Modelling stock market volatility

What this book actually teaches

Modelling Stock Market Volatility (1996), edited by Peter E. Rossi, is an academic collection assembling key empirical and theoretical contributions to the econometrics of financial volatility, drawn primarily from the Journal of Business and Economic Statistics. The book's purpose is not to teach investing but to document the state of the academic literature on volatility modeling at the point when ARCH (Autoregressive Conditional Heteroskedasticity) and its extensions had become the dominant framework in financial econometrics. It is a scholarly reference, not a practitioner's guide.

The collection's centerpiece is the family of ARCH models introduced by Robert Engle (who would later receive the Nobel Prize in Economics partly for this work) and their generalizations — GARCH, EGARCH, and related specifications. Each model attempts to capture a stylized fact about financial return series that standard linear models fail to accommodate: volatility clustering, where high-volatility periods tend to follow high-volatility periods, and the leverage effect, where negative returns tend to generate larger volatility increases than equivalent positive returns. The papers collected here develop the statistical machinery for estimating, testing, and comparing these models on real financial data.

The empirical contributions include papers on daily stock returns, options pricing implications of stochastic volatility, and the relationship between volatility and trading volume. The methodological contributions cover likelihood estimation, model selection criteria, and the challenge of separating conditional variance from changes in the underlying return distribution. Readers familiar with time series econometrics will find this a useful reference; those without that background will find it inaccessible.

The book's limitations are substantial from a practical standpoint. Published in 1996, it predates the high-frequency data era, the computational tools that made large-scale volatility estimation routine, and the post-2008 integration of systemic risk and liquidity into volatility frameworks. The models it covers have been extended substantially; realized volatility measures, HAR models, and machine learning applications to volatility prediction are all absent. Practitioners in quantitative finance today use this literature as a foundation, not a current reference.

For financial economists and econometricians studying the history of volatility modeling, or graduate students who want primary sources for the ARCH/GARCH literature rather than textbook summaries, this collection has genuine research value. For anyone else — including most investors, financial advisors, or quantitative practitioners seeking current tools — it is a historical document rather than a working reference.

Read next

About Peter E Rossi

Read more from Peter E Rossi and explore the full bibliography on ClearValue Books.