About me

Households form long-term relationships with a main bank and concentrate their other financial products within it. I study how these features of household demand shape competition in retail banking. Theory does not sign their effect. Inertia creates a trade-off between investing in new relationships and harvesting existing ones. Demand complementarities across products allow banks to price the relationship product below cost and recover margins on add-on products, or the reverse. Which force dominates is an empirical question. I use monthly data on the bank and product choices of nearly two million UK households over the 2022–23 rate-hiking cycle, a period in which banks also offered cash incentives for switching. I estimate a dynamic model of these choices and use it to measure the value of a customer relationship to a bank and how this value shapes pricing over time and across products. 

In 2025-26, the USD exchange rate risk rose and financial intermediaries’ balance sheets were stressed by arbitraging deviations from UIP. And yet, the quantity of FX hedging rose and the cost of hedging stayed small. This paper proposes an explanation: intermediaries take uncovered FX positions. When the currency is depreciating (appreciating), taking these positions leads them to endogenously increase (reduce) the supply of forwards, which causes the basis to narrow (widen). We find evidence in the trade-level micro data that supports this hypothesis, and show that it can explain the time-series correlations between UIP wedges, CIP wedges, the quantity of hedging, and the change in the exchange rate. By taking FX risk, financial intermediaries attenuate the depreciation and the flow of capital away from the US during USD sell-offs.

This paper develops a theory of how deposit insurance shapes banks’ lending decisions when deposit supply is not perfectly elastic. Expanding deposit insurance coverage can lead to either excessive or insufficient lending, depending on the banks’ marginal source of funding. Deposit insurance acts as an implicit subsidy that lowers the cost of insured funding. When marginal lending is deposit-funded, this subsidy induces risk-shifting and excessive lending; when marginal lending is wholesale-funded, it creates an overhang problem that leads banks to pass up profitable loans. Evidence from U.S. bank- and mortgage-level data is consistent with the model mechanism.

In this paper, we investigate the relationship between the banks’ solvency ratio and their funding costs using a proprietary dataset from Banco de Portugal for 21 Portuguese banks from 2006 to 2020. In light of the discussion on impediments to capital buffer usability by banks, we focus on the importance of market discipline to this relationship. Our results suggest that the relationship between solvency and funding costs is negative and state-dependent, i.e. market participants become more sensitive to changes in solvency during economic downturns. The relationship is stronger for market-based financing sources in comparison to deposits. Finally, we use a breakpoint analysis and find that investors are more likely to penalize the same absolute deterioration in solvency levels when banks are already in a fragile position. Our findings support the hypothesis that fear of market stigma may make banks reticent to use buffers in times of stress.