Measurement of Intraday Volatility in Stock Markets

• Applicant: Professorin Dr. Franziska Peter
• Subject Area: Accounting and Finance ; Statistics and Econometrics
• Term: from 2018 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 389577820

• Final Report Year: 2023

Final Report Abstract

Accurate assessment of current and future equity market volatility is of great importance, not only for individual investors, but also for the stability of the financial market system. Today's trading has a speed that requires faster, more flexible risk management that can reflect rapidly changing market situations. This research project aims to fill gaps in the literature and investigate volatility at a high frequency in the intraday context and for individual stocks. Since volatility itself is a latent variable that is only implicitly derived from observable time series such as stock returns, an adequate quantification is usually bound to parametric models. However, the current financial market literature shows that the options-based measure of implied volatility contains information related to future stock market volatility that cannot be derived from historical stock prices. In the context of this project, the extent to which such a measure based on high-frequency option data can also be implemented beyond the index context, i.e. for individual stocks, is examined. The evaluation of a large options data set of US companies shows that an implementation of the options-based measure on the stock-specific intra-day context is indeed possible. However, due to the liquidity and breadth of the strike prices of the traded options, it is only recommended for a group of about 100 underlying stocks. For these, the model-free implied volatility measures can be calculated on an intraday frequency. These measures were next used to empirically analyse the relationship between volatility and return on an intraday frequency. The results confirm a negative relationship between stock returns and volatility, which is stronger for extreme returns. However, empirical evidence for the volatility -eedback or leverage effect documented for lower frequencies is not found. Furthermore, during the project - deviating from the original key questions - a focus developed on the influence of sentiment and attention measures based on proxies derived from data from social media platforms or prevalence in Google searches on key variables of stock trading, especially volatility. Several empirical studies have been conducted to measure these effects. For example, the results of a study of the interaction of volatility and a sentiment and attention measure based on Twitter data show a significant correlation between intraday volatility and information from stock-related tweets. However, from an economic perspective, the effects are negligible and out-of-sample forecasting performance is not significantly improved. From a practical perspective, therefore, this study finds that high-frequency Twitter information is not particularly useful for investors with access to such data for intraday volatility assessment and forecasting, and results from other studies for stock indices and lower frequency do not translate the context of individual stocks and intraday frequencies.

Publications

• The Twitter myth revisited: Intraday investor sentiment, Twitter activity and individual-level stock return volatility. Journal of Banking & Finance, 96, 355–367.
  Behrendt, Simon & Schmidt, Alexander
  
• An encyclopedia for stock markets? Wikipedia searches and stock returns. International Review of Financial Analysis, 72 (2020, 11), 101563.
  Behrendt, Simon; Peter, Franziska J. & Zimmermann, David J.
  
• No Model No Cry? Intraday Model-Free Implied Volatility and the Leverage Effect for Individual Equities. SSRN Electronic Journal (2020).
  Peter, Franziska & Haas, Martin G.
  
• Does online investor attention drive the co-movement of stock-, commodity-, and energy markets? Insights from Google searches. Energy Economics, 99, 105282.
  Prange, Philipp
  
• Nothing but noise? Price discovery across cryptocurrency exchanges. Journal of Financial Markets, 54, 100584.
  Dimpfl, Thomas & Peter, Franziska J.
  
• Structural breaks in Box-Cox transforms of realized volatility: a model selection perspective. Quantitative Finance, 21(11), 1905-1919.
  Behrendt, Simon
  
• What are you searching for? On the equivalence of proxies for online investor attention, Finance Research Letters, 38.
  Behrendt, Simon & Prange, Philipp
  
• Intraday Value at Risk Estimation with Multivariate Intensity Models: An Application to Cryptocurrencies. SSRN Electronic Journal (2023).
  Patino, Mariana & Peter, Franziska