Financial Intelligence Research

1
Financial Intelligence Research

• Prof. Huaiqing Wang, PhD
• Dept of Information Systems
• College of Business
• City University of Hong Kong
• iswang_at_cityu.edu.hk
• http//144.214.5.243/iswang
• Sept. , 2009

2
Agenda

• Introduction
• Intelligent Financial Process Monitoring and
  Management
• Financial Process Risk Monitoring
• Exception management in securities trading
• Money Laundering Monitoring and Detecting
• Investment Intelligence
• Intelligent Family Wealth Management
• Change detection in data-mining for financial
  securities databases
• On-going Research (My PhDs)

3
Why?

• Modern Finance Needs
• Smart financial institutions need intelligent
  activities
• Mass, complex and incomplete financial data and
  information
• Smart thinking, deep analyzing and reasoning

4
How?

• Applying Innovative Technologies
• Artificial intelligence Intelligent agents
• Data mining Knowledge discovery
• Modeling Ontology
• Service oriented computing Web services
• Knowledge management

5
Definitions

• Financial Intelligence gathering of information
  about the financial affairs of entities of
  interest, to understand their nature and
  capabilities, and predict their intentions.
  --- Wikipedia
• Intelligent Finance emerging from the confluence
  of several distinct disciplines in financial
  market analysis, investing and trading, etc. It
  is conceived as the science, technology and art
  of the comprehensive, predictive, dynamic and
  strategic analysis of global financial markets,
  --- Intelligent Finance an
  emerging direction, Quantitative Finance, Vol.
  6, No. 7, 2006, 273-277

6
Financial Process Risk Monitoring

• Huaiqing Wang, and Chen Wang, Visibility of the
  Compass Rose in Financial Asset Returns a
  Quantitative Study, Journal of Banking and
  Finance, Vol. 26, No. 6, June 2002 pp 1099-1111.
• Huaiqing Wang, John Mylopoulos and Stephen Liao,
  "Intelligent Agents and Financial Monitoring
  Systems", Communications of the ACM, Vol. 45, No.
  3, March 2002, pp 83-88.
• Many spectacular losses in the financial industry
• The Baring Lost Case (Nick Leeson and the
  Collapse of Barings Plc) , 1995
• These (financial) problems reflect inadequate
  monitoring and supervision -- Henry Kaufmann,
  Business Week

7
Financial Risk Monitoring

• With the rising tide of financial fraud and the
  unmonitored undertaking of financial risks,
  institutions are enforcing stricter rules to
  ensure the monitoring and the supervision of
  financial transactions (Paul Munter, 1998)

8
Research Approach
9
Monitoring Agent

• A Monitoring agent performs continuous monitoring
  of the activities of trading agents automatically
  during the trading process by recording,
  detecting and analyzing the trading data.

10
Simulation

• Simulates two month main transactions by Nick
  Leeson
• In this simulation, our system gave three
  warning messages
• unauthorized trade warning, because the
  trading amount 6,000 is over the amount-limit.
• position warning, because the contract 43,320
  is over the contract-limit.
• dangerous loss warning, because the total loss
  (14.65 billion) is over the loss-limit.
• Also, the supervisor agent sends a message to ask
  the top-manager to stop the trade.

11
Contributions

• Demonstrates the benefits of a light-weight
  architecture
• Provides the benefit of scalability in terms of
  dealing with the business activity monitoring.
• By utilizing a formal conceptual model, the
  intelligent agents are capable of carrying out
  unambiguous and concise communication by a common
  knowledge repository.

12
Exception management in securities trading

• Selected Publications
• Minhong Wang, and Huaiqing Wang, From Process
  Logic to Business Logic -- A Cognitive Approach
  to Business Process Management, Information
  Management, Vol.43, No.2, 2006, pp.179-193.
• Minhong Wang, Huaiqing Wang, and Dongming Xu,
  "Intelligent Workflow Monitoring with Agent
  Technology", Knowledge-based Systems, Vol.18,
  No.6, 2005, pp.257-266.
• Minhong Wang, Huaiqing Wang, Dongming Xu, Kwok
  Kit Wan and Doug Vogel, A Web-service Agent-based
  Decision Support system for Securities Exception
  Management, Expert Systems with Applications,
  Vol.27, No.3, 2004, pp.439-450.

