Job Market Paper · 2026
Branch Closures and Spatial Competition in Deposit Markets
Abstract
This paper studies how the decline of physical bank branches reshapes
spatial competition and household access in retail deposit markets. I
combine branch-level deposits, bank balance-sheet information, and
census-tract demographics to estimate a spatial model of deposit demand
and a dynamic model of branch-network contraction. Households choose among
nearby branches and an outside option, while banks decide which branches to
close, trading off deposit franchise value against own-network cannibalization,
post-merger related integration costs and strategic costs
of local network overlap. I document that closures are more likely for small
branches, recently acquired branches, branches near other offices of the
same bank, and branches serving older households. Using individual closure
decisions to estimate supply primitives, I solve counterfactual dynamic
closure games in representative local banking markets at the county level.
The counterfactuals show that ownership consolidation can accelerate
network contraction, but that its effects depend critically on the spatial
configuration and bank affiliation of the branches involved. Targeted access
rules can preserve important coverage margins with substantially fewer
retained offices than a strict no-closure policy.
2026
Outsourcing and competition in the banking sector: The rise of Cloud Service Providers
With Peter Eccles and
Paolo Siciliani
Abstract
Cloud outsourcing may alter competition in banking by allowing smaller
competitors to access scalable digital infrastructure. This paper studies
the effects of banks’ outsourcing agreements with Cloud Service
Providers (CSPs) in the UK banking sector using proprietary bank-provider
contract data. We find that CSP spending is associated with lower operating
costs and higher deposits, with reduced-form effects concentrated among
large institutions. We also find that increases in capital requirements are
associated with higher CSP spending, consistent with large institutions
using CSP adoption to reduce dependence on legacy IT systems, improve
operational efficiency, and strengthen long-term franchise value. We then
estimate a structural model of competition in the UK deposit market to
quantify depositor-demand effects from CSPs. We find that the demand-side
benefits of CSP adoption are substantially larger for small and medium banks
and building societies. We use the model to conduct two counterfactual
analyses. First, we simulate a scenario in which cloud outsourcing was
restricted prior to its widespread adoption. The counterfactual implies
higher market concentration, lower market shares for smaller institutions,
and lower depositor welfare. Second, we analyse a reduction in capital
requirements. While lower capital requirements directly increase welfare
through funding-cost effects, they also reduce incentives to invest in CSP
adoption, offsetting roughly 32% of the direct welfare gain. Our findings
suggest that cloud outsourcing has partly reduced technological barriers to
competition in banking markets.
2026
Countercyclical Capital and Reserve Requirements as Macroprudential Stabilizers
With Diego Bohorquez
Abstract
This paper studies the interaction between reserve requirements and
countercyclical capital requirements in stabilizing the business cycle. We
develop a small open-economy DSGE model with nominal and financial frictions
and a parsimonious banking sector. The model captures the distinct
transmission channels of these instruments through interest-rate spreads
and banks’ balance-sheet composition, and allows us to trace their
effects on credit conditions, output, and inflation. We characterize optimal
simple rules under alternative central-bank objectives. Countercyclical
adjustments to both instruments help stabilize the economy, and become
especially valuable when financial stability is an explicit policy goal.
However, unlike capital requirements, tighter reserve requirements tend to
raise inflation and have ambiguous effects on output, potentially conflicting
with traditional monetary policy goals. A shock-specific decomposition shows
that capital requirements are generally stronger at leaning directly against
credit fluctuations, but that the preferred division of tasks between the two
instruments depends on the shocks hitting the economy.
