My current research explores the intersection of macroeconomics, financial market imperfections and technological change,
with a particular focus on how idiosyncratic and systemic shocks propagate across production networks and credit markets.
Fernández-Villaverde et al. (2024) propose a dynamic hierarchical factor model to measure geoeconomic fragmentation as a latent construct designed to capture its inherently multifaceted nature. This paper examines the robustness of that framework through four complementary exercises. First, we offer a reconstructive replication of the model and assess the sensitivity of the resulting estimates to alternative state-space representations and identification choices, with particular attention to issues of replicability and scalability. Second, we examine the framework's capacity to accommodate new sources of information by replacing an energy-related uncertainty indicator with a market-based measure constructed from the sensitivity of defense-sector equity returns to geopolitical news. Third, we investigate the role of taxonomy by reassigning selected indicators across the four domains underlying the fragmentation index. Fourth, we examine the behavior of the estimated factors across a set of historically grounded regimes spanning segmented integration, hyperglobalization, slowbalization, and strategic fragmentation. Across all exercises, the aggregate fragmentation factor displays a remarkably high degree of stability. By contrast, the interpretation of domain-specific factors proves considerably more sensitive to modeling choices, proxy selection, and the classification of indicators. The evidence suggests that aggregate measures of geoeconomic fragmentation are robust to a wide range of reasonable specifications, whereas their decomposition across channels remains closely tied to the informational structure through which the latent construct is identified.
We develop an agent-based model in which the architecture of credit intermediation — the split between regulated and shadow banking — emerges endogenously from regime-switching decisions by heterogeneous banks. Intermediaries are periodically allowed to adjust their business model by choosing between a regulated mode subject to capital requirements and Cournot oligopoly pricing, and a shadow mode funded via asset-backed securities with higher lending intensity. Three channels drive the dynamics: i) a Cournot mark-up channel that makes regulated banking more profitable as competitors migrate to shadow; ii) a balance-sheet contagion channel through which shadow stress spills over to regulated funding costs; iii) a procyclical lending channel enabling co-expansion of both credit volumes during booms. Simulations show that tighter capital requirements induce persistent migration toward shadow intermediation, with wavelet analysis revealing an endogenous financial cycle at lower frequencies. We document co-expansion of regulated and shadow credit during booms, and a markup effect whereby reduced competition raises the profitability of remaining regulated banks, partially offsetting the regulatory burden that triggered the migration. Monte Carlo experiments confirm robustness across regulatory parameterizations. These results suggest that microprudential capital requirements alone cannot control systemic leverage when the regulatory perimeter is permeable.
Deposit reallocation, liquidity and systemic risk in banking networks (with Lucio Gobbi and Alessandro Sulas) Slides here
We study how deposit runs on individual banks, and the ensuing reallocation of funds across institutions, interact with interbank market access to shape systemic fragility. We develop an agent-based liquidity framework in which banks hit by a run lose deposits that are endogenously reallocated toward alternative banks according to proximity and capacity, while all institutions can borrow or lend in an interbank funding market subject to liquidity hoarding and payment frictions. Monte Carlo simulations show that such reallocations can generate an endogenous leakage channel when recipient banks reach absorption limits, amplifying system-wide stress. Liquidity hoarding has non-monotonic effects, initially constraining interbank funding and increasing defaults, but potentially stabilizing outcomes at higher levels. Network structure further conditions these dynamics, with core–periphery configurations and endogenous concentration of inflows shaping the timing and synchronization of cascades. Overall, systemic risk reflects the joint evolution of deposit flows and interbank linkages, implying that policies affecting liquidity provision or deposit mobility have regime-dependent effects.
The double-edged sword of banking supervision: costs, constraints, and strategic complementarities (with Thomas Carraro and Marco Gallegati).
This paper develops a model in which strategic complementarities emerge in a banking system constrained by supervisory frictions. The supervisory authority operates under a binding budget constraint that limits its monitoring capacity, while compliance with regulatory oversight imposes costs on banks. Within this setting, macroprudential policy seeks to regulate aggregate credit dynamics by preventing the system from converging toward extreme equilibria - either excessive credit expansion or a credit freeze. We show that achieving this objective requires a two-pronged policy approach: countercyclical capital buffers effectively curb excessive credit growth during booms, while loan-support programs sustain credit provision in downturns. No single instrument is sufficient to ensure financial stability; instead, a combination of regulatory tools is required to address distinct phases of the financial cycle and counteract the amplification mechanisms driving systemic risk.
New Entry or Reactivation? Machine Learning Evidence of Firm Dynamics (with Giacomo Caterini, Matteo Cristofaro and Mattias Martini).
Accurate and timely statistics on business dynamics are essential for analyzing firm turnover, job creation, and sectoral contributions to aggregate activity. A key challenge in this context is distinguishing genuine new entrants from reactivated firms—entities that re-emerge after apparent exit but retain strong continuity with previously closed businesses. Official data sources often conflate these categories, leading to biased estimates of entry and exit rates. This paper proposes a NLP-based methodology to improve the classification of firm transitions by combining high-frequency administrative records with unstructured textual data. We first use supervised learning applied to firms’ self-declared business descriptions to impute missing sectoral classifications and introduce a filtering algorithm to isolate economically informative text. Secondly, we estimate a Random Forest model to predict firm reactivation and distinguish true entrants from regenerated firms. The resulting framework enhances the quality and timeliness of firm-level statistics and offers new tools for studying business demography and its micro- and macroeconomic implications.
I am also working on the following projects:
The structural origins of macroeconomic time scales: From equilibrium propagation to wavelet evidence (with Marco Gallegati)
Technological collateral and the financial limits of creative destruction (with Artur M. Boscaglia)
The Great Divergence as symmetry breaking: Land, credit and human-capital formation
The temporal organization of markets: Technology, institutions, and macroeconomic coordination