风险学会季刊 2011年五月

|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| IN THIS ISSUE Intelligent risk INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 LETTER FROM THE CHAIRMAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 LETTER FROM COO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 VISIONS OF RISK Systematic Generation of Plausible Scenarios: . . . . . . . . . . . . . . . . . . . . . 4 The Fusion of Best Practices from the Intelligence and Financial Communities Risk-Adjusted Performance: Lessons from the Financial Crisis . . . . . . 7 Stress Testing Trident . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 ISO 31000 Q & A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Fit for Purpose? Reassessing the Role of Risk Management . . . . . . . . 18 ACADEMIC PARTNER PROFILE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Kellogg School of Management WHAT’S ON THE WEB? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Enterprise Risk Management CHAPTER REPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 PRMIA Toronto ANNOUNCEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 PRMIA Awards LEARNING OPPORTUNITIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 PRMIA LEADERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Board of Directors, Founders, Committees EDITORIAL BOARD EXECUTIVE EDITOR Carol Alexander chair@prmia.org EDITOR Michael H. Martin editor@prmia.org PRODUCTION EDITOR Christine Bernier Lienke christine.lienke@prmia.org CHINESE TRANSLATION Xi Chen translator@prmia.org REGULARS VISIONS OF RISK Bob Mark treasurer@prmia.org WHAT’S ON THE WEB Tom Day thomas.day@prmia.org ACADEMIC PARTNER PROFILE Christine Bernier Lienke apc@prmia.org CHAPTER REPORT Amanda Ritari amanda.ritari@prmia.org Thanks to our sponsor, the exclusive content of Intelligent Risk is freely distributed worldwide to more than 75,000 PRMIA members. If you would like more information about sponsor- ship opportunities contact cheryl.buck@prmia.org SunGard’s Adaptiv provides enterprise-wide credit and mar- ket risk management and operations solutions for financial services institutions. Adaptiv assists institutions of varying size and complexity to deploy technology to meet both internal and regulatory requirements for risk management and operational control. Adaptiv helps financial services institutions from the banking, hedge fund, asset management, insurance and corporate sectors with our deep understanding of risk management and operational processes. Visit http://www.sungard.com/enterpriserisk. MAY 2011 S P O N S O R E D B Y : K N O W L E D G E F O R T H E P R M I A C O M M U N I T Y Copyright (c) 2011 Professional Risk Managers’ International Association, All Rights Reserved |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| We all know uncertainty has a way of materializing at the worst possiblemoment. As I was drafting this note a pipe over my kitchen broke andwater started seeping out of the ceiling and dripping on the table. When sending it as an email, my wireless internet access failed at two different airports. And I’m just one part of the mix. Preparing this issue of Intelligent Risk combined efforts of multiple individuals overcoming various obstacles. As you read this issue please remember that it reflects the work of skilled designers and patient (!) coordinators pulling the pieces together as well as the authors and contributors. There is a quote from the movie Shakespeare in Love that I think of whenever Murphy’s Law seems to have become as fundamental as the law of gravity. Henslowe: Mr. Fennyman, allow me to explain about the theatre business. The natural condition is one of insurmountable obstacles on the road to imminent disaster. Fennyman: So what do we do? Henslowe: Nothing. Strangely enough, it all turns out well. Fennyman: How? Henslowe: I don't know. It's a mystery. Putting on a play or publishing a newsletter have their own obstacles, rewards and mysteries. Of course, in most fields “I don’t know. It’s a mystery,” is not enough. Risk managers are expected to look deeply into that “natural condition” to manage the unknown with all of the tools, knowledge and capabilities available. For an editor, effective communications are always center stage. Conveying accurate, timely information in ways that the relevant audience understands and appreciates is significant in itself. At its best it also drives clearer thinking, new insights, better decision-making and—one hopes—timely action. We believe this issue contributes in all of these roles. PRMIA Chair Carol Alexander describes PRMIA’s new organizational structure and the organization’s vision for the future. Jodi Lundell reports on our recent membership survey. (Thank you!) A profile of the Toronto chapter illustrates how that group pursues the vision in their footprint. (Keep up the good work!) Also included in this issue is a useful Q&A discussion of a new set of risk management standards (ISO 31000) with two experts in the field. David Rowe discusses three important types of stress-testing based on history, imagination and “pessimisation.” And Chris Ray identifies several methods and analytical tools the intelligence community uses to analyze complex, adaptive systems that are increasingly applicable to financial services. Clifford Rossi argues that a simplistic focus on metrics like P/E leads to risky decisions and behavior and points out ways that regulatory capital calculations lead to decisions dramatically different than those based on economic calculations. Risk measures like RaROC may be more complicated but should improve the appreciation of risk and the decision-making necessary for its management. Enjoy the issue and please let us know what we are doing right and what we can improve. Risk management is an evolving and improving field. We look forward to everyone playing their part. Strangely enough, it may all turn out