Presentations - PhD students - EM - second year

29 September 2020
Versione stampabile

9.30 AM, Laboratory 4 - Doctoral School of Social Sciences, via Verdi 26

To book your place, please contact school.socialsciences [at]


Networks: sparse penalized estimation methods and applications in economics
Representing connections among variables of interest using a graph could be incredibly useful for understanding more clearly the relations among them. In addition to a clearer representation, knowing the network topology which links a set of variables could help us to understand and model better how a shock in one or more variables propagates into a system. Penalized estimation methods are used today in statistics to estimate sparse and more interpretable models. One can see these as noise filter techniques able to retrieve the relevant information from data. In the context of network structure estimation, different estimation tools with penalty are used to produce sparse estimated graphs, thus performing a selection in the edge set. The direction of my research is twofold. First I want to develop and test the performance of modifications of the available statistical tools to produce sparse estimates of a network. Possible modifications can be related to different types of penalties, such as elastic-net penalty, or to different joint multivariate distribution, such as the multivariate t or multivariate normal inverse Gaussian distribution. Second I want to apply these methods to study financial and economic networks, eventually using simulations to understand better the diffusion of specific shocks. It can be also useful to understand if and how they evolve over time, in an evolutionary fashion, in response to past shocks. A first possible application is related to the estimation of banks’ network and how important events, like last financial crisis or the more recent pandemic, can change its structure. A second possible application is linked with the estimation of the trade network at regional or national level when fine-grained data, useful to precisely reconstruct it, is not (publicly) available. Being able to retrieve a network structure can help us to better predict the evolution of an economic system. But, thanks to instruments developed in network analysis, one can also study the properties of the specific estimated topology understanding how critical they are for a system.

  • GUIDA Vittorio

Exploration and Exploitation: Behavioural Insights and Individual Decision-Making
The concepts of exploration and exploitation were initially introduced in the field of managerial and organizational studies by James G. March (1991). Following an initial consistency in study approaches - especially in the field of organizational learning - the body of literature on this topic has become increasingly vast, complex, and heterogeneous. However, exploration and exploitation are still of great relevance both scientifically, since there are still many opaque areas in the literature (e.g., research on the antecedents of the switch from exploration to exploitation and vice versa), and strategically.
This research proposal addresses individual decision-making, in line with a relatively new research flow in the exploration and exploitation literature. To such purpose, the main theoretical apparatus which this proposal is based on includes behavioural theory and bounded rationality.
Methodologically, the present proposal relies mainly on laboratory experiments, nevertheless, some additional tools and techniques are adopted (e.g., agent-based modelling and computer simulations).
The scientific contributions this proposal aims to provide are strictly related to the following objects of investigation: i) the possible switch between individual exploration and exploitation and its antecedents (e.g., an update of the aspiration levels); ii) whether social and institutional elements influence individuals explorative and exploitative choices; iii) alternative experimental designs and
methods compatible with the fundamental paradigm of “adaptive rationality theories”.
Finally, the management implications of this proposal are related to opportunities for nudging, redefinition of organizational design, and possible new strategic visions for organizations.

  • GOUDARZI Mostafa

High frequency trading and market quality

As a whole, this research studies high frequency trading (HFT) and its impact on market quality. The first stage of the research aims at the identification of HFT by developing a probabilistic model by machine learning techniques that supports the classification using order book data. Although the advent of HFT is back to the last decade, it attracted considerable attention and criticism in both the scientific and the professional community. However, the lack of a standard model for HFT identification made scholars use a wide variety of proxies for HFT with different types of data, from account-level data to anonymous public data, resulting in inconsistent conclusions. Since this study's probabilistic model is developed using data accessible for academics, it is expected to provide more consistency and reproducibility to further research on HFT.
The second stage focuses on the impact of HFT on market quality. It considers liquidity, volatility, and market efficiency as proxies of market quality. The developed model from the previous stage supports the classification in this stage, and the result contributes to the growing literature of market quality from three abovementioned aspects.
The last stage incorporates macroeconomics news with HFT activities and investigates the impact of their activity on market equilibrium. In fact, it seeks to clarify whether high frequency traders play the role of informed traders and if so, what their effects in a zero-sum game.

  • KORCA Blerita

The Impact of Regulation in Social and Environmental Disclosure and Financial Performance of European Companies

This Ph.D. project aims to investigate the impact of the Directive 2014/95/EU on both the quantity and quality of social and environmental disclosure (SED) and corporate financial performance (CFP) in the European large undertakings.
From a methodological viewpoint, the Ph.D. thesis will utilize a mixed-methods approach. The Directive 2014/95/EU marks an important step of the EU towards greater corporate transparency thus, it is crucial to show what the research has already offered and what yet has to be examined. Therefore, a systematic literature review will be conducted to assess the current state of research concerning this regulation and draw out future research avenues. Second, a case study will be conducted with one of the biggest groups in Italy which have to comply with the regulation. This case study aims to examine how the Directive impacted both the quality and quantity of SED. To this end, content analysis is performed in eight sustainability reports (2013-2020), which cover the period before and after the regulation. To complement the longitudinal analysis, interviews inside the group are conducted and seminars are attended. Finally, the third study adopts a more quantitative approach by considering all listed banks in Europe which have to comply with the regulation. This study will examine the relationship between SED in two regimes (voluntary and mandatory) and CFP by using three different proxies.
The overall contribution of this project is to shed light on the ongoing debate if SED should be mandatory or otherwise. Findings are expected to inform practitioners and policymakers on the impact of regulation in fostering or not the quantity and quality of SED. Finally, the analysis of the relationship SED - CFP will offer insights on the potential benefits of organizations in terms of financial performance if disclosing social and environmental information.