ETF Time Series Analysis — Univariate & Multivariate GARCH with VAR Modeling

Overview

This academic project simulates a client-oriented analysis, performing a time series study on three ETFs representing different asset classes and geographic exposures. The objective is to assist a hypothetical client in portfolio diversification by studying return dynamics, modeling volatility using univariate and multivariate GARCH models, and analyzing interdependencies through a Vector Autoregressive (VAR) approach.

ETFs Analyzed

• CW8.PA - Amundi MSCI World UCITS ETF (Global Equities)

• CRP.PA - Amundi EUR Corporate Bond Climate Paris Aligned UCITS ETF (Corporate Bonds, Climate-aligned)

• LGQM.DE - Amundi Pan Africa UCITS ETF (African Equities)

Objectives

• Analyze historical returns and basic statistics of selected ETFs.

• Model conditional volatility with GARCH(1,1) models.

• Explore joint volatility dynamics with multivariate GARCH-BEKK models.

• Analyze interdependencies with VAR models and study cross-market reactions via Impulse Response Functions (IRF).

Methods & Models

Descriptive Analysis

• Annualized returns

• Volatility (standard deviation)

• Skewness and kurtosis

• Autocorrelation functions (ACF)

GARCH(1,1) Modeling

• Estimate conditional volatility per ETF

• Study shock sensitivity (alpha) and volatility persistence (beta)

Multivariate GARCH-BEKK

• Capture joint volatility and time-varying correlations between ETF pairs

VAR Modeling & Impulse Response Functions

• Study interdependencies between ETFs

• Analyze how shocks propagate across markets

Data

• Source: Yahoo Finance

• Frequency: Weekly data

• Period covered: April 2018 — March 2025

Tool

• R (RStudio)

Key Findings

• Global equities (CW8.PA) show persistent but moderate volatility.

• Corporate bonds (CRP.PA) are relatively stable but can experience extreme events.

• African equities (LGQM.DE) exhibit high volatility and strong sensitivity to shocks.

• Diversification benefits reduce significantly during global crises (e.g. COVID-19 2020).

Thank you for taking the time to read and explore this project. I am constantly learning and improving my quantitative finance skills. I warmly welcome any feedback, suggestions, or recommendations that could help me enhance this project and develop further. Feel free to reach out!

Gianni Marchetti

Master 1 student in Finance at Aix-Marseille School of Economics

[email protected]