well. MICHAEL H. MARTIN, Editor michael.martin@centuryrisk.com EDITOR’S NOTE Michael H. Martin is a sustaining member of PRMIA, a senior examiner for Office of Comptroller of Currency and founder of a management consulting firm. He has written on business and risk management for Fortune, BNet and other publications. The views expressed do not necessarily represent those of the OCC. 1 INTELLIGENT RISK — MAY 2011 2 INTELLIGENT RISK — MAY 2011 PRMIA is finalizing its business plan for strategic growth. Our vision is tobecome the globally recognised professional association for risk managers offinancial markets — at the individual level and in both financial and non- financial institutions — of equal standing and stature to other professional organisations such as the CFA Institute, the Institute of Actuaries or the Association of Chartered Accountants. PRMIA will provide its members with the resources needed throughout their careers: certification, chartering, professional development, ethics and governance standards, and guidance for best practice. We will increase the portability of risk careers and act as a voice for excellence in risk management. Plans to restructure PRMIA’s committees have been informed by the recent membership survey (see next page). The revised structure appears in the diagram below. Over 600 like-minded, ethical and altruistic volunteers from all over the world work on these committees, united by PRMIA in their commitment to an enlightened future for financial risk management. That so many individuals whose talents are in demand devote so much of their precious time to PRMIA is indeed an endorsement for our collective goals. The membership survey showed that the areas that bring the most value to our members are (1) chapter activities; (2) webinars, training and global events; and (3) publications and certification. Our financial position continues to strengthen. So, as we aim for our one-year strategic goals, we are initiating some key business developments in each of these areas. Number one is increased support for chapter development. PRMIA has hired new staff for this role, and we LETTER TO PRMIA MEMBERS |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| Committees shown in blue are newly-formed. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| Our March membership surveycame at a critically importanttime as PRMIA pursues its strategic goals supported by the organization’s 2011–2012 business plan. We asked members to provide feedback on PRMIA’s brand, organizational structure, products and services. We also requested feedback on the PRM™/Associate PRM programs. Responses came from students, academics and professionals representing all levels in the corporate hierarchy from across a broad range of industries. As expected from PRMIA’s global constituency, we received input from around the world. Your responses made several things clear: � PRMIA is seen as a trustworthy, credible source for risk management information and education. Nearly 50 percent of respondents joined PRMIA at the height of the recent financial crisis, a strong indicator that professionals were looking to PRMIA for guidance through the challenging economic times. � Chapter events and webinars are valuable resources for sharing information and ideas. Almost 50 percent of all respondents identified local chapter events or webinars as the PRMIA tools providing the most value. � Publications and other published works, such as case studies, articles, and research rated among the most highly valued resources available on the PRMIA website. � Risk certification is a valuable and necessary component for career advancement. Most respondents holding the PRM certification reported that the program significantly increased their knowledge, and about 50 percent report career or salary advancement as a direct benefit. As a not-for-profit, member-led association dedicated to promoting sound risk management standards and practices globally, your insight plays a vital role in building PRMIA’s future. We asked for your time, candor and ideas. You gave us all those things. Thank you. JODI LUNDELL Chief Operating Officer jodi.lundell@prmia.org 3 INTELLIGENT RISK — MAY 2011 PRMIA MEMBERSHIP SURVEY REPORT have spoken with Regional Directors and Steering Committee members from every chapter, to discuss the functions and operations of the new Regional Director Committees (RDCs) and the sub-committees that they may appoint. The new Global Council of Regional Directors (GCRD) has been designed to improve communications and empower our chapters. Going forward, the RDC in each region will take a greater role setting and maintaining standards, while the GCRD replaces the Support and Standards Committee and assumes greater strategic responsibility. Regarding PRMIA’s global training, webinars and events, during the first half of this year we presented 30 webinars, including a successful series on the Dodd-Frank Act from Clifford Rossi, Keith Ligon, Tom Day and Christopher Laursen. We also rolled out a new series of webinars for Associate PRM candidates. The Complete Course in Risk Management remains popular in multiple cities in the US and Europe, and going forward the new Global Events Committee (GEC) will enhance PRMIA offerings at major conferences and other forums. The GEC will collaborate with the reinstated Blue Ribbon Advisory Council and build on partnerships with Sponsors. Finally, we are increasing our commitment to risk education at the institutional level, with corporate clients such as Bank of America, AIG, Ernst and Young, Standard Bank and Murex leading the way. PRMIA’s Education Committee has been reviewing its exam content and structure to ensure that it is up to date. When this is done, a new suite of supporting materials will be developed, to replace the current PRM Handbook and our online learning resources. Other exciting PRMIA