Deposit Calculations on Demand

The FDIC has resolved hundreds of banks since the crisis, in nearly all cases making insured deposits available to the failed bank’s customers by the next business day. Although US depositors have come to rely on the efficiency and seamlessness of this process, the work that goes on behind the scenes to determine exactly which deposits are insured (and for how much) can be very complex.

To facilitate this process, the FDIC proposed a rule last month that would create new recordkeeping requirements for its 36 largest supervised banks. [1] Originally released last April as an Advanced Notice of Proposed Rulemaking (“ANPR”), the proposal is largely consistent with the ANPR and essentially shifts the responsibility of calculating deposit insurance payouts from the FDIC to the banks, in order to help ensure a timely payout to depositors if a bank fails.

The proposed rule requires banks to have complete records on each deposit account, account owner, and beneficiary in order to: (1) accurately classify each account into the FDIC’s account ownership taxonomy, [2] (2) consistently assign a unique identifier to each depositor, [3] and (3) match each depositor to all the accounts in which they have an ownership or beneficiary interest. Based on that information, banks must be able to quantify the total amount of insured and uninsured deposits in all deposit accounts on a daily basis. Currently, most firms have the ability to do this only for selected types of accounts.

Additionally, the proposed rule requires annual attestation of compliance by the Board, a requirement that was not included in the ANPR. Banks also would be subject to onsite testing and reviews, although no more often than annually. A bank can apply for an extension or an exception from certain provisions for particular accounts if it is unable to meet the proposal’s requirements. If granted an exception, a bank would be required to alert affected holders that receipt of deposit insurance proceeds may be delayed in the event of bank failure and to have the ability to put holds on those accounts in case of resolution.

Banks have approximately 30 months to implement changes to meet the proposal’s requirements (assuming roughly six months for the rule to be finalized and a two-year period for banks to come into compliance), which is a short time considering the scope of work that needs to be done. The FDIC estimates the necessary enhancements will cost the affected banks $328 million over the two-year implementation period, although this estimate may be low given the system upgrades that will be required.

This post offers relevant background, our view of the proposal’s key challenges, and our suggestions of what banks can start doing now.

Background

The Federal Deposit Insurance Act requires the FDIC to pay out deposit insurance to eligible depositors “as soon as possible” following a bank failure. [4] Although the law does not prescribe a specific timeframe, the FDIC has traditionally aimed to make payouts within one business day (usually by the Monday after a Friday failure), largely to maintain public confidence in the banking system. The FDIC’s ability to act promptly in this regard depends on its ability to accurately calculate the amount owed to each depositor.

For each insured bank, deposit insurance coverage is provided: (1) per depositor, and (2) per ownership category. In theory, an individual can be insured for $1.5 million at a single bank so long as the funds are distributed such that $250,000 [5] is placed in each of six separate individual ownership categories (see the Appendix of this brief, available here, for a list of individual ownership categories).

To facilitate insurance determination, the FDIC issued requirements in 2008 for larger banks [6] to maintain the technological capability to provide the agency with basic account information (e.g., account type and owner’s name/address) at the close of each business day. In addition, to prevent deposit insurance overpayment, the 2008 rule requires banks to have an automated process for placing holds on account balances that exceed certain thresholds. [7]

The FDIC has for some time been concerned about its ability to accurately calculate deposit insurance payouts in a timely manner following the failure of a large bank. These concerns are in part fueled by the current trend of deposit concentration at the largest banks, as well as the banks’ (and perhaps the FDIC’s) inadequate technological capability to timely process significant volumes of data.

This proposal is the FDIC’s latest attempt to improve the resolvability of financial institutions. [8] Accurate estimates of insured deposits play a central role in the FDIC’s choice of strategy to resolve a failed bank (e.g., liquidation versus sale to other banks), especially with respect to the decision to establish a bridge bank and transfer some or all deposits to it, as noted in resolution planning guidance issued in December 2014. [9]

Key challenges

The core requirement of a daily insured deposit calculation assumes high quality data, a complete and accurate classification system, and automation of end-of-day account balance reconciliation. Compliance costs are likely to be disproportionately borne by institutions with:

• Data and other recordkeeping deficiencies.
• Misalignments relative to the FDIC’s updated guidance on account classification. [10]
• Significant reliance on month-end manual reconciliations.
• The greatest account and customer complexities, including third party-placed deposits.

Determining the total insured deposit amount by each individual depositor in a precise and timely manner relies heavily on the quality and completeness of deposit records. One of the primary causes of missing or incomplete data is the use of different deposit platforms which have not been fully integrated, perhaps as a result of past acquisitions. Even when systems are fully integrated, it can be difficult for banks to keep records complete and accurate given their high volumes of deposit accounts as well as a lack of electronic records for older accounts. Such issues will make it difficult for banks to prepare their data to meet the proposed requirements, especially when the necessary remediation cannot be fully automated (e.g., obtaining missing customer signature cards).

Further, many banks did not design their internal account holder taxonomies with the FDIC categories in mind, and those with significant misalignments will have difficulty complying with the rule. Even small nuances in how an account was documented or structured can create misalignments with the FDIC’s 14 ownership categories, [11] and change the insurance treatment. For example, merely naming a beneficiary to an Individual Single Account will cause that account to be treated as a Revocable Trust.

The challenges of daily insurance calculation increase exponentially with the complexity of deposit product offerings. Consider, for example, customers that have many accounts at one bank, some having multiple owners and/or beneficiaries, or others involving a trust structure. In these cases, banks that lack a single, consistent customer identifier will face a major impediment in aggregating account balances across all databases and systems. Additionally, some banks rely on manual month-end reconciliation of balances across a customer’s different accounts and will need to enhance their systems and processes in order to perform the on-demand daily reconciliation that the proposal requires.

Finally, complications arise from accounts that are channeled into the bank from third party agents, such as trustees or brokers. While the third party agent often maintains a master or omnibus account, the composition of the underlying owners or beneficiaries changes daily and this information is often held by the broker or trustee in its own systems. To determine the insured total for each customer, the information in third party agent accounts must be considered on a fully integrated basis with the bank’s other deposit accounts.

What can banks do now?

The scope of the enhancements required for some banks suggests that the two-year implementation period is not particularly generous. As such, impacted banks should begin to assess existing data, recordkeeping, and end-of-day account balance determination capabilities (especially those requiring manual intervention) against the proposed requirements to better understand the elements that will be most challenging to meet. This assessment should be followed with documented plans to address any identified issues in detail, including an estimate of the amount of deposits affected and total costs. This step will not just help with implementing the requirements—the FDIC is more likely to look more favorably upon an exception or extension request that is supported with a well thought-out and detailed rationale, including a cost-benefit analysis.

Most banks would benefit from prioritizing work that is valuable regardless of whether it is required by the proposed rule. For example, missing or inaccurate data diminishes the value of business-as-usual tools such as customer relationship management systems and improvements in the quality of the underlying customer account data would provide a business value beyond rule compliance. As a result, prioritizing data quality improvement for individual and joint accounts, which are generally the least complex and most voluminous, should be considered.

Large banks will only be able to demonstrate that they have near on-demand capability to identify and quantify insured and uninsured deposits if the process is fully automated. Accordingly, many banks will need to automate manual steps and identify high-level potential architectures for extracting, housing, and transforming account data and required attributes. Like data quality improvement, the efficiency benefits to other operational processes may take some of the sting out of the proposed rule’s costs. Much of this planning should begin as soon as possible due to the significant IT investment needed to design an enhanced infrastructure which meets both recordkeeping and customer service needs.

* * *

The complete publication, including appendix, is available here.