Soala Ekine

MIT Department: Economics

Undergraduate Institution: Bryant University

Faculty Mentor: Daron Acemoglu

Research Supervisor: Mert Demirer

Website: LinkedIn



I am an international student from Nigeria that is a rising senior at Bryant University and there, I’m an Applied Economics major with a political science minor. After my degree, I plan to pursue a PhD in development economics. I’m passionate about trying to understand why low income countries are poor and how they can develop. I’m currently fascinated with the value of randomized controlled trials in the field. My interests include reading whatever books I can get a hand on, being with friends and watching a lot of basketball.

2018 Research Abstract

Investigating the link between banking networks and bank health

Soala Ekine1, Mert Demirer2 and Daron Acemoglu3

1Department of Economics, Bryant University

2, 3Department of Economics, Massachusetts Institute of Technology

Ever since the 2008 banking crisis, there has been an increase of interest in understanding the factors of financial crises, one of the key ones being the spillover effects of risk in interconnected banks. However, it is incredibly difficult to analyze current banking networks because of their complicated connections and the banks’ refusal to be forthcoming with internal data. In this paper we analyze the Panic of 1896 using banking data on 187 Massachusetts savings banks from 1892-96. We hypothesize that banks that are more connected can spread their risk and would therefore be safer than banks that aren’t as connected. As a proxy for bank health, we measure the percent growth in closed accounts in savings banks. We measure bank connectedness by: taking account of the weighted growth of connected national banks’ individual deposits; finding an average of the growth of national banks’ deposits tied to each savings bank; and counting the number of national banks each savings bank had deposits in. We run different regression models using these three measures each as an independent explanatory variable. We find that the variables of interest are not statistically significant.