“Liquidity is oxygen for a financial system.” – Ruth Porat
In a previous article we discussed the precarious position of cash liquidity in the current economic climate. As intraday liquidity represents up to a third of reserves at banks, millions of dollars in annual funding costs are tied to it. In this regard, we highlighted the practical benefits of network analytics and simulation for modelling and optimizing payments and liquidity.
The decentralized nature of payment system designs at banks, pressures for more up-to-date and granular data from regulations (i.e. BCBS248), and the current Covid environment have made having a robust liquidity management system more essential than ever. In this article we will use a network representation of payment interdependencies and incorporate Liquidity Saving Mechanisms (LSMs) to improve intraday liquidity usage and to reduce cost within the treasury department of banks. This proven analytical approach enables bank treasurers to observe clusters and interconnectedness of the payments network helping to identify key entities and sources of risk. As the network representation lends itself to large data sets of payment flows with tools to emphasize throughput and measures of concentration, it can also highlight previously unseen opportunities and channels for greater operational efficiency. Finally, by incorporating proven techniques in LSMs, such as more advanced queueing strategies, partial netting, bilateral offsets, and payment splitting treasury departments can manage and balance faster speed of settlement with lower intraday liquidity requirements.
To illustrate these practical financial benefits of network analysis and simulation we will examine a simple representation of an intrabank payments network using FNA’s Intraday Liquidity Optimization solution.
In the screenshot above blue circles (or nodes) represent fictitious bank subsidiaries and nostro accounts. The size of the circle represents the average delay of payments (so that larger circles such as ‘Branch B – Nostro GBP’ take longer to make payments). The directed arrows (or links) between entities represent the direction of flow of payments with the thickness of the line showing the value of payments between any two entities. This way we are able to visually identify and quickly direct attention to areas of high exposure, focusing on the largest nodes and the thickest links to see intrabank entities with the largest potential for improvement. For instance, in the above example we can easily see the significance of the thick link extending from ‘Subsidiary B’ as well as the large node ‘Branch B – Nostro GBP’ indicating liquidity pressures.
Next we use our simulation approach to incorporate LSMs to see if we are able to improve intraday liquidity usage and therefore reduce costs. Simulations enable us to experiment and very simply examine the effects of different system designs, disruptions to payment flows, and liquidity shortages, equipping users to closely replicate the actual operational environment and to develop optimal approaches specific to their payments network traffic.
LSMs may use a variety of policies such as queuing, bilateral offsets, multilateral queue optimization, and payment splitting. But there are inevitable trade-offs. Policies will be largely dependent on the operational model of the bank and factors such as the set of accounts, transactions, and business operations also need to be considered. All this highlights the importance and the practical benefits of having a safe, accurate and realistic way to test different scenarios, helping to determine what policies would be most effective given the features and foundations of the infrastructure.
What is the effect on liquidity demands of current policies? Should the institution consider different queueing mechanisms as opposed to simple FIFO? Should tighter overdraft limits and bilateral netting be considered? How does the flow of intraday liquidity improve when incorporating payment splitting regimes?
In our original network example above we used FIFO without making use of overdraft limits, offsets, payment splitting, or queueing optimization schemes. These policy changes can be examined individually or in combination to evaluate the impact on the flow of payments and intraday liquidity. Using more sophisticated approaches we were able to reduce the intraday liquidity requirements of this network by 24% and reduce payment delays by 51%. This analysis can easily be extended to CBDCs and blockchain as it is agnostic to the underlying technology for payments that inform liquidity issues and as such the network approach can simulate blockchain based systems as well.
In this article we discussed how to manage intraday liquidity to save on significant carry costs of liquidity reserves. Applying network analysis to identify potential LSMs enables bank treasury departments to identify significant savings and to reduce payment delays.
The FNA Intraday Liquidity Optimization solution enables treasurers at banks to map their network of payments and to run simulations to optimize intraday liquidity. To learn more about our capabilities around liquidity optimization, please get in touch.
Mohsen Namazi (firstname.lastname@example.org) is a Managing Director of FNA based in North America. FNA’s intraday liquidity optimization solution is based on more than 20 years of relevant experience, initially with national payments systems and more recently with banks.