Tools for Enhancing Efficiency of Interbank Payment Settlements
Mathematical Methods and Software Products for Enhancing Efficiency of Interbank Payment Settlement Systems
Tech Area / Field
- INF-SOF/Software/Information and Communications
- INF-OTH/Other/Information and Communications
3 Approved without Funding
National Academy of Sciences of the Republic of Belarus / Institute of Informatics Problems, Belarus, Minsk
- National Bank of the Republic Belarus, Belarus, Minsk\nBelarussian State University / Institute of Technologies of Informatization and Management of BSU, Belarus, Minsk
- Clark Strategies Inc., USA, MA, Lexington\nCentrale des Reglements Interbancaires, France, Paris\nUniversity of Southampton, UK, Southampton\nRutgers University / Center for Operations Research, USA, NJ, Piscataway\nUniversity of Graz / Institut fur Statistik Okonometrie und Operations Research, Austria, Graz\nCastlerock Corporation, USA, MA, Stow\nInstitut für Betriebliche Informationsysteme und Expertensysteme, Germany, Dortmund
Project summaryThe Project's Purpose
Development of tools for enhancing efficiency of interbank payment settlement systems by minimizing requirements of banks in intraday liquidity. These tools include models, methods and software products for forming an optimal subset of payments for bilateral and multilateral netting subject to restricted predeposited funds allocated by banks for a clearing session as well as some other constraints.
The State of the Art in the Field
Minimizing the intraday liquidity requirements of banks for executing interbank funds transfers and settlements is one of the ongoing pressing problems. Solution of such problems could enhance efficiency of banking business. Facilities for decreasing banks’ requirements in intraday liquidity are available in some interbank settlement systems, including EAF-2 (Deutsche Bundesbank)..1. Gntzer M.M., Jungnickel D., Leclerc M. Efficient algorithms for the clearing of interbank payments. European Journal of Operational Research, 1998, 106, p.212 – 219. , PNS (Centrale des Reglements Interbancaires, France) ..2. The PNS System. La Lettre de la Centrale des Rglements Interbancaires No. 12, 1999.. The Bank of Finland conducts research aimed at developing such tools..3. Leinonen H. Interbank funds transfer systems: liquidity needs, counterparty risks and collateral. Bank of Finland Discussion Papers, 1999, No. 16/98, 38 p. [3,4]. Analysis of the above mentioned facilities demonstrates that all of them are based on solving the problems of forming a subset of payments for netting subject to restricted funds allocated by banks for a clearing session as well as under some other constraints. Results presented in  show that even the most simple problems of this type are NP-hard. This fact does not leave a hope for a possibility to create fast algorithms for solving these problems to the optimum. At present, the existing interbank payment settlement systems employ relatively simple approximate algorithms for solving problems in question..4. Leinonen H., Soramki K. Optimizing Liquidity Usage and Settlement Speed in Payment Systems. Bank of Finland Discussion Papers, 1999, No. 16/99, 39 p. [1,2]. These algorithms are sufficiently effective when banks predeposit high or medium funds balance for a given clearing session (from 5 to 50 % of the total payments value) . At low level of predeposited funds (1 - 2 % and less of the total payments value) the accuracy of these algorithms is greatly decreasing. Besides, specific features of these algorithms disregard some additional constraints such as, for example, bilateral credit limits in multilateral netting schemes.
The Impact of the Proposed Project on the Progress in the Field
Realization of the proposed project will provide an expansion of the variety of solved problems, increase accuracy of algorithms for solving the problems. Besides it will enable to create the system for perfecting methods of solving the problems in question, selection of the most efficient algorithms, and for perfecting technologies of using the developed methods and software products in hybrid systems of interbank payment settlements.
The project will provide a number of economic beneficial effects. These effects include mainly practical cost savings that will be achieved by banks participating in the interbank settlements schemes due to higher value and number of settled payments and, as a result, lower costs for reserving and predepositing required intraday liquidity.
Project team includes 7 scientists directly involved in the development of computer-aided designing systems for missiles, development of radio engineering equipment and testing devices for missiles, development of systems and devices for tracking missile targets in counter missiles facilities. From specialization point of view the project participants could be subpided into two groups. The first group consists of 8 specialists in the field of discrete optimization and development of software products. They have published more then 120 scientific papers and 4 monographs.
The second group includes 3 specialists in the area of interbank payment systems. They are practitioners in payments and settlement operations and, in particular, they are coauthors of the “Concepts for Development of the National Payment System of the Republic of Belarus”. They have published more than 20 articles and reports in specialized industry journals.
Expected Results and Their Application
Fulfillment of the project will provide the following results.
(1) Mathematical models of decision-making situations in interbank payment settlement systems, related to forming a subset of payments for netting.
(2) Fast approximate methods and algorithms for forming a subset of interbank payment orders for netting with maximum total value and number of settled payment orders in multilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session.
