Research

Working Papers and Ongoing Research

[1] Modeling Interdependent Assets: A Global Perspective

Abstract:

Existing literature acknowledges two key characteristics of asset returns' dynamics: they are interconnected and feature drifting prices. We propose modeling these features jointly using a Global Error Correction Model and demonstrate that this novel methodology systematically improves the fit of buy-and-hold strategies across asset classes. We establish the equivalence between Generalized Impulse Response Functions and the conditional sensitivity of asset returns to unexpected global financial market fluctuations—analogous to the Beta in the CAPM framework. Notably, portfolios with low exposure to systemic market shocks (i.e., low beta) outperform traditional strategies aimed at reducing market risk, such as minimum variance or diversified bond portfolios, particularly over annual holding periods.

Conference presentation:

International Conference in Finance, Accounting, and Banking (Southampton, UK, September 2024)
41st International Conference of the French Finance Association (AFFI) (Dijon, France, May 2025)
29th International Conference on Macroeconomic Analysis and International Finance (Rethymno, Greece, May 2025)
Annual Conferences of the International Association for Applied Econometrics (Turin, Italy, July 2025)
41st meeting of the European Economic Association and the 77th European meeting of the Econometric Society, (August 17-21, Ireland, Dublin)

[2] Wildfires, Collection Capacity, and Loss Given Default

joint with Walter Distaso, Angelo Luisi, and Wolfgang Lefever

Link to UGhent WP

Abstract:

Credit losses are not settled at default, yet most of the literature focuses on default risk. The existing research on loss given default (LGD) focuses on prediction rather than causal relationships. A central challenge to isolating causal effects is that factors affecting LGD may also have influenced selection into default in the first place. We address this challenge by exploiting wildfires occurring during the recovery process of already-defaulted accounts. Combining data on more than three million defaulted unsecured telecommunications and utility consumer debts handled by an Italian debt collection agency with geospatial data on burned areas, we find that wildfires occurring in a debtor’s province decrease the probability of full recovery by an average of 0.34 percentage points, or roughly 2.0% of the average recovery rate among exposed debts. Operational congestion within the collection agency amplifies the adverse effect of wildfire exposure on recoveries, consistent with limited servicing capacity constraining the agency's ability to respond to the shock. At the same time, consistent with prioritization of collection effort where scarce collection resources are redirected toward more valuable claims, the effect is weaker for larger claims and higher-value portfolios. Overall, these findings highlight the importance of collection-side operational constraints in determining how adverse shocks translate into recovery outcomes.

Conference presentation:

30th Annual Conference of the European Association of Environmental and Resource Economists (Bergen, Norway, June 2025)
EFic - Conference in Banking and Corporate Finance (Rimini, Italy, June, 2025)
Conference on Sustainable Banking & Finance (Naples, Italy, July 2025)
GRASFI Annual Conference for Sustainable Finance and Investment (Paris, France, August 2025)
XIX Credit Scoring and Credit Control Conference (Edinburgh, UK, August 2025)

[3] Pricing Interconnected yet Heterogeneous Assets: Evidence from the Art Market

joint with Marie Blum and Marc Joets

Abstract:

Markets for unique, infrequently traded assets, such as art, real estate and private equity, are typically studied through aggregate indices. Such indices average out cross-sectional heterogeneity and conceal interdependence among related assets. We propose a reduced-form asset pricing framework that embeds data-driven measures of asset proximity in a dynamic spatial panel model. Applying it to monthly auction returns for 283 artists over 1990–2022, with proximity estimated from return comovement and from textual information on artists' careers, we find that artist returns co-move far more with a proximity-weighted portfolio of peers than with the aggregate market index. Disaggregation also alters inference on macro-financial exposures. Short-term interest rates, producer price inflation and financial uncertainty are associated with artist-level returns, whereas only financial uncertainty matters for the aggregate index. These sensitivities intensify during periods of stress, and interdependence among artists has weakened since 2015.

Conference presentation: