The new Instant Payment scheme, launched by EPC in November 2017 has raised an intensive discussion on how it impacts bank’s information system, introducing real-time and 24/7/365 processing in a legacy batch-oriented architecture. Nevertheless, information system topic could be the tree that hides the forest. Bank’s processes are also deeply impacted and being one of them is bank’s liquidity management. Banks have currently 2 types of proposal to exchange Instant Payments:
- Using an ACH like RT1 from EBA Clearing or SCTInst service from STET, settling IP with commercial money collaterized with central bank money
- Using TIPS, the brand-new IP settlement service from Eurosystem and embedded in TARGET2, settling PI in central bank money directly
There is a high probability that most of the banks, or at least the key players in payment, will have to participate in both. Hence, that introduces a double challenge for liquidity management.
Complexity to make treasury prediction
The first challenge relates to the Bank treasurer’s activity, which is already a very complex puzzle to solve: comply with the different businesses constraints and needs for their payment flows, including securities in T2S, limiting the intraday liquidity requirement and optimizing the end-of-day utilisation of liquidity excess. This extremely balanced act could be achieved as of today with a smart flow monitoring and release management, based on back-end notifications. But in a real-time context, it seems impossible to maintain the same logic. First phase of IP implementation being only focused on peer-to-peer payments, this is not yet critical. But this will shortly be an obstacle to reach a critical mass when implementing IP on B2C and B2B to replace legacy instruments like cards, cheques and even cash.
Fragmentation of liquidity
From the ECB point of view, and particularly regarding Monetary Policy, this new IP framework based on multiple prefunded IP accounts will create fragmented unused pools of liquidity. Participating bank, considering that the reliability of IP service is critical and any failure in settlement in not an option, will tend to fund their ACH mirror account over their real need. Since the cost of liquidity is not a big issue today, banks will for sure accept to pay for this security. Therefore, a significant amount of liquidity will not benefit to the economy, hindering the result of negative deposit facility rate policy initiated by ECB in 2014. Even if at one bank scale, the amount to be considered is not serious issue, it becomes more sensitive at European banking community scale.
Artificial Intelligence for adaptive predictive patterns
Hopefully, the choice of ISO20022 standards to support the IP scheme provides the ecosystem with a very good opportunity to leverage on (attending that the quality of data provided is good enough). Using this rich and structured source of information in payment messages, combined with customer’s profile information and open data, could lead to better understand why this payment has been made, in which context. Through a smart clustering approach and with the help of machine learning and AI, it is then possible to imagine new predictive patterns for treasurers. Its sounds logic, as the payment act becomes more and more transparent and integrated in the customer journey, to focus now on business activity prediction instead of cash flow forecast. SKAIZen Group’s specific research focus on AI, Open Data and Machine learning provides its customers with a relevant enabler to start moving forward.