As interest rates rise, deposit costs lag and mixes change. It’s a concept that bankers are schooled in early in their careers – and they re-learned this lesson during the 2022-2023 tightening cycle. But this time it was a bit extreme.
According to Darling Consulting Group’s Deposits360°® analytics, interest bearing non-maturity deposit costs rose 108 bps as the Fed actively tightened from March 2022 through July 2023. However, they went up another 43 bps in the 12-months following the last Fed hike. Additionally, NMDs fell from 86% of total deposits to 73% of total deposits from March 2022 through July 2024.
This affected interest rate risk (IRR) reporting as positions changed rapidly. What was once asset-sensitive turned liability-sensitive. Confidence in ALCO diminished. And therefore, the ability to make confident, risk-prudent decisions from the ALCO report also diminished.
A highly anticipated Fed easing cycle may create its own surprises. Deposits360° analytics suggest that deposit rates may lag as the Fed cuts rates, which may soften the expected margin relief many bankers are hoping for.
So how is a deposit study supposed to capture such dynamic movements and provide meaningful clarity to ALCOs? Here are four strong considerations for your next deposit study:
1. Capture the lag... but not necessarily timing lags
A common “lag” assumption in risk models is to delay the effective start time of beta for three months. Focus on spread lags to the market index as well as their post-cycle impact.
For example, MMDA rates have increased by 58 bps since the last Fed hike in July 2023. The lag effect is real and not captured in most IRR models. Additionally, when rates fall, DCG’s models project that non-maturity deposit rates will decline at a slower pace for the first 100 bps than the next 100 bps. The same concept can be applied to decay rates. Capturing these lags in the model is critical in gaining confidence in the results.
Bonus content: Darling Consulting Group’s webinar What Is Your Deposit Study Missing?
2. Don't fail to connect Beta & Volume
A basic deposit study may run a single factor regression, testing an MMDA rate change vs. a market rate change (e.g., Fed Funds). That beta may then be applied to the up and down rate shocks in the model. This is not technically incorrect, but could be misleading.
Here’s an example. A new ALCO client provided DCG’s analyst team with their most recent study. The study showed MMDAs have experienced a full cycle beta of 15%. This was technically not wrong. MMDAs moved 15% with Fed Funds through this cycle. However, MMDA balances also declined nearly 30% during that time frame. Was the institution accurately capturing sensitivity in this analysis?
Common factors to analyze behavior patterns can be account size, vintage, and pricing spreads relative to peers, while analyzing these changes continuously.
3. Get smarter about forecasting customer behavior
Decay rates accelerated as the spread between deposit costs and the Fed Funds rate widened during this cycle. This highlighted how dynamic average lives behave as rates change. DCG believes that the Federal Reserve Bank of Dallas Oct 2023 research paper on Deposit Convexity [1] should be required reading for risk managers. The authors state, “…we show empirically that the “beta” of deposit rates to market rates increases as market rates rise, causing the duration of deposits to fall.” Advanced modeling is coming.
4. Don't stop with a "one-and-done"
A “deposit study done every two years” works when rates are relatively static. We have been living in anything but a static rate environment. The example below shows the sensitivity change in average lives for a client through the rising rate cycle and how continuous tracking can lead to proactive planning. Looking ahead, decays on savings and checking accounts may be much different a year from now as the Fed starts to ease.
Image source: Darling Consulting Group Deposits360°®
The banking industry is steeped in data analytics to improve performance. We live in an era where advanced analytics can inform risk modeling, and therefore, the ability to make confident, risk-prudent decisions.
It’s time to regain trust in ALCO.
[1] Federal Reserve Bank of Dallas, “Deposit Convexity, Monetary Policy, and Financial Stability,” October 2023. https://www.dallasfed.org/-/media/documents/research/papers/2023/wp2315.pdf
Darling Consulting Group’s Deposits360°® enables financial institutions to clearly see and understand depositor trends to support strategic decision-making.
Deposits360°’s predictive deposit study delivers continuous and more forward-looking analytics to help institutions tell their risk story and develop appropriate balance sheet strategies. With Deposits360°, the data leads and instinct follows.
ABOUT THE AUTHOR
Joe Kennerson is a Managing Director at Darling Consulting Group. In this capacity, he works directly with financial institutions by providing solutions for their asset/liability management process in the areas of interest rate risk, liquidity risk management, ALM modeling, regulatory compliance, and executive-level education. He is a frequent speaker and author and directly advises clients in all aspects of ALM.
Contact Joe Kennerson at jkennerson@darlingconsulting.com or 978-499-8150 to work collectively to transform ALCO into a profit center.
© 2024 Darling Consulting Group, Inc.
