Abstract: We examine whether election periods are associated with increased systemic risk. Our analysis includes a global sample of banks from 22 advanced economies from 2000 to 2023, covering a total of 147 national elections. The findings indicate that systemic risk increases during election and post-election periods, while it is lower in the pre-election period in the case of end-of-term elections. More specifically, the year in which elections occur is associated with a 3.74% higher systemic risk compared to the overall average. The results can be attributed to the suppression of negative information and expansionary fiscal policies in the period before elections. Notably, the impact is more pronounced for snap elections and when the incumbent government was not re-elected. In addition, we find that macroprudential policies, strong economic growth and trust in the current government and banks’ financial health can partially mitigate the impact of elections on systemic risk. Finally, to alleviate endogeneity concerns, we employ two instrumental variables, namely, term times and an election uncertainty index based on Google Trends, in a 2SLS model and the results hold and confirm our previous findings, further validating the robustness of our analysis

Abstract:  We study how bank branch closures affect local economic activity, using satellite-based nighttime lights (NTL) as a high-resolution proxy for changes in economic vibrancy at the census-tract level. Our sample covers over 11,000 US tracts in the period of 1994-2020. To isolate plausibly exogenous variation in branch closures, we exploit within-county, tract-level exposure to post-merger branch consolidation, that is, the closure of redundant branches arising from greater network overlap rather than from local economic conditions. The resulting IV estimates show that branch closures depress local economic activity, as proxied by changes in reduced nighttime lights intensity. The results reveal significant heterogeneity as they are most pronounced in communities with older populations, lower income, lower density, and lower Black population share, as well as in high-mobility counties. A spillover analysis further indicates that the effects of branch closures are highly localized, with significant impacts confined to adjacent tracts and dissipating rapidly over distance.

Abstract:  This paper examines whether standard measures of systemic risk are prone to survivorship bias. Using data on more than 2,500 institutions across 34 countries from 1995 to 2025, we find that relying only on surviving firms overstates systemic risk by around 10 basis points or 5.19% at the country level. The bias can be decomposed into two complementary effects. A firm-level weighting effect (WE), unique to systemic risk estimation, arises because institutional exits reassign system-index weights to surviving firms, mechanically inflating their measured contributions. A country-level omitted-firms effect (OE) occurs because exiting institutions were, on average, less systemically important than survivors, reflecting sample selection bias from excluding systemic events and the fact that many exiting institutions were “too small to save.” Overall, survivorship bias systematically inflates the perceived importance of long-lived institutions while understating risks from newly established firms or those that exited early. The magnitude of the two effects varies across countries, sectors, and time periods. Survivor-only datasets can therefore distort stress tests, cross-country comparisons, and the identification of systemically important institutions.