Abstract
dc:descriptionThis dissertation investigates issues in high-frequency financial econometrics, focusing on the effects of market microstructure noise and price staleness on volatility estimation and jump detection in financial markets. The increasing availability of high-frequency financial data has revealed complex features of price dynamics that pose significant challenges to reliable econometric analysis. In particular, the presence of market microstructure noise and zero returns can substantially affect the performance of conventional volatility estimators and the reliability of statistical procedures for jump detection. The simultaneous presence of these phenomena constitutes the central research problem of this dissertation, which aims to evaluate existing methodologies and develop corrections that improve the reliability of volatility estimation and statistical inference concerning jumps. The research is organized around three main objectives. First, non-parametric jump detection methodologies are evaluated under different simulated conditions, with particular emphasis on the effects of zero returns. Second, the estimation of integrated variance is examined in the simultaneous presence of market microstructure noise and price staleness, with the aim of developing a corrected version of the Multipower Variation estimator. Third, the dissertation investigates whether the use of the corrected variance estimator can improve the performance of established jump detection tests under the same conditions. Chapter 2 presents a systematic review and comparative evaluation of widely used non-parametric jump detection methodologies. Their statistical properties are examined through an extensive Monte Carlo simulation study incorporating different stochastic price processes, alternative discretization schemes, and varying threshold parameter choices, where applicable. Particular attention is devoted to the empirical size and power of the tests, as well as to the effects of zero returns on their ability to identify jumps accurately. Based on the simulation results, the competing methodologies are compared and ranked, highlighting their respective strengths, limitations, and robustness across different data-generating environments. The chapter also provides practical guidance for their implementation. Chapter 3 focuses on the estimation of integrated variance in the simultaneous presence of market microstructure noise and price staleness, the latter manifesting itself through zero returns in the observed return series. After examining the econometric challenges arising from these two sources of distortion and discussing the limitations of existing estimators, a corrected version of the Multipower Variation estimator is proposed to improve the reliability of volatility estimation under these conditions. Its theoretical and finite-sample properties are investigated, while its performance is evaluated through an extensive Monte Carlo simulation study and compared with that of conventional estimators. The results highlight the effectiveness of the proposed correction in addressing the effects of market microstructure noise and price staleness under the conditions considered. Chapter 4 builds on the methodological developments of the preceding chapter and examines their implications for statistical inference concerning jumps. Specifically, two established jump detection tests are revisited, and appropriate modifications are proposed to account for the simultaneous presence of market microstructure noise and price staleness. Through an extensive simulation analysis, the performance of the modified tests is evaluated, and the relative effectiveness of the competing methodologies is reassessed. The results suggest that appropriately addressing these sources of distortion can substantially alter conclusions regarding the frequency of detected jumps and their contribution to total price variation. In particular, under the conditions considered, jumps are detected less frequently, and their estimated contribution to total variation is smaller than that implied by conventional methodologies. Overall, this dissertation contributes to the study of reliable volatility estimation and jump detection in high-frequency financial data by highlighting the importance of jointly accounting for market microstructure noise and price staleness. The findings underscore the need to consider these features when evaluating the performance of econometric methods and drawing conclusions about the continuous and discontinuous components of financial price variation.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Daousis, Emmanouil
- Δαούσης, Εμμανουήλ
Subjects
dc:subject × 15- Χρηματοοικονομική οικονομετρία υψηλής συχνότητας
- Μέτρηση μεταβλητότητας και εκτίμηση ολοκληρωμένης διακύμανσης
- Άλματα τιμών
- Στασιμότητα
- High-frequency financial econometrics
- Volatility measurement and integrated variance estimation
- Market microstructure noise
- Price jumps
- Staleness
- Οικονομικά και Επιχειρήσεις
- Κοινωνικές Επιστήμες
- Οικονομετρία
- Economics and Business
- Social Sciences
- Econometrics
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/62850