Rising interest rates around the world can be helpful to banks, but they come with challenges—especially after an extended period of low borrowing costs. Aside from the direct impacts of monetary tightening, which may have positive or negative effects, banks face increased uncertainty around customer behaviors, in both their loan books and deposits. Amid heightened regulatory risk, leading institutions are now revisiting the impact of these changes across data, models, and risk management activities.
Fast changes in central bank interest rates can lead to significant shifts in customer behaviors and bank risk exposures. In the loan book, drawing and repayment patterns may change, and there can be significant disruptions in mortgage pipelines. There is a higher risk of deposit decay-rate instability, which can require increased use of decay and hazard models to fix repricing tenors.1 Meanwhile, a more intense competitive environment can cause individual players to see declines in liquidity coverage and gains in deposit beta—the percentage of rate change passed on to customers.
In Europe, many of these challenges are subject to regulatory oversight through the European Banking Authority’s (EBA) new guidelines for interest rate risk in the banking book (IRRBB), published in October 2022. The expanded framework motivates banks to balance risks against key metrics, including net interest income (NII) and economic value of equity (EVE), which represents the net value of assets and liabilities. It also provides rules and recommendations on how to calculate key metrics, for example, with respect to the modeling and composition of balance sheet exposures, yield curve scenarios, behavioral assumptions, and subrisk categories (Exhibit 1). To meet IRRBB obligations, many banks are now abandoning the lens of the past 15 years—predicated on “lower for longer”—and taking action to manage the impacts of higher rates across the business.
IRRBB’s new definition of risk to net interest income
The current test for IRRBB supervisory outliers is focused on changes in banks’ EVE—the difference in the value of assets and liabilities before and after hypothetical rates shocks. Going forward, however, the EBA’s proposed test would equally assess the impact of shocks on NII. Methods for measuring NII risk are evolving. The new supervisory outlier test requires a calculation based on a static balance sheet, but many banks (around one in three) are starting to use the more realistic assumptions and responses to market scenarios represented by a dynamic balance sheet.2 Moreover, rising numbers of banks report that a combination of higher rates and the new methodology are causing them to breach the test limit.
McKinsey’s most recent Treasury Survey also reveals some potentially damaging approaches to measurement of delta NII, with the majority of respondents applying floors on deposit and mortgage rates. This has the effect of producing higher levels of variability when rates move up or down. A minority of banks apply floors on market rates.3
In calculating the impact of interest rate shocks, the EBA guidelines say banks should adopt an expanded definition of risk to NII that includes market value changes in the other comprehensive income (OCI) category (for example, revenues, expenses, gains and losses), a correction position on a bank’s common equity tier-1 (CET1) capital. Just 56 percent of banks currently impose limits on OCI, according to McKinsey’s survey. In addition, banks should consider increases or declines in profit or losses and capital over a longer time horizon. This, in turn, impacts how they should treat and interpret behavioral models.
To understand how best to manage IRRBB exposure in the new regulatory environment, banks need to gauge trade-offs in the relationship between bank EVE and NII. The basic rule is that if a bank’s NII declines, it is less able to retain earnings. In a positive rate environment, an upward shock to interest rates yields a negative delta EVE and a positive delta NII. Therefore, a critical task is to ensure that behavioral models can impart the insights into deposits to minimize NII volatility while remaining delta EVE neutral. In practice, this means optimizing reinvestment and hedging activities. Ideally, then, it makes sense to agree on an optimal delta EVE/delta NII position.
McKinsey research shows that banks are taking a range of approaches to balancing delta EVE against delta NII in various rate scenarios. The differences are usually associated with variables that include currency mix, yield curves, behavioral assumptions, and pricing. However, the basic rule is that a bank’s ability to immunize NII across rate scenarios will be contingent on its ability to manage EVE, as well on as the modeling choices it makes.
A new approach to deposit modeling and hedging
The EBA guidelines provide clarifications and extensions for modeling, first relating to maximum tenors and then to the scope of relevant deposits, which have been expanded to include operational deposits by financial institutions.
Best practices for deposits modeling and hedging include the following:
- Customer segmentation in line with regulatory classifications. Balances should be assigned to distinct segments. Behavioral and regulatory features can be used for segmentation, complemented by advanced analytics and expert judgment.
- Core balance modeling. Banks must determine their long-term stable balances, taking into consideration migration between current accounts, term deposits, and savings deposits.
- Deposit volume modeling. Efforts should be made to measure the evolution of deposit volumes. Industry best practice is to use the age-period-cohort model, taking into account the survival rate and expected volume.
- Deposit beta. This is defined as the sensitivity of client rates to changes in market rates or the pass-through rate. There is a trend toward regime-based deposit betas to better capture the variability of market rates being passed through to customers in different interest rate regimes. Calculation of deposit beta should inform hedging strategy.
- Hedging strategy. Risk profile of modeled liabilities can be covered by different hedging instruments. The hedging strategy can focus on economic value or net interest margin, or it can target the optimization of the risk-return profile. An increasing number of institutions are using stochastic models to test hedge ratios in the presence of convexity and optionality.
An increasingly common approach is to apply advanced analytics to the modeling task, for example, by using machine learning to estimate classification probabilities and predict allocations as rates and regulatory treatments change. A random forest model, for example, creates multiple binominal regression-based decision trees and simultaneously selects variables. Often the analytics will point to previously unconsidered drivers, leading to higher-than-expected prediction accuracy.
The current rate environment also requires heightened attention to modeling and management of mortgages and other term loans. Acceptance rates may become more volatile due to changing prepayment behavior, as can loan life spans.
