Modeling short-term

bank deposits
In a post-Silicon Valley Bank, Signature Bank world

March 2023
When NMD management unravels
Traditional banking business models rely on creating short-term liabilities via deposits and funding, and investing in
long-term assets to make profits.

The Silicon Valley Bank (SVB) was no different. What mattered most was a strong, robust asset and liability
management (ALM) framework that manages the risks emanating from dynamic interest rate environments,
fluctuating deposit volumes, and changing investment profiles.

The collapse of SVB and other regional banks such as Signature highlights how porous risk-management
processes and weak modeling, especially for non-maturity deposits (NMDs), can wreak destruction.

This whitepaper discusses the key elements of interest rate risk in banking book (IRRBB), best practices in ALM,
and regulatory expectations for the banking industry, specifically in the context of non-maturing deposits (NMDs).

NMDs are a key source of funding for banks that must be dealt with at a granular level, from a risk-management
perspective. As NMDs can be withdrawn at any time, banks must carefully manage their liquidity to address
depositors’ demands.

Categorisation, classification, and structure

The primary identification and classification of deposits into ‘wholesale’ and ‘retail’ are important for risk modeling
and monitoring. This is largely because banks typically use different funding strategies for wholesale and retail
deposits. For instance, banks rely heavily on wholesale funding for short-term liquidity and retail deposits for long-
term funding.

Wholesale deposits are typically larger and more volatile, and have a larger impact on a bank's liquidity and funding
costs as they are typically more sensitive (higher deposit beta) to interest rate changes.

Retail deposits are further split into transactional and non-transactional deposits. Based on historical data (typically
more than 10 years), banks can distinguish between stable and non-stable NMDs. A stable NMD is defined as a
portion that remains undrawn, with a higher degree of likelihood of being drawn. Core deposits are the portion of
stable NMDs that are unlikely to reprice even if there are significant changes in interest rates, and thus have lower
deposit beta.

IRRBB measures that capture inherent risk

The Basel Committee for Banking Supervision (BCBS) has provided guidelines to enhance control framework for
managing IRRBB, which is getting increased attention from regional regulators as they revise their respective
prudential standards. In line with BCBS recommendations, The asset-liability management committee (ALCO) in a
bank is required to have a strong oversight on policies, procedures, assumptions, and risk-mitigation plans. The
treasury typically calculates the net interest income (NII) and economic value of equity (EVE).

Net Interest Income (NII) — the difference between interest received from asset and interest paid to liabilities

Economic Value of Equity (EVE) — the difference between present value of all asset cash inflows and the
present value of all liability cash outflows

2
Stress testing interest rate risks is the key and for this, BCBS recommends that banks apply six prescribed interest
rate shock scenarios to measure IRRBB risk:

• Parallel shock up
• Parallel shock down
• Steepener shock (short rates down and long rates up)
• Flattener shock (short rates up and long rates down)
• Short rates shock up
• Short rates shock down
Changes in interest rates impact the present value and timing of future cash flows, thereby affecting the underlying
value of the NMDs and their EVE. Changes in interest rates also impact banks’ earnings by changing interest rate
expenses, thereby affecting the NII of NMDs.

The BCBS has identified three risk sub-types within IRRBB: gap, basis and option

Gap risk arises from the term structure of instruments. It exists because rate resets occur at different tenors
Gap risk for these instruments, and there is a risk to the bank when the rate of interest paid on liabilities increases
before the rate of interest received on the liabilities.

Basis risk arises due to interest rate changes for instruments with similar tenors but priced in different
Basis risk indices. Option risk arises due to the elements embedded within a bank’s assets and liabilities, where the
bank or its customers can alter the timing of cash flows.

Optional risk is a major concern within the NMD portfolio due to the potential impact of depositors exercising
Option risk
their rights to withdraw their investments due to the underlying economic environment.

The BCBS guidelines states that banks should consider the following dimensions that can influence exercising the
behavior options within the NMD portfolio:

Responsiveness of product rates to changes

Current level of interest rates
in market interest rates

Spread between the bank’s offer rate and market rate Competitive situation and competitive forces

Geographic characteristics Demographic characteristics

Although several metrics converge on interest rate risk management between the Federal Reserve (Fed) in the US
and the Prudential Regulatory Authority (PRA) in the UK (since they both draw from BCBS), there are differences in
how the two regulators treat interest rate risk management:

Interest Rate risk treatment US regulatory treatment UK regulatory treatment

Framework Regulatory guidance issue by Office of the EBA and Prudential Regulation Authority
Comptroller of the Currency (OCC) and Federal (PRA) has set out elaborate IRRBB
Deposit Insurance Corporation (FDIC) for banks to framework to report key metrics NII and
assess and manage their exposure to interest rate EVE and governance frameworks to be
risk in the banking book. mandatorily followed.

