Jeremy Clark, head of asset & liability management, group market risk, OCBC, discusses the accurate half-life functions of addressing CASA profiles.
April 11, 2014 | Jeremy ClarkStress Liquidity modelling for retail and commercial banks can be a real challenge for liquidity managers. There is a plethora of different ‘narratives’ to define different scenarios for different uses, ranging from the familiar idiosynchratic and/or market-wide scenarios to regulatory-driven industry-wide scenarios, through to other short (‘acute’) or longer, more drawn out (‘chronic’) scenarios.
The building and defence of key assumptions can be a real issue, not least where the overall cumulative stress cash outflows are so nearly as large as the inflows. In such cases, the overall result, including possible limit breaches, can hinge very much on the assumptions put into the models themselves. ‘Reverse stress testing’ becomes critical.
One of the biggest challenges is how to treat non maturity demand deposits – like current and savings accounts (‘CASAs’). These products have a degree of behavioural inertia, even in a stress scenario. More to the point, however, their volumes tend to be very large as a proportion of the overall commercial funding base of banks. Getting the assumptions even slightly wrong here can have a huge impact on the overall stress liquidity result.
CASA accounts can be further complicated by issues such as ‘burnout’ in stress scenarios, where the rate of loss of deposits slows as the scenario goes on for no other reason than the ‘hottest’ deposits have already left the Bank, leaving proportionately more ‘colder’ deposits behind. Liquidity managers need to model cash outflows in a simple, comparable way – but also in a way which results in a predefined daily cash outflows.
Some banks have turned to half-life functions to address these issues. With two simple variables (the half-life – defined in in days, and the proportion of the CASA portfolio to which the function is applied) then nearly any cash outflow profile can be built. This can then be calibrated or compared from on...
Categories:
Credit Risk, Liquidity Risk, Risk and RegulationKeywords:OCBC, CASA, Half-life
