Grant detail

GAUK no. 910680: Asset pricing and portfolio selection in frequency domain

Principal investigator: Mgr. Martin Hronec
Collaborators: doc. PhDr. Jozef Baruník Ph.D.
Mgr. Josef Kurka
Description: Factor investing is based on the identification of risk factors, and proper allocation of risk between them in the portfolio. We argue that both, asset pricing and risk parity models are connected and highly influenced by asymmetric distribution of returns. Hence, higher moments and co-moments of return distribution should serve as proper risk factors. Moreover, heterogeneous preferences of economic actors also have a significant impact on variables entering both asset pricing and portfolio selection. However, most current models working with over-the-frequencies aggregated information do not reflect this heterogeneity. We plan to contribute to the current state of art by proposing models incorporating this heterogeneity. We will develop a theoretical framework for frequency-specific realized measures of higher moments of distribution and then explicitly include them into risk parity portfolio selection methods. We will build on the research of Baruník & Vácha (2015, forthcoming 2018) who formalized frequency domain counterparts of the first and second moments, and Chaudhuri and Lo (2016) who present frequency domain counterparts of number of other risk measures used in portfolio selection models.
We believe that introducing frequency domain counterparts of higher moments and co-moments of distribution can bring very interesting insights to the explanation of cross-section of returns, thus factor investing, as well as risk allocation. To test the usefulness of theoretical framework, we plan to undertake number of empirical tests on financial markets data.
Participation:
Work in grant:
Web link:
Finance:
End date: 2020
Publications:

Does It Pay to Follow Anomalies Research? Machine Learning Approach with International Evidence

Conferences:

Partners

Deloitte

Sponsors

CRIF
McKinsey
Patria Finance