How to Manage Uninsured Deposit Liquidity Risk

Liquidity risk management for risk practitioners remains one of the most difficult areas in asset-liability management, complicated by limited data on how certain types of depositors respond to bank event risk.

One simmering problem that has been lurking below the surface of bank balance sheets for some time is how to assess the risk of uninsured deposits as part of a robust liquidity risk management process.

Clifford Rossi

Indeed, uninsured depositors pose a knife-edge problem to banks that few fully appreciate and understand in their liquidity risk management processes. During most times, uninsured deposit risk remains relatively dormant. But at the whiff of problems at a bank, these are the first deposits to go – amplified by the speed of information flows via social media. Compound that with an insufficient liquidity buffer and, well, you have the makings of a bank failure.

The 2023 madness that took down several large regional banks was an excellent example of the impact of such low frequency, high-severity events. Those failures abruptly jolted the financial services industry and the regulatory community out of complacency over how, exactly, to treat uninsured deposit risk.

Greater emphasis on the timing of deposit runs and outflows of uninsured deposits in liquidity stress testing, combined with the use of better definitions of high-quality liquid assets, will help banks mitigate the risk of uninsured deposits.

In a minute, we’ll delve into greater detail about the specific steps banks can take to more effectively measure and manage this risk. But first we need to understand the kinds of events that can trigger a bank run, as well as why uninsured deposits present a significant threat.

The GFC vs. 2023: Event Risk and Uninsured Deposits

The more uninsured deposits a bank has, the greater the risk it has from a banking panic. Events that can trigger a bank run include mounting unrealized losses on a bank’s balance sheet, subprime mortgage losses and surprise capital hikes.

Diamond and Dybvig’s seminal article describing the nature of deposit runs and liquidity risk set the stage for understanding the risk of uninsured deposits. A key tenet of their analysis is that depositors, especially uninsured depositors, are incented to run on a bank whenever there is a real or perceived adverse change in the valuation of deposits relative to assets over time. The crushing impact of accelerating interest rates in 2022 on bank security portfolio values is a good example of this tenet.

For liquidity risk managers, one key question is how to treat deposit outflows of uninsured deposits in their liquidity stress tests. The first insight from the 2023 bank failures is that the timing of deposit runs appears to be quite different from other tumultuous periods, such as the 2008 GFC. Two of the big bank failures of that period, WaMu and Wachovia, had an uninsured deposit to total deposit ratio averaging about 33% – far below the nearly 84% average for Silicon Valley Bank, Signature Bank of New York and First Republic Bank.

Another major difference between the GFC and the 2023 failures was the timing of deposit runs. Wachovia and WaMu saw 4.4% and 10.1% of their deposits, respectively, run off in a matter of a couple of weeks. In contrast, First Republic experienced 57% deposit outflows over a two-week period. Even more startling was the fact that one-day outflows for SVB and SBNY were 25% and 20%, respectively.

Three factors appear to distinguish the banks runs of 2023: the high percent of uninsured deposits, an increased concentration of large depositors and the increased presence of social media. Considering these factors, what should banks be doing to better manage their risk from uninsured deposits going forward?

Measuring and Managing Uninsured Deposit Risk

To better assess and mitigate the risk of uninsured deposits, banks should (1) build liquidity stress scenarios that align with uninsured deposit risk; (2) leverage the definitions of high-quality liquid assets from the liquidity coverage ratio (LCR) requirements to develop adequate liquidity buffers; and (3) consider the use of innovative products, such as reciprocal deposits, to reduce incentives for uninsured deposits to run in the first place.

Elaborating on step number one, it is logical to build liquidity stress scenarios that involve multiple time horizons. A scenario that involves a severely adverse event for the bank (as described earlier) should include one-day, two-week, 30-day and one-year horizons. For each period under this scenario, the bank must show a positive net liquidity position.

The second recommendation – to leverage the LCR – can help shape a bank’s liquidity profile. While the LCR applies to only the largest banks, the ratio’s definition of high-quality liquid assets is especially useful, because it can help banks set clear definitions built on well-established practices.

Regarding the third potential step, banks with uninsured deposit exposures may want to consider strategies to mitigate some of the risk of these deposits by selling them to the reciprocal deposit market. Uninsured deposits can effectively be parceled out into segments of less than $250,000 to other participating banks, reducing the uninsured risk to the depositor and their incentive to run in the future.

There are limits, though, on how much reciprocal deposits can be used. For well-capitalized banks, such deposits must be the lesser of 5% or 20% of the bank’s liabilities.

Additional Ideas and Issues

Another recommendation is to align uninsured deposit outflow assumptions in liquidity stress tests to conform with or exceed what was observed from deposit outflows from the 2023 bank runs. Banks with significant exposure to uninsured deposits must have even greater vigilance in building their liquidity risk management infrastructure – and should only take on those deposits after such capabilities are in place.

Banks should also strengthen their understanding of the risk from large depositor concentrations. One way they can achieve this is by imposing realistic risk limits that are closely managed and not just monitored.

Effective interest rate and liquidity risk management also requires the development of analysis built on realistic assessments of a bank’s balance sheet. In this area, non-maturity deposits (NMDs) pose one of the biggest analytical risk management challenges.

Since NMDs have no stated maturity, deriving a deposit decay curve can be problematic and fraught with management overlays – including the use of industry practice truncation points that are really nothing more than fudge factors. This lack of sophistication in measuring NMDs is rather appalling, given the importance of this deposit segment in managing bank risk.

It’s important to keep in mind that all the practices we’ve outlined are fraught with a high degree of subjectivity that requires a strong measure of conservatism, especially when uninsured deposits are involved.

Parting Thoughts

The March 2023 banking crisis put a spotlight on liquidity risk management practices, particularly those used to assess the risk of uninsured deposits. Regardless of what customer relationships exist, uninsured deposits pose significant flight risk to banks.

There are, however, steps that banks with large uninsured deposit exposures can take to better manage that risk. Developing liquidity stress test scenarios over multiple time horizons, using severe outflow assumptions on uninsured deposits, is a good first step. Banks also should aggressively manage large depositor concentrations, leverage LCR definitions of high-quality liquid assets and consider the use of reciprocal deposits.

Clifford Rossi (PhD) is the Director of the Smith Enterprise Risk Consortium at the University of Maryland (UMD) and a Professor-of-the-Practice and Executive-in-Residence at UMD’s Robert H. Smith School of Business. Before joining academia, he spent 25-plus years in the financial sector, as both a C-level risk executive at several top financial institutions and a federal banking regulator. He is the former managing director and CRO of Citigroup’s Consumer Lending Group.