An interest rates cluster analysis

TL;DR

This paper conducts an empirical cluster analysis of 34 weekly interest rate time series over 16 years, revealing a hierarchical structure of 6 main clusters without prior assumptions.

Contribution

It introduces a data-driven clustering approach to identify meaningful groupings in interest rate fluctuations over a long period.

Findings

Six main interest rate clusters identified

Hierarchical organization of interest rate groups

No prior assumptions used in clustering

Abstract

An empirical analysis of interest rates in money and capital markets is performed. We investigate a set of 34 different weekly interest rate time series during a time period of 16 years between 1982 and 1997. Our study is focused on the collective behavior of the stochastic fluctuations of these time-series which is investigated by using a clustering linkage procedure. Without any a priori assumption, we individuate a meaningful separation in 6 main clusters organized in a hierarchical structure.