market: Option<f64>,
    pub oas: Option<f64>,
}

pub enum DayCount { Act360, Act365Fixed, ActActICMA, ActActISDA, Thirty360, ThirtyE360 }
pub enum Compounding { Simple, Annual, SemiAnnual, Quarterly, Continuous }
pub enum InterpolationMethod { Linear, LogLinear, CubicSpline, MonotoneConvex }

Validation Targets

  • Roundtrip: Input swap rates must reprice exactly (< 1e-10)
  • Bloomberg parity: Z-spread within ±0.05 bp, OAS within ±0.5 bp
  • Arbitrage-free: Positive forwards, monotonic discount factors