Reporting requirements Banks should report these metrics as part of the Banks required to report NII and EVE
Consolidated Reports of Condition and Income (Call using (Common Reporting) COREP
Reports) or CCAR submissions. templates

3
Modeling for NMDs
Key inputs for calculation of IRRBB metrics

Inputs Description

Deposit balances The total amount of balances held by the bank at any given point in time

Maturity The time until deposits are paid back to customers

Deposit rates The interest rates paid on deposits to customers

Deposit rate elasticity (DRE) Sensitivity of deposit flows to changes in deposit interest rates

Discount rate The rate used to calculate the present value of future cash flows, including interest payments
on deposits

Customer rate lag The number of periods between a change in the market rate and the corresponding change in
deposit rates.

Market data scenarios The interest rate scenarios under which the IRRBB metrics are to be forecast or simulated.

It is imperative for banks to prepare these inputs at the right level of granularity — in preparation of developing risk-
models. Treasury ALM team in G-SIBs monitor delta net interest income (ΔNII) and delta economic value of equity
(ΔEVE) for NMDs based on a five-stage modeling framework.

Step 1 Step 2 Step 3 Step 4 Step 5

Product Behavioral Calculation of Implementation Use cases
segmentation characteristics and ΔNII and ΔEVE
(based on their overlays
contractual features
and calculation of
DRE)

NMDs are classified based on contractual features i.e., fixed, or floating and the DRE is calculated. The DRE
measures sensitivity of deposit flows to changes in deposit interest rate. If a bank’s DRE is high, then depositors
are more sensitive to changes in interest rates, and they are more likely to shift their funds to other banks or
alternate investments, which will result in a decrease in the bank’s deposit base.

Impact of DRE on NII

The impact of DRE on NII is significant. If a bank’s DRE is high, it must increase deposit rates more than it would
generally like to, to retain customers, which in turn will increase funding costs and lower the NII. Conversely, if
interest rate falls, the bank should lower the deposit rates to reduce funding costs and increase NII.

Competitive forces thus shape a bank’s ability to set funding rates, and hence there is a need to carefully manage
deposit rates and monitor DRE to optimise NII.

4
Indicative impact of increasing DRE on Bank NII, based on CRISIL’s
implementation for a leading bank

Impact of DRE on EVE

Changes in interest rate also have a significant impact on a bank’s EVE. If the DRE is high, even a small change in
interest rate can lead to significant changes in the quantum of deposits held by the bank, which will ultimately affect
the bank’s EVE. In addition to customers’ behavioral factors, the impact on EVE depends on several factors such
as the bank’s balance sheet structure, interest rate volatility, and the duration of deposits.

The next part of the modeling framework requires the creation of a replicating portfolio. In case of NMDs, the
replicating portfolio can be created using interest rate-sensitive assets (such as bonds) and interest rate-sensitive
liabilities (such as loans). As a precursor, the expected deposit balances, the interest rates that will be paid on
deposits, and the expected deposit behavior need to be estimated. The replicating portfolio approach can help
banks manage the interest rate risk associated with NMDs by providing a hedge against changes in interest rates.

Indicative impact of increasing DRE on bank EVE, based on

CRISIL’s implementation for a leading bank

5
Incorporating the impact of behavioral aspects
Deposits behavioral modeling involves predicting depositor behavior, such as deposit run-offs. To build an effective
deposit model, a combination of internal and external factors should be considered, and regression techniques are
typically used to model depositor behavior. Most common inputs and regression techniques used in deposit
behavioral modeling include:

Inputs

• Demographic data: Age, gender, income, education level

• Historical deposit data: Deposit amounts, frequency, duration
• Product data: Type of deposit products used, interest rates, fees
• Economic data: Interest rates, inflation rates, GDP growth
• Competitor data: Interest rates, marketing strategies, product offerings

The choice of modeling technique depends on the specific objectives of the bank's deposits modeling efforts and
data constraints.

Technique Effective when Important considerations

Logistic Most popular technique used in deposit behavioral For this technique to be effective, it is critical that
regression modeling for the objective of predicting the customer data should be accurate, complete, and
likelihood of depositor actions, such as opening a representative of the target population.
new deposit account, making a deposit, and
withdrawing a deposit account.

Time series Most effective for modeling trends and patterns in Requires data from multiple sources with a large
analysis deposit behavior over time; especially useful to sample size to build a comprehensive time series
identify seasonal or cyclical patterns in deposit model and use of statistical techniques to fill in
behavior and to predict future deposit trends. missing data or outliers to account for model bias.

Linear Less popular choice for deposit modeling because It can be particularly useful for identifying potential
regression it cannot capture non-linear effects like time series outliers or influential observations that may have a
models. significant impact on deposit behavior. Or, for
instance, to examine one-to-one relationships
between interest rates and deposit flows.

Cluster analysis Deployed to identify segments of depositors with Depending on the complexity of dataset and size,
similar behaviors. The dataset of each segment of clustering requires sufficient computational resources
customers is utilised to model the behavior of and high-quality data representative of the relevant
depositor actions. features of deposits.