developments are on the horizon, some of which you will hear about in my next letter to members. I look forward to providing more information once the business plan has been agreed in June. PROFESSOR CAROL ALEXANDER Chair of the Board chair@prmia.org |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 4 INTELLIGENT RISK — MAY 2011 SYSTEMATIC GENERATION OF PLAUSIBLE SCENARIOS: The Fusion of Best Practices from the Intelligence and Financial Communities CHRISTINA RAY The financial and intelligence communities have much in common. Each performs surveillance,gathers intelligence, and analyzes massive amounts of real-time data using highly quantitative models and methods. And whether an analyst’s task is called “threat assessment” or “risk manage- ment,” the objective is the same: to identify all possible outcomes, and to devise strategies that mitigate the likelihood or consequences of the most adverse. Nevertheless, the two communities are largely unaware of each other’s best practices. The analytical methods that each employ are based on fundamentally different views of reality. Still, with the increasing focus within the financial community on stress testing and dynamic modeling,1 the methods of the intelligence community have increasing applicability to the financial risk manager’s mission. Recent experience indicates that there is good reason to view the global financial system—in the language of network analysis—as complex, adaptive, and self- organizing. As such, the system can exhibit emergent behavior; that is, rapid transformation to a new state in reaction to exogenous or endogenous shocks. Such a system can at times be both robust (e.g., remarkably resilient to simultaneous Japanese earthquakes and a third US military action in Libya) and fragile (e.g., sen- sitive to a single order touching off a flash crash). So, analysis of the global financial system requires quantita- tive approaches that accommodate emergent behavior. The two world views each require their own analyti- cal methods. Experts in statistics come in two flavors: “frequentists” and “Bayesians.” The financial commu- nity is most heavily populated with frequentists (with a twist2) while the intelligence community consists almost exclusively of Bayesians. Frequentists focus on the probability of events, while Bayesians instead focus on their plausibility. To a frequentist, the proba- bility of an event is the frequency at which it occurs over the long run. That is, as time goes to infinity, the observed frequency converges to some theoretical probability that is an intrinsic characteristic of the sys- tem being observed. Conversely, to a Bayesian, there is no such thing as truth—instead there is only a degree of belief about reality. A frequentist implicitly assumes system stability, while a Bayesian does not. In a sense, a Bayesian explicitly acknowledges the existence of model error. Historically, the two communities chose their diver- gent paths decades ago. In the financial community, Modern Portfolio Theory and the efficient market hypothesis laid the intellectual foundation for the use of stochastic models. And, triggered by advances in technology in the 1970s, the original “rocket scien- VISIONS OF RISK THOUGHT PIECES FROM PRMIA LEADERS 1 For a discussion of current best practice in risk management, refer to Essentials of Financial Risk Management (McGraw Hill, 2006, which is a course text for the Associate PRM certificate). 2 The “with a twist” is implicit acknowledgement that the financial system is unstable. For example, even frequentists limit their use of histori- cal data to a period they consider representative of the future (e.g., the last three years) and acknowledge non-normal behavior such as jump processes. Similarly, factor-based models acknowledge the existence of fundamental drivers, although they employ statistical models informed by history. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 5 INTELLIGENT RISK — MAY 2011 tists” took new advantage of the ready availability of historical pricing data in creating valuation and risk measurement models. Conversely, in the intelligence community, history was both sparse and often a liar—especially in the face of adaptive ene- mies. So, that commu- nity instead focused on structured, causal analy- sis. But in spite of this philosophical rift, best practice in financial risk management is starting to integrate the two approaches. For exam- ple, there is now an increased emphasis on scenario analysis. However, methods used to generate the set of sce- narios can be spotty; they may come from history, or they may be identified on an ad hoc basis. As a result, some scenarios may be included in analyses, while equally plausible scenarios may be ignored.3 For exam- ple, many were caught unaware by the events of 2007- 20084. Yet extreme scenarios are rarely complete sur- prises; rather than spontaneously springing into exis- tence in full flower, they are instead the logical conse- quence of exogenous or endogenous events that trig- ger a cascade of consequences. So, causal approaches are ideally suited for systematic generation of all plau- sible scenarios. As economist David Hendry pointed out, “Causality is a property of nature, not a model.” Failure to consider causality can result in a worldview that, while elegant, may be inaccurate. Fortunately, in the words of Judea Pearl,5 “Put simply, causality has been mathematized.” The language of causality is a natural language for risk management. Examples of causal concepts are influence, confounding, intervention, and explanation. Compared to stochastic models, in causal models elegance may be exchanged for accuracy. The simplifying assumptions that rely on continuous distribu- tions are replaced by dis- crete-time and discrete-out- come modeling that is far more realistic. More importantly, they can also be used to inform risk mitigation decisