- Understanding the interest rate risk of deposits requires more than a view of deposit beta – banks need a pulse on deposit weighted average life (WAL) to complete the picture.
- In 2023, high interest rates led to significantly shorter WAL across consumer and commercial segments, including online/direct high-yield savings.
- The value and stability of deposits will likely change by the end of 2024 – data and analytics can help banks manage the transition.
Bank treasury teams focused on interest rate and liquidity risk management, business line teams focused on the competitive environment, and investor/analyst teams focused on bank valuation continue to devote considerable attention to deposit repricing betas, seeing it as the key interest rate risk (IRR) indicator.
In our view, this is an incomplete assessment of IRR and the quality of a bank’s deposit funding. A more well-rounded take would consider deposit weighted average life, or WAL, which is a measure of deposit balance stickiness – i.e., how long an acquired dollar of deposit balances will remain on a bank’s balance sheet.
Changes in WAL have significant implications for IRR measurement, balance sheet management and profitability assessment. Unlike beta, however, this metric cannot be derived directly from public or regulatory financials, and is frustratingly difficult to infer from bank IRR disclosures.
Generally speaking, longer deposit lives translate to more stable (and thus, more valuable) funding for banks. From an IRR perspective, longer lives mean a less liability-sensitive balance sheet, which translates to increased net interest margin expansion when rates rise. Of course, if shortening in deposit life is not analyzed and tracked proactively, it means that institutions may have inaccurate assessments of their IRR profile and deposit profitability.
As in previous rising-rate periods, flow of funds in 2023 accelerated toward higher-earning savings and term-deposit products – this significantly shortened consumer and commercial deposit WAL, according to the Curinos Treasury Analyzer, which provides deposit study analytic benchmarks from an account-level industry dataset covering more than $7 trillion in U.S. deposit balances on a monthly basis.
Traditional retail savings saw the largest drop in WAL at 3.6 years (see Figure 1)1, while direct savings, which represent higher-rate savings products originated by online channel players, also experienced a significant decrease compared to core transaction accounts.
Figure 1. Change In Trailing 12 Month By Segment Observed In 2023
This shortening, and the variation across segments, carries several implications. First, WAL shortening suggests that banks may be less asset-sensitive than expected, implying a less rosy profitability outlook. As rates remain elevated, it will be more costly to replace runoff balances at prevailing market rates.
In addition, differences in the magnitude of shortening observed across segments confirm the general intuition that different deposit segments have different funding value to banks. This is critical knowledge for strategic business planning, tactical deposit pricing and product design initiatives. More simply, not all deposits are created equal from a funding value perspective.
Improved deposit analytics are needed
To manage IRR in the current rate environment, banks must invest in improved deposit analytics that feed asset/liability modeling and balance sheet management, including interest rate hedging. Granular insights delivered in a timely fashion can help treasury teams navigate fast-changing industry dynamics and gather funding from unfamiliar sources, such as new entrants to the high-yield direct/online deposit scene.
The most critical areas of improvements should be:
- Timeliness. Observations from dated deposit studies, “through-the-cycle” observations that over-index to flat-rate environments or observations skewed by COVID-era stimulus effects may be unrepresentative of current behaviors and lead to suboptimal internal rates of return. In our view, these metrics should be monitored quarterly, including at the executive-level balance sheet management committees, rather than the annual (or sometimes less frequent) reviews in place at many institutions.
- Segmentation. Overly aggregated results (such as those that clump together pools of balances from a range of customer types into broad product categories) lead to inaccurate measurement and suboptimal decisions. We believe the most useful deposit studies will segment around customer types, focusing not just across major lines of business, but also on the primacy of customers or clients within each business. A critical example of this is across subsegments within a savings portfolio, where the behavior of rate-based savings acquired through an online or direct-banking channel will contrast with that of traditional savings balances acquired through a branch network.
In an information age where vast troves of data are available in near real time, critical information around customer deposits – the primary source of funding and arguably the primary driver of shareholder value for the industry – should not be months or years old.
The value and stability of deposits, given interest rate and inflationary effects, stand to be markedly different by the end of 2024 from where they stand today. Banks that can understand and react to the trends as they emerge will have a real competitive and risk management edge.
1WAL calculated based on balance decay curve analysis with decay rates observed over a 12-month rolling historical observation window to understand current deposit stickiness (vs. longer-term through the cycle averages) with 10-year cash flow truncation assumed.
