Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

VOLatility Archive for Realized Estimates (VOLARE)

Fabrizio Cipollini, Giulia Cruciani, Giampiero M. Gallo, Alessandra Insana, Edoardo Otranto, Fabio Spagnolo

Abstract

VOLARE (VOLatility Archive for Realized Estimates - https://volare.unime.it) is an open research infrastructure providing standardized realized volatility and covariance measures constructed from ultra-high-frequency financial data. The platform processes tick-level observations across equities, exchange rates, and futures using an asset-specific pipeline that addresses heterogeneous trading calendars, microstructure noise, and timestamp precision. For equities, price series are cleaned using a documented outlier detection procedure and sampled at regular intervals. VOLARE delivers a comprehensive set of realized estimators, including realized variance, range-based measures, bipower variation, semivariances, realized quarticity, realized kernels, and multivariate covariance measures, ensuring methodological consistency and cross-asset comparability. In addition to bulk dataset download, the platform supports interactive visualization and real-time estimation of established volatility models such as HAR and MEM specifications.

VOLatility Archive for Realized Estimates (VOLARE)

Abstract

VOLARE (VOLatility Archive for Realized Estimates - https://volare.unime.it) is an open research infrastructure providing standardized realized volatility and covariance measures constructed from ultra-high-frequency financial data. The platform processes tick-level observations across equities, exchange rates, and futures using an asset-specific pipeline that addresses heterogeneous trading calendars, microstructure noise, and timestamp precision. For equities, price series are cleaned using a documented outlier detection procedure and sampled at regular intervals. VOLARE delivers a comprehensive set of realized estimators, including realized variance, range-based measures, bipower variation, semivariances, realized quarticity, realized kernels, and multivariate covariance measures, ensuring methodological consistency and cross-asset comparability. In addition to bulk dataset download, the platform supports interactive visualization and real-time estimation of established volatility models such as HAR and MEM specifications.
Paper Structure (57 sections, 34 equations, 22 figures, 20 tables)

This paper contains 57 sections, 34 equations, 22 figures, 20 tables.

Table of Contents

  1. Introduction
  2. The Raw Data from Kibot
  3. Price Analysis
  4. Volume Analysis
  5. Some Insights on Odd-Lot Trades
  6. The VOLARE Back-end: Data Preparation
  7. Asset Classes
  8. Stocks
  9. Exchange Rates
  10. Futures
  11. Download Raw Data
  12. Data Cleaning Procedure for High-Frequency Prices
  13. From Tick Data to Regular Intervals
  14. The VOLARE Back-end: Realized Measures
  15. Univariate Series
  16. ...and 42 more sections

Figures (22)

  • Figure 1: Boxplot of absolute percentage errors across MSFT price series. One year of data (2023-01-03 to 2023-12-29).
  • Figure 2: Boxplot of percentage errors across MSFT volume series. One year of data (2023-01-03 to 2023-12-29).
  • Figure 3: Intraday price dynamics for MSFT, with and without odd-lot transactions (volume $<$ 100), displayed for different time windows at various trading times on December 30, 2022.
  • Figure 4: Intraday price dynamics for JNJ, with and without odd-lot transactions (volume $<$ 100), displayed for different time windows at various trading times on November 20, 2024.
  • Figure 5: Percentage difference between original and odd-lot-adjusted price series for MSFT, JNJ, GE, MAT and LOCO over the period 2024-01-02 to 2024-12-31. Both series were sampled at 1, 5, 10, 15, and 20-minute intervals using the last available price.
  • ...and 17 more figures