Agent-based simulation offers a way forward for banks seeking to get an holistic perspective on the interest rate risk on their books. The Basel Commission on Banking Standards (BCBS) has placed increased emphasis on asking banks to study how movements in interest rates affect banks’ profits and capital. This has proved challenging for even the largest banks.
Broadly, we can think of Interest Rate Risk in the Banking Book (IRRBB) as coming in three parts:
- Gap risk — which arises from changes in the term structure of interest rate changes
- Basis risk — arising from similar instruments being priced off of different interest rate indices.
- Option risk — due to optionality embedded in assets or liabilites. Some of this risk will be automatic, and some of it will be a function of customer behaviour.
To properly study interest rate risk, a bank needs to be able to assess the cash flows on both its assets and its liabilities, and to be able to understand how these will move as the yield curve changes. The agent-based simulation approach is an innovative solution for banks seeking to get a better handle on the interest rate risk they run.
Using this, banks can model each financial instrument as an agent, and aggregate their cash flows from the bottom-up. Changes in rates will have heterogeneous impacts on different instruments, as determined by the interest rate index, part of the curve they are priced off of, or embedded optionality — both contractual and behavioural.
Modeling cash flows from the bottom-up, and aggregating up the impact on the bank’s P&L and capital makes it possible to study which features of financial instruments are generating interest rate risk under a given interest rate scenario. Doing the analysis from the bottom-up also means that risk can be studied at different levels of aggregation: balance sheet level, within a specific portfolio, and all the way down to the individual instrument level.
Coupling a bottom-up modeling approach with simulation gives risk managers the ability to generate large numbers of scenarios, including alternative behavioural scenarios, to assess interest rate risk under a wide range of interest rate environments. For example, banks could simulate a range of alternative behaviours depositors will display as rates normalize. This analyse can be used to ascertain whether there are latent behavioural risks embedded in their funding structures which could crystallize in the medium-term.
More powerfully, banks can use simulation under a range of alternative future business models to guide their strategic thinking and decide between alternative funding and lending models.
It is also interesting to note that the BCBS itself even goes so far as to hint at more advanced simulation approaches along these lines:
“…banks should use a variety of methodologies to quantify their IRRBB exposures under both the economic value and earnings-based measures, ranging from simple calculations based on static simulations using current holdings to more sophisticated dynamic modelling techniques that reflect potential future business activities.”
As rates start to creep back up, interest rate risk is an increasingly important source of uncertainty. Existing statistical models simply won’t be able to offer useful prescriptive insight for risk managers in tomorrow’s interest rate environment.
Some of the world’s largest banks are using computational simulation to better prepare for tomorrow’s world. Should yours be too?