How COVID-19 and low interest rates force banks to change their Non-Maturing Deposits modeling

How COVID-19 and low interest rates force banks to change their Non-Maturing Deposits modeling

Managing interest rate and liquidity risk on savings and current accounts is a hot topic for banks in 2021. Risk, ALM, and treasury managers have to navigate changing regulatory requirements, changing withdrawal behavior and deposit pricing strategies due to COVID-19, and decreasing market rates.

To identify (best) market practices for addressing these challenges, Zanders has initiated a mini-survey of ten European banks. This article summarizes the main insights gathered from this survey.

Intensified monitoring and expert overlays due to COVID-19

Since the start of the COVID-19 crisis in March 2020, most surveyed banks have been noticing increasing volumes for savings and current accounts. Banks’ risk management and ALM teams responded by intensifying the monitoring of interest rate and liquidity risk in Non-Maturing Deposits (NMD) portfolios. Some banks opted to recalibrate the parameters of their NMD models (earlier than planned), others chose to apply an expert overlay on the replicating portfolio to adjust for the increased liquidity in the NMD portfolios.

The current challenge for ALM and risk managers is to decide whether the additional liquidity in NMD portfolios is of a structural or temporary nature. Considering the additional volume as ’structural’ instead of ’temporary’ would increase NMD durations for most banks, because less volume will be invested in overnight (or one-month) tenors. The increasing NMD duration will in turn force banks to rebalance their interest rate hedges in order to achieve their target duration of equity.

Redevelopment of NMD models to adapt to low interest rate environment

Especially the Dutch banks act fiercely on the decreasing (or even negative) interest rate levels in the financial markets. Like the excess liquidity impact from COVID-19, it caused banks to intensify the monitoring of their interest rate and liquidity risk. In addition, banks recalibrated the parameters of their NMD risk models to account for the changing pricing and withdrawal behavior. Driven by model validation findings, some banks even completely redeveloped their NMD models. Next to including the most recent volume and pricing data, these banks also changed their NMD models as follows:

  • Incorporate non-linear or asymmetric pricing of NMD portfolios, allowing for modeling a different response from the bank towards increasing and decreasing interest rates.
  • Incorporate forward-looking expectations into the final model parameterization.

In the coming years, we expect more banks to intensively redevelop their NMD models, as legacy NMD models are failing backtests due to linear repricing assumptions. Incorporating non-linear pricing dynamics and forward-looking expectations in model parameter calibration will indeed prove useful in solving the current backtesting struggles.

More explicit management of core and non-core volume

Explicitly modeling and hedging the core and non-core volume in NMD portfolios can be motivated by either regulatory requirements or internal visions regarding the differing behavioral maturity between core and non-core portions. Only half of the banks participating in our mini-survey apply a distinction between core and non-core volume in managing interest rate and liquidity risk in their NMD portfolios.

Many banks still struggle with the interpretation and practical implications of core/non-core NMD volume. EBA guidelines describe core volume of savings or current accounts portfolios as the portion of volume that is not expected to flow out at short notice and does not reprice under changing interest rate levels. Due to the lack of specific (disclosure) requirements, many banks are awaiting more specific and technical guidelines or lead examples before implementing a core/non-core framework for their NMD portfolios in their monthly interest rate risk hedging.

Banks that do explicitly separate core and non-core in their interest rate risk hedges, model the non-core volume either overnight or with a very low duration (e.g., one month). When looking into the ratio between the core and non-core volume, the core portion seems to be somewhat higher for current accounts portfolios, as compared to savings portfolios. This is probably due to the more stable pricing of current accounts, even though the volume tends to exhibit more volatile behavior. Banks take a wide variety of approaches for core/non-core measurement, although most are utilizing NMD model metrics such as client rate repricing speeds and outflow rates. How banks convert these metrics into a core/non-core measure depends on their interpretation of the EBA guidelines.

Zanders & Savings modeling

In the past few years, we have supported many European banks in the development, validation and outsourcing of risk models for non-maturing deposits. We offer support through the entire modeling cycle on topics such as:

  • Calibration of models based on historical data and expert judgement
  • Substantiation of the bank’s model methodology and key assumptions
  • Continuous model updates on recent market or portfolio developments

Our in-house developed Savings Modeling Solution services banks digitally on non-maturing deposits risk modeling and management using efficient calibration, state-of-the-art tooling and longstanding subject-matter expertise. You can read more on how we can support you and/or request a demo on our website. For more information, do not hesitate to contact Bas van Oers.