EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3477. Forecasting the worst: is implied volatility forward-looking enough?

Invited abstract in session TB-63: Risk Management and Cryptoassets, stream OR in Banking, Finance and Insurance: New Tools for Risk Management.

Tuesday, 10:30-12:00
Room: S14 (building: 101)

Authors (first author is the speaker)

1. Carlo Confalonieri
Department of Management "Valter Cantino", University of Turin
2. Paola De Vincentiis
Management, University of Torino

Abstract

The paper backtests the accuracy of 1-day Value-at-Risk (VaR), computed with variance-covariance methodology, using various alternative methodologies for estimating the risk factor’s volatility. The aim is to gauge whether using option-implied volatility (IV) produces superior results compared to other time-series based measures. In particular, we want to test if a forward-looking estimate – reflecting the sentiment of market investors – is more capable of capturing the tail behaviour of returns distribution during crisis periods, characterized by volatility spikes. The empirical analysis is performed on the S&P500 Stock Index, using both the VIX Index and VIX-1day Index, recently introduced by the Chicago Board Options Exchange. The crisis analysed are the pandemic period and the months following Russia’s invasion of Ukraine. Our results contradict our initial hypothesis, showing that IV-based VaR tend to be underestimated during market turbulences, both in terms of frequency of realized losses exceeding the threshold and in terms of average magnitude of the excess losses. However, quite a different picture emerges when we use the new VIX1D index, instead of the traditional VIX. The preliminary evidence indicates that the 1-day implied volatility derived from near term options may be effective for estimating VaR and superior to alternative methodologies.

Keywords

Status: accepted

Back to the list of papers