Best practices on quantification, hedging, and pricing of prepayment risks are evolving. They include:
- Customer segmentation. Banks can divide the mortgage portfolio into customer segments through analysis of behavioral features.
- Prepayment behavior. Banks should quantify “constant” prepayments and prepayments subject to specific criteria, such as interest rate level, prepayment penalties, age of mortgage, and additional borrower background. They should adjust expectations to reflect likely shorter tenors.
- Interest rate scenarios. Banks should model a range of scenarios and simulate potential prepayment behavior for each scenario.
- Hedging ratios and strategy. Banks should evaluate the value of mortgages under various rate scenarios and derive sensitivities to economic value and P&L. They should select hedging instruments reflecting fair value and P&L changes.
- Pricing. Institutions should adjust pricing based on analysis of maturities and prepayment behaviors.
In a more volatile rate environment, pipeline risk increases, with acceptance rates tending to be less predictable as prices move between first lock and full drawing. In fixed-rate mortgages, meanwhile, prepayment rights can lead to significant reductions in repricing tenors. At the time of writing, average repricing tenors are well below the regulatory cap. Meanwhile, just a quarter of banks hedge pipeline risk, our most recent Treasury Survey shows.
There are related uncertainties around the quantum and timing of drawdowns. Again, some banks are tackling the challenge with advanced analytics, which they apply at each step in the process, from the initial lock, through client acceptance, and at full approval. Early indications are that use of supervised and unsupervised learning provides powerful insights into acceptance rates, contingent on the banks’ ability to marshal sufficient data across financial, behavioral, and macroeconomic dimensions. The analytics can also provide vital insights into potential hedging ratios and hedge timing.
Focusing on credit spread risk
The new IRRBB guidelines expand the perimeter for credit spread risk in the banking book (CSRBB) and set higher expectations for bank governance with respect to credit decisions. CSRRB is defined as a combination of two elements: changes in “market credit spread” and changes in “market liquidity spread,” representing the liquidity premium that sparks market appetite for investment and creates willing buyers and sellers.
There is still uncertainty regarding the scope of CSRBB. However, the guideline asks banks to consider all instruments that may be subject to credit spread risk, including off-balance-sheet items such as loan commitments. Assets at fair value should always be included, while changes in the institution's own funding rate cannot be used to offset credit risk. Indeed, institutions should not exclude any instrument in the banking book from the perimeter of CSRBB ex ante, including assets, liabilities, derivatives, and other off-balance-sheet items. Potential exclusion of instruments should only be done in the absence of sensitivity to credit spread risk and should be appropriately documented and justified (Exhibit 2).
It is not always the case that credit spread volatility is directly correlated to credit rating, and higher-rated companies are often more volatile, McKinsey research shows. A key but sometimes ignored driver is debt tenor. To offset the risks associated with the above approaches, the EBA has introduced an idiosyncratic component to measures of credit spread risk in the banking book.
It is not always the case that credit spread volatility is directly correlated to credit rating, and higher-rated companies are often more volatile.
With banks starting to implement the CSRBB alongside IRRBB rules, some have adopted strategic change programs, allowing them to simulate CSRBB for the entire balance sheet, including issuances. Consequently, they can measure the full impact of changes in market liquidity and credit spread for both assets and liabilities. This helps them reflect a more dynamic view of the impact of changes to funding spreads in internal risk management frameworks and make economic capital calculations that go beyond the regulatory interpretation of the EBA guideline. Currently, just 28 percent of banks measure the risk of variation in their funding rates over time, McKinsey’s Treasury Survey shows.
Balancing the framework with six priorities
The EBA’s new standards for managing IRRBB are designed to help banks navigate the impacts of shifting rate environments on securities portfolios, pensions, and fair-value accounting. Meanwhile, the new supervisory outlier test threatens to capture many more banks than the existing method.
To tailor the operating model to the demands of IRRBB, we recommend an approach focused on governance, organization, and processes. As a first step, many banks review committee structures, and terms of reference, benchmarked against peers. A common action is to review organizational capabilities (size, skills, mandates) and responsibilities. For processes, some leading banks apply a twin strategic risk management and operational risk management lens, taking into account escalation processes and remediation playbooks. Another common strategy is to apply dedicated KPIs to process efficiency, again ensuring alignment with peer groups.
Once a baseline is established, banks should seek out ways to manage changes in NII, including using advanced analytics and potentially switching to fair-value accounting for securities. It is increasingly common to fully integrate credit spreads into the steering process, with the focus on fair-value securities, while refining behavioral models to reflect higher levels of rate elasticity. It will also be imperative for banks to measure and manage operational deposits. All of this will best be achieved under a holistic governance framework that balances delta NII and EVE through six key priorities:
- Data integrity: ensuring completeness of data, exhaustive and correct data attribution, reliable data transfer, and central storage
- Behavioral models: employing a full suite of behavioral models for deposits, mortgages, and committed credit lines to capture changes to client behavior in different rate scenarios
- Supervisory outlier test: allowing for rapid and frequent calculation of supervisory outlier testing for economic value and net interest income, including sensitivity and scenario analyses
- Business and hedging strategy: reviewing repricing profile of assets, liabilities, and off-balance-sheet position, aligning average tenors, as well as considering application of derivatives and hedge accounting
- Credit spread risk: allowing for calculation CSRBB for a wide range of balance sheet items, which can also be leverage for a broader management of spread risk, including funding-spread risk
- Reporting: establishing a flexible and dynamic reporting framework that is easily accessible for multiple users and allows for drill downs
By paying close attention to management of these six areas, leading banks have shown that they can more accurately gauge the impact of rising interest rates and credit spread risk across key business lines, meet regulatory expectations, and create the impetus for competitive advantage.