When deposits are debt with random maturity
Deposits of US banks increased by $865 billion just in April 2020 alone. Are these large deposit inflows an opportunity or also a challenge for banks?
In this paper, Bolton et al. treat banks’ deposit base as a persistent and stochastic process subject to both inflow and outflow shocks in a fully dynamic setting. As a result, deposit inflow becomes burdensome as it further reduces the equity capital-to-deposit ratio, making risk management more challenging. In their model, banks adjust deposit flows through deposit rate: they set a higher rate when they are well capitalized and want to attract cheap financing for future risk taking; but when they are undercapitalized, they lower their deposit rate to discourage deposit inflows in order to mitigate unintended expansion of their leverage. The other two important elements of their model are: a lower bound in their deposit rate which becomes binding in a low interest rate environment; and the requirement to sustain a level of equity capital that is above a regulatory minimum in order to continue to operate. Their model yields two important conclusions:
- When the bank’s equity capital is depleted, the marginal value of deposits can be negative. The reason is that, near the (lower) boundary of costly equity issuance, uncontrollable deposit inflows destroy value for the bank’s shareholders by increasing the bank’s leverage and amplifying the likelihood of costly equity issuance. The bank then wants to deleverage and turn away deposits. However, the bank can only go as far as setting the deposit rate at the lower bound.
- Capital regulation that imposes a cap on leverage amplifies the cost of deposit-taking because unexpected deposit inflow drives up leverage, pushing the bank closer to the violation of regulation and costly recapitalization.
Dynamic Banking and the Value of Deposits
Authors: Patrick Bolton, Ye Li, Neng Wang, Jinqiang Yang
From: Columbia University, Ohio State University, Shanghai University of Finance and Economics
When deposits are debt with endogenous maturity
The majority of U.S. banks’ liabilities are deposits as they accounted for 78% of banks’ balance sheets at the end of 2019, according to the FDIC. Time deposits with limited maturities account for 15% of all deposits while the remaining deposits are non-maturing i.e. they have no explicit maturity date and are typically not withdrawn for extended periods. Because these deposits can be withdrawn at any time, their maturity depends on market conditions.
This is the idea that Jermann and Xiang explore in this paper: unlike existing literature that models deposits as one-period debt, their model treats redeemable deposits as long-term debt with endogenous maturity and hence exposed to debt dilution. The intuition behind is that there are frictional costs that prevent deposits from being withdrawn and repriced every period. In such a setting how do banks respond to interest rate shocks and optimal bank leverage policies from a regulator’s perspective?
- Following a rate reduction, deposit withdrawals decline and lengthen the maturity which leads to stronger debt dilution incentives, driving up leverage and default rates;
- After several periods of low interest rates, banks are more levered in the model, and this increases the sensitivity of default rates to interest rate moves;
- A regulator who can commit to future policies can impact banks’ and depositors’ values in a more powerful way than one without such a commitment;
- The main finding of their analysis is that accounting for endogenous deposit dilution is important for modeling the effects of monetary policy.
Dynamic Banking with Non-Maturing Deposits
Authors: Urban Jermann, Haotian Xiang
From: University of Pennsylvania, Peking University