(3) Fast approximate methods and algorithms for forming a subset of interbank payment orders for netting with maximum total value and number of settled payment orders in bilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session and under bilateral credit limits.
(4) Fast approximate methods and algorithms for forming a subset of interbank payment orders for netting with maximum total value and number of settled payment orders in multilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session and under bilateral credit limits.
(5) Methods and algorithms for computing upper bounds of objective functions’ values in the problems under consideration. These bounds will be used for estimating the quality of solutions obtained by various approximate methods. This, in turn, will permit to select the most efficient methods.
(6) Software applications like stand-alone programs implementing the developed algorithms.
(7) Simulating system with the following basic functions:
- generation of interbank payment orders flow with controllable stochastic characteristics, initial funds allocated by banks for a given clearing session, and bilateral credit limits;
- forming a subset of interbank payments for netting subject to a given constraints using the relevant stand-alone program;
- computing bounds of objective functions’ values and estimating deviation of the obtained variant of the netting subset from the optimal one;
- operational simulation of the hybrid type interbank settlement systems aiming to select the most appropriate combination of gross-based settlement procedures with the net-based settlements.
(8) Principles of implementing developed algorithms and software products in interbank payment settlement systems.
Developed methods, algorithms and software products for forming subsets of payments for netting could be used as a basis for the relevant marketable software products.
Simulating system could be used to analyze efficiency of various methods of forming subsets of payments for netting as well as to perfect alternative strategies of employing netting procedures in settlement systems of hybrid type.
Meeting ISTC Goals and Objectives
Project activities are in compliance with the ISTC objectives and targets because their implementation
- will provide a possibility for seven specialists earlier involved in the development of missiles technologies, to redirect their professional skills and knowledge to solving problems related to enhancing efficiency of banking systems functioning;
- will provide a possibility to create intellectual products in the form of algorithms and related software products oriented to potential end users and will form a basis for developing marketable products;
- will provide a possibility for weapons scientists to integrate into the world scientific community in the area of discrete optimization and its applications.
Scope of Activities
Estimated project duration is 2.5 calendar years. It is planned to involve in the project 11 scientists and specialists and 4 support personnel from three institutions. The total number of persons-months of effort is 165.75. The following tasks are supposed to be solved during the project implementation.
(1) Development of mathematical models of decision-making situations in interbank payment settlement systems related to forming a subset of payments for netting.
(2) Development of fast approximate methods and algorithms for forming a subset of interbank payments of maximum total value and number in multilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session.
(3) Development of fast approximate methods and algorithms for forming a subset of interbank payments of maximum total value and number in bilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session and subject to bilateral credit limits.
(4) Development of fast approximate methods and algorithms for forming a subset of interbank payments of maximum total value and number in multilateral netting schemes under restricted predeposited funds allocated by banks for a given clearing session and subject to bilateral credit limits.
(5) Development of methods and algorithms for computing upper bounds of objective functions’ values for problems of forming a subset of payments described in items (2)-(4).
(6) Development of software applications like stand-alone programs implementing the developed algorithms.
(7) Development of simulating system.
(8) Conducting experiments with stand-alone application programs and simulating system aimed at selecting the most efficient application programs, working out recommendations for employing different application programs depending on values of initial parameters of the problems as well as recommendations for employing netting procedures for better operation of the hybrid type settlement systems.
Role of Collaborators
The principal role of the collaborators consists in the following:
- participation at the stage of problems formulation, working out requirements to algorithms and software applications and in choosing approaches to development of solution methods;
- evaluation of intermediate and final results;
- advising on copyrights of intellectual property created as a result of the project fulfillment.
Collaborators will receive the following results: description of models and methods for solving problems, executable application programs, simulating system, and principles of using software facilities.
Possible ways of collaboration are exchange of information, consultations, joint developments, publications, shared activities aimed at delivering project results to the market.
Technical Approach and Methodology
The team of the project participants has an experience and skills in solving problems of forming a subset of payments for netting. They have developed a variant of algorithm based on search of contours in a directed graph representing a set of payments. First experiments with the software application have demonstrated high quality of resulted solutions.
In the course of the project implementation it is supposed to use methods of discrete optimization, graph theory, theory of network flows, theory of algorithms complexity.
The main supposed trend is to construct graph models (task 1) and to use procedures of searching contours in directed graphs while forming a subset of payments for netting (tasks 2 - 4). In order to construct upper bounds of objective functions’ values (task 5), methods of constraints relaxation and methods of searching flows with prescribed features in networks will be used.
For developing stand-alone application programs (task 6) and simulating system (task 7) the developers will use programming language C++.
The International Science and Technology Center (ISTC) is an intergovernmental organization connecting scientists from Kazakhstan, Armenia, Tajikistan, Kyrgyzstan, and Georgia with their peers and research organizations in the EU, Japan, Republic of Korea, Norway and the United States.
ISTC facilitates international science projects and assists the global scientific and business community to source and engage with CIS and Georgian institutes that develop or possess an excellence of scientific know-how.