Use of AI-controlled dynamic periodic net settlement mechanisms in real-time payment market infrastructures
Since the 2008 global banking crisis, it has become increasingly important to ensure that prominent retail payment market Infrastructures such as national automated clearinghouses or real-time payment platforms or systems are fully collateralized and no longer dependent on the taxpayer to cover default risk. This has led to an increased variety of models and a move away from traditional batch deferred net settlement systems (DNS) toward line-by-line direct settlement (DS) models that ensure transactions pass through only when liquidity is available to cover them.
In this paper, we propose a new model where the netting period used in a DNS model is dynamic to manage liquidity efficiently based on market variables. Essentially, the system will alter, monitor and optimize itself, as required, based on the prevailing conditions. In this way, the system can retain a higher liquidity efficiency while ensuring that risk levels remain constant regardless of transaction volume or velocity.