Unraveling financial risk

Research output: ThesisPhD Thesis

70 Downloads (Pure)


This thesis consists of three self-contained studies on the statistical analysis of
financial time series. All studies concern financial time series of which the risk
must be estimated. The main emphasis is on the econometric methodology to
measure financial risk and, in particular on the development of financial risk
measures which can be unraveled.
The first study introduces a new sampling technique for jumps in stock prices.
Stock prices often react sluggishly to news, producing gradual jumps and jump
delays. By locally rearranging the jumps in time, and changing the time labels
of some our observations, we can synchronize the scattered jumps and better
approximate the true common stock jump component.
The second study introduces a new tool to control for false discoveries and
identify individual signals when the test statistics are correlated and the signals
are sparse. In such situations, the Cauchy combination test aims for a global
statement over a set of null hypotheses by transforming and summing individual
p-values. We unravel the combination test to find out which of the p-values
trigger the global rejection. We also revisit a multiple hypothesis testing
problem in high-frequency financial econometrics for which the test statistics
are constructed from rolling windows.
The third study proposes a new decomposition of portfolio risk measures into
the component’s higher moment contributions. The decomposition is based on
the covariance, coskewness and cokurtosis matrix and uncovers the connection
between the summary statistics of portfolio returns, like asymmetry and fat-
tailedness, and the respective risk contributions.
Original languageEnglish
Awarding Institution
  • Vrije Universiteit Brussel
  • KU Leuven
  • Boudt, Kris, Supervisor
  • Smedts, Kristien, Supervisor, External person
Award date4 Oct 2022
Publication statusPublished - 2022


Dive into the research topics of 'Unraveling financial risk'. Together they form a unique fingerprint.

Cite this