13
Background

• Financial markets face rising trading volumes and
  increasing risks caused by inefficiencies in
  transactions
• The 80-20 rule (Wall Street Technology Online,
  2000)
• STP - Straight Through Processing
• (U.S. Securities and Exchange Commission,
  Securities Industry Association)
• Exception Management
• STP requires trade information passed within the
  trade lifecycle timely and accurately (Simmons
  2001, Flectcher et al. 2002)
• Major issue of exception management has not been
  adequately addressed in securities trading
  (Guerra 2002)
• Complex process in exception management

14
Analysis

• Exception Analysis

• Deployment of Agents
• Monitoring agents
• Trade details monitoring agent, Trade status
  monitoring agent
• Diagnostic agent
• Resolution agent
• Interaction agents
• Trading interaction agent, Settlement interaction
  agent

15
Cognitive Agents and Web Services Supported
Exception Management (EM)

• EM is a complex process, in which multiple
  organizations and mixture of human activities and
  automated tasks may be involved.
• Agent technology is proposed by modularizing the
  complex problem in terms of multiple autonomous
  components
• e.g. monitoring agents, diagnostic agents,
  resolution agents, etc.
• Cognitive agents supported approach is used to
  manage complex tasks in EM (Wang et al. 2004
  HICSS-38)
• Web Services enable integration models for
  facilitating program-to-program interactions

16
System Architecture
17
Cognitive Agents inComplex Exception Management
18
Knowledge Engineering of Agent
-- Diagnostic Agent

• Situation knowledge
• Error reports, special trades, diagnosing
  activities, etc.
• (event (e-type error_report) (err-no 102))
• Process knowledge
• Rules for scheduling diagnostic tasks, monitoring
  diagnostic activities, resolving special errors,
  etc.
• (defrule report-emergent-error
• (event (e-type error_report) (err-no ?e_no)
  (trd-no ?t_no))
• (large-trade (trd-no ?t_no))
• gt (assert (task (t-type emergent_report) (err-no
  ?e_no) (start-time (time)))))
• Problem solving knowledge
• Rules for exception analysis, etc.
• (defrule check-missing-items
• (trade (trd-no ?n) (security ?s) (price ?p)
  (quantity ?q) (custodian ?c)
• (trade-date ?t ?t1) (value-date ?v ?v1))
• gt (if (or (eq ?s nil) (eq ?p nil) (eq ?q nil)
  (eq ?c nil) (eq ?t nil) (eq ?t1 nil)
• (eq ?v nil) (eq ?v1 nil)) then (assert
  (item_missing ?n))))

19
Evaluation Demonstration

• User satisfaction

20
Evaluation Demonstration (cont)

• User feedback

21
Intelligent Money Laundering Monitoring and
Detecting

• Group of Seven Industrial Democracies (G-7)
  Financial Action Task Force (FATF)
• FATF 40 Recommendations (1990)
• Know-your-customer (KYC)
• Public explored person (PEP) List
• Non-corporative Countries and Territories (NCCT)
• Annual Report on Money Laundering Typologies

22
The Global Anti-Money Laundering Regime
Pillars of the Anti-money Laundering Regime
(Adaptived from Reuter and Truman, 2004)
23
Research Motivation

• Weakness of common bank practice assisted with
  computer systems
• The amount of detected ML is inadequate and
  incompatible (Levi Reuter, 2005)
• Our Research Focus
• Electronic ML methods Internet facilitate ML
  (Philippsohn, 2001)

24
Decision Making Process Model for IMLMDS
Figure 1. Decision Making Process Model for IFDSS
25
Multi-agent System Architecture for IMLMDS
26
Intelligent Family Wealth Management (FWM)

• Shijia Gao, Huaiqing Wang, Dongming Xu, and
  Yingfeng Wang, Intelligent Agent Assisted
  Decision Support for Family Financial Planning,
  Decision Support Systems, 44 (1), 60-78, November
  2007.
• Shijia Gao, Huaiqing Wang, Yingfeng Wang, Wenqi
  Shen, Siubun Yeung, Web-service-agents-based
  Family Wealth Management System, Expert Systems
  with Applications, 29 (1) 219-228 JUL 2005.
• Family Wealth capital (intellectual, emotional,
  financial, or otherwise) used to secure a given
  quality of life for the whole family members.
  (Financial Advisor Pages http//www.fapages.com)

27
Research Objective

• Apply innovative technology
• to develop an
• integrated and comprehensive FWM system that can
  overcome previous problems.
• Technologies applied
• Intelligent Agents Multi-agent system
• Web-services
• Integrated aspects
• Financial Portfolio Risk Management
• Accounting Cash Flow Management
• Legal Tax Estate Planning

Web-service Intelligent Agents
28
Agent Hierarchy
29
System Architecture Intelligent FWM System
(IFWMS)
30
Multi-agent Interaction
31
Implementation

• Java Web-services Development Package (JWSDP)
  (Java.sun.com, 2004)
• Using SOAP (Simple Object Access Protocol) for
  agents communication (SOAP objects are done by
  JAXM), while KQML (Knowledge Query and
  Manipulation Language) is wrapped by SOAP
• Agent reasoning (i.e., knowledge level) is
  developed by Java Expert system Shell and
  Scripting language (JESS)
• Using WSDL (Web-services Description Language) to
  describe and publish agents
• Using UDDI (Universal Description, Discovery and
  Integration) to create directory of agents

32
System Operation a Use Case

• A wealthy investor with moderate risk-tolerance
  level, currently hold an investment account
• A friend of him told him that it is a good time
  to buy a new-issued China stock
• ? How could IFWMS advise him?
• His family ordered a new car, which requires him
  to pay the down payment
• ? How about the cash-flow management and estate
  planning?

