450 likes | 693 Views
Design of Clearing and settlement operations: A case study in business process modelling and evaluation with Petri nets. Introduction. Name : Pieter Kwantes Work at ING Bank, Wholesale Securities department
E N D
Design of Clearing and settlement operations:A case study in business process modelling and evaluation with Petri nets
Introduction • Name : Pieter Kwantes • Work at ING Bank, Wholesale Securities department • My contribution is based on my master thesis Computer Science (Leiden University. Jetty Kleijn) finished May 2006
Content of this presentation • Motivation behind the case study • Case study • Problem statement • Approach and research questions • Execution of the case study • Concluding remarks
Motivation • Choice between different process design alternatives, is often based on intuition, expert opinion. Tools: MS-word and power point. • The question is whether a formal approach based on Petri nets might do better • Try it out in a realistic case study
Problem statement (1) • Clearing and settlement = processing of transactions on the secondary capital market • Primary capital market : issue of new securities • Secondary capital market : trading of already issued securities
Trading Clearing Settlement Custody Problem statement (2)processing of transactions on secondary capital market
Problem statement (3) • Problem : how to design a business process able to deliver Clearing and settlement services ? • This resulted in 2 design alternatives • How to decide which one is the best, before starting implementation?
Approach: the basic idea • Make a Petri net model of each alternative process design • Calculate the performance of each model • Choose the design showing the best performance for implementation
Approach:steps • Acquire informal description process designs • Establish performance criteria • Formalize descriptions with Petri nets • Validate formal description • Measure performance • (Implement best process design)
Questions • Questions: • Are Petri nets suitable to describe critical properties of the subject domain : C&S ? • Are graphical representation and animation enabled by Petri nets helpful for validation purposes? • Is it possible and feasible to decide between two process designs of C&S using Petri nets • (Can PN-based models be used as a specification/blueprint to guide implementation?)
Content of this presentation • Motivation behind the case study • Case study • Problem statement • Approach and research questions • Execution of the case study • Concluding remarks.
Approach:steps • Acquire informal description process design • Establish performance criteria • Formalize description with Petri nets • Validate formal description • Measure performance • (Implement best process design)
Informal description (1) • Two different institutional arrangements for secondary capital markets: • “Over the counter” markets • Formalized exchanges
Trading Clearing Settle ment Custody Informal description (2) Trading Trading Member Firm 1 100 units Shell Trading Member Firm 2
Trading Clearing Settle ment Custody Informal description (3) Trading Trading Member Firm 1 100 units Shell Trading Member Firm 2 Clearing 100 units Shell Central Counter party 100 units Shell
Trading Clearing Settle ment Custody Informal description (4) Trading Trading Member Firm 1 200 units Shell Trading Member Firm 2 200 units Shell 200 units Shell Clearing Clearing Member Firm B Clearing Member Firm A 200 units Shell Central Counter party 200 units Shell
Trading Clearing Settle ment Custody Informal description (5) 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Clearing Member Firm Clearing Member Firm Central Counter party Custody Power of attorney 0 0 100 100 0 Custodian Custodian -Invariance 200 0 0 CSD 0 0
Trading Clearing Settle ment Custody Informal description (6) 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Clearing Member Firm Clearing Member Firm Central Counter party Custody Power of attorney 100 0 0 100 0 Custodian Custodian -Invariance 100 100 0 CSD 0 0
Trading Clearing Settle ment Custody Informal description (7) 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Clearing Member Firm Clearing Member Firm Central Counter party Custody Power of attorney 0 0 0 100 0 Custodian Custodian -Invariance 100 0 0 CSD 100 0
Trading Clearing Settle ment Custody Informal description (8) 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Clearing Member Firm Clearing Member Firm Central Counter party Custody Power of attorney 0 0 0 100 0 Custodian Custodian -Invariance 100 0 100 CSD 0 0
Trading Clearing Settle ment Custody Informal description (9) 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Clearing Member Firm Clearing Member Firm Central Counter party Custody Power of attorney 0 100 0 100 0 Custodian Custodian -Invariance 100 0 0 CSD 0 100
Trading Clearing Settle ment Custody Informal description (10)design alternative 200 units Shell Trading Member Firm 2 Trading Member Firm 1 Trading Member Firm 3 Central Counter party Clearing Member Firm Custody Power of attorney 100 100 0 Custodian Custodian -Invariance 200 CSD 0 0
Trading Clearing Settle ment Custody Informal description (11)design alternative Trading Member Firm 1 200 units Shell Trading Member Firm 2 Trading Member Firm 3 Central Counter party Clearing Member Firm Time : S Custody Power of attorney (-100) 100 0 Custodian Custodian -Invariance -Timing -Power of attorney CSD 0 200 0
Trading Clearing Settle ment Custody Informal description (12)design alternative Trading Member Firm 1 200 units Shell Trading Member Firm 2 Trading Member Firm 3 Central Counter party Custody (-100) 100 Custodian 200 Custodian CSD 0 0 200
Trading Clearing Settle ment Custody Informal description (13)design alternative Central Counter party Trading Member Firm 1 Trading Member Firm 3 Trading Member Firm 2 Custody 0 0 Custodian 200 Custodian CSD 0 0 200
Informal descriptiondifference between design 1 and 2 Late received OTC-trades S - t S – ½ t S • Design 1 has 2 extra steps before final settlement with CCP, so • decision to borrow earlier • more failure due to late receiving OTC-trades CCP Design 1 CCP Design 2
Informal description DFD : design alternative 1 OTC- Market Exchange Lending Firm 7 1 10 8 9 13 Central Counter party Clearing Member Firm 12 Custodian 11 2 6 14 15 Central Securities Depository 5 4 3
Informal description DFD: design alternative 2 OTC- Market Exchange Lending Firm 23 25 1 22 24 21 20 Central Counter party Clearing Member Firm Custodian 13 2 31 30 27 29 28 26 Central Securities Depository 4 3
Approach:steps • Acquire informal description process design • Establish performance criteria • Formalize description with Petri nets • Validate formal description • Evaluate process design • (Implement process design)
Performance criteria • Basic purpose of Clearing and settlement to complete a transaction according to what has been agreed upon (amount, prices, time of delivery) • Performance criterium is • % of traded securities delivered in time • amount of borrowing
Approach:steps • Acquire informal description process design • Establish performance criteria • Formalize description with Petri nets • Validate formal description • Evaluate process design • (Implement process design)
Approach:steps • Acquire informal description process design • Establish performance criteria • Formalize description with Petri nets • Validate formal description • Evaluate process design • (Implement process design)
Approach:steps • Acquire informal description process design • Establish performance criteria • Formalize description with Petri nets • Validate formal description • Evaluate process design • (Implement process design)
Evaluate process design Measurements design 1 (2 accounts) Borrowing cost Settlement failure
Evaluate process design Measurements design 2 (1 account) Borrowing cost Settlement failure
Content of this presentation • Motivation behind the case study • Case study • Problem statement • Approach and research questions • Execution of the case study • Concluding remarks.
Concluding remarks • Suitability of HTCP-nets for modelling Clearing and settlement process • Decentralized control/communication • Complexity • Time critical • Issues: • Modelling of priorities • Modelling of dataflow/datastructure
Concluding remarks • Validation of the HTCPN model • Graphical representation • Animation • Issues: • Snap shot animation, lose view of entire OS • Local state change, snap shots of entire net • Animation of high level nets • Domain specific presentation layer • Extraction of performance information
Concluding remarks • Suitability for design evaluation • What-if analysis • Visualisation of alternatives • Issues • Support for model variants • Reusable model fragments