Parsimonious HJM Modelling for Multiple Yield-Curve Dynamics

TL;DR

This paper introduces a simplified multi-curve interest rate model extending the HJM framework, enabling the derivation of yield-curve dynamics from limited market data, addressing liquidity issues and calibration challenges.

Contribution

It presents a parsimonious multi-curve model based on observed rates, reducing calibration complexity compared to existing multi-curve models.

Findings

Model successfully calibrates to market data

Reduces need for extensive yield curve quotes

Demonstrates effective capture of basis-swap spreads

Abstract

For a long time interest-rate models were built on a single yield curve used both for discounting and forwarding. However, the crisis that has affected financial markets in the last years led market players to revise this assumption and accommodate basis-swap spreads, whose remarkable widening can no longer be neglected. In recent literature we find many proposals of multi-curve interest-rate models, whose calibration would typically require market quotes for all yield curves. At present this is not possible since most of the quotes are missing or extremely illiquid. Thanks to a suitable extension of the HJM framework, we propose a parsimonious model based on observed rates that deduces yield-curve dynamics from a single family of Markov processes. Furthermore, we detail a specification of the model reporting numerical examples of calibration to quoted market data.