33
IFWMS Web-service Agents Collaboration
Interaction
34
Change detection in data-mining for financial
securities databases

• Xiaoyan Liu, Zhenjiang Lin and Huaiqing Wang,
  Novel Online Methods for Time Series
  Segmentation, IEEE Transactions on Knowledge and
  Data Engineering, Dec 2008, pp 1616-1626.
• Xiaoyan Liu and Huaiqing Wang, A
  Discretization Algorithm Based on a Heterogeneity
  Criterion, IEEE Transactions on Knowledge and
  Data Engineering, Vol. 17, No. 9, September 2005,
  pp1166-1173.

35
Contributions

• Technology Proposed a novel data reduction
  method and a set of effective online change
  detection methods for data streams, that support
  building the data mining models in data streams
  more effectively and accurately
• Finance provided a tool to reduce the
  information asymmetry in the securities markets
  and support the market short-term players making
  short-term trading decisions.

36
On-going research (Ye Kang, BSc USTC)

• Knowledge Level Modeling for Systemic Risk
  Management in Financial Institutions
• This model has comparable quality, e.g. semantic
  quality
• Benefit to enhance interchange of information and
  knowledge sharing, assist KB design, and support
  inter-organizational coordination
• Kang Ye, Shanshan Wang, Jiaqi Yan, Huaiqing Wang
  and Baiqi Miao, Ontologies for crisis contagion
  management in financial institutions, Journal of
  Information Sciences, 35(5), OCT 2009, pp
  654-666.
• Working in Shanghai Securities Exchange after PhD

37
On-going Research(Wang Shanshan, BSc USTC)

• Data mining for causal relationships between news
  and financial instruments
• An ontology of such causal relationships
• An ontology based data-mining framework
• Shanshan Wang, Kang Ye, Zhe Zhang, Huaiqing Wang,
  and Xiaojian Chen, An ontology for causal
  relationships between news and financial
  instruments, Expert Systems with Applications,
  Vol. 35, Iss. 3, 2008, pp 569-580.
• Working in Inner Mongolia University after PhD

38
On-going Research(Wang Yingfeng, BBAIS, MPhil,
CityU)

• Multi-intelligent-agent-based Decision Support
  System for Target Selection in Hedge Fund
  Activism
• To improve the decision making process in target
  selection and to enhance the decision quality
  through negotiation, compromise and collaboration
  among agents.

39
On-going Research(Terrance Fung, BSc, MSc, Univ
of Wisconsin)

• Illegal Trade-based Manipulation and Surveillance
  in Financial Markets
• Modeling trade-based manipulations and the
  mechanisms in detecting them

40
On-going Research(Lai Hok Yin, BSc PolyU, MSc
CUHK)

• Detecting Market Manipulation in the derivatives
  market
• Design an intelligent market manipulation
  detection system, especially for the derivatives
  market at the first stage.

41
On-going Research(Zonghang Yang, BSc Fudan,
MPhil CityU)

• Kernel-based simulation and application in
  financial market
• An efficient mesh-free method is proposed to
  price interest rate derivatives under the LIBOR
  market model

42
On-going Research(Chen Kun, BSc, MSc, Shangdong
U)

• News sensitive risk management for family
  financial plan a general knowledge supported
  approach

43
On-going Research(Cheng Lin, BSc Anhui FEU )

• World Knowledge-based Risk Management in Family
  Financial Plan

44
On-going Research(Wu Jiani, BSc Hefei U of Tech.)

• Modeling tunneling and propping in pyramidal
  structure using social ontology
• Strategic Dependency (SD) model which describes
  the network of dependency relationships among
  actors in pyramid structure
• Strategic Rationale (SR) model which provides a
  more detailed level of modeling by looking
  inside actors to model intentional
  relationships

45
On-going Research(Yan Jiaqi, BSc USTC)

• Belief-Desire-Intention Modeling for Quality
  Management in Cooperative Supply Chain
• Model the potential intentions or behaviors of
  upstream producers during the quality management
  process
• A foundation to design intelligent systems to
  monitor and detect quality problems timely and
  effectively in inter-organizational business
  process

46
The End