Calculation of ΔNII and ΔEVE

The calculation of ΔNII and ΔEVE can be implemented in third-party ALM tools such as QRM, SunGard, Misys,
and IBM. In our experience, there have been notable challenges while using third-party systems, and most of them
are linked to a less flexible, rigid implementation that leaves little room for accommodating bank-specific nuances.

6
The computations can also be implemented in an in-house risk-calculation engine

ΔNII is used in the following submissions:

• Pillar 2 reporting
• Pillar 3 disclosures, which include NII sensitivity
• Internal and regulatory stress tests
• NII forecasting
ΔEVE is used for several purposes, including:
• Internal and regulatory stress tests
• Pillar 2 reporting
• Pillar 3 disclosures
• Economic capital
• MI and reporting

Notable challenges in implementing ΔNII and ΔEVE modeling frameworks

In our experience of working with several large and regional banks in the US, the UK and the EU, we have come
across themes that either significantly delay or create a weak implementation of modeling frameworks in banks. We
highlight a few of them below:

• Unavailability of key data inputs for modeling — missing granularity, standardisation and pre-processing across
portfolios and legal entities
• Ineffective ALCO — In banks where implementation was weak, ΔNII and ΔEVE were viewed as compliance
metrics, and no decisions were directly linked to them
• Inadequate management attention and bandwidth

Conclusion
Calculating NII and EVE is key to not only efficient risk management, but also for the effective survival of a bank, as
evidenced by the collapse of SVB and other banks. It is time that management and Boards of banks pay adequate
attention and provide sufficient resources to upgrade risk frameworks to implement ΔNII and ΔEVE calculation
frameworks.

The frameworks have been implemented in larger banks, and significant learnings from those can be drawn to
ensure that the upgraded framework is effective and efficient. In addition, without a strong management will to
accept these metrics in business management, any amount of modeling and/or risk-framework implementation will
be inadequate.

7
Enquiries:

US UK/Europe

Alex Paladino Mandy Singh

Head of Business Development, Americas Head of Business Development, Europe
CRISIL Global Research & Risk Solutions CRISIL Global Research & Risk Solutions
alexandra.paladino@crisil.com maninder.singh@crisil.com

Analytical contacts

Jan Larsen Sarang Bhutada

President Director
CRISIL Global Research & Risk Solutions CRISIL Global Research & Risk Solutions

Dinesh Kumar Yatharth Narang

Associate Director Lead Analyst
CRISIL Global Research & Risk Solutions CRISIL Global Research & Risk Solutions

8
 About CRISIL Limited

CRISIL is a leading, agile and innovative global analytics company driven by its mission of making markets function
better.

It is India’s foremost provider of ratings, data, research, analytics and solutions with a strong track record of growth,
culture of innovation, and global footprint.

It has delivered independent opinions, actionable insights, and efficient solutions to over 100,000 customers
through businesses that operate from India, the US, the UK, Argentina, Poland, China, Hong Kong and Singapore.

It is majority owned by S&P Global Inc, a leading provider of transparent and independent ratings, benchmarks,
analytics and data to the capital and commodity markets worldwide.

About Global Research & Risk Solutions

Global Research & Risk Solutions is the world's largest and top-ranked provider of high-end research and analytics
services. We are the world's largest provider of equity and fixed income research support to banks, and the
foremost provider of end-to-end risk and analytics services to trading and risk management functions at world's
leading financial institutions. We provide corporate research and analytics solutions to operations, strategy, and
sales and marketing teams of corporations globally. Coalition provides analytics and business intelligence to 14
leading global investment banks. We operate from 8 research centers in Argentina, China, India and Poland,
working with clients across time zones and languages. Being part of CRISIL enables us to attract and retain top
quality talent. We have over 2,300 employees, 75% of whom hold advanced degrees in finance, accounting and
management. We employ the largest number of CFAs and CFA aspirants in India. We have won top honours at the
World HR Congress on Talent Management and HR Project for the year 2015. We have also won the NASSCOM
Exemplary Talent Practices Award (NExT Practices) for skill development for two years in succession in 2011 and
2012. The award recognises us as a firm that has the vision to proactively invest in its people and get them future-
ready.

We are committed to delivering cutting-edge analysis, opinions, and solutions. This underscores our proposition of
being the best people to work with.

CRISIL Privacy Notice

CRISIL respects your privacy. We may use your contact information, such as your name, address, and email id to
fulfil your request and service your account and to provide you with additional information from CRISIL. For further
information on CRISIL's privacy policy please visit www.crisil.com.

Argentina | China | Hong Kong | India | Poland | Singapore | UK | USA | UAE

CRISIL Limited: CRISIL House, Central Avenue, Hiranandani Business Park, Powai, Mumbai – 400076. India
Phone: + 91 22 3342 3000 | Fax: + 91 22 3342 3001 | www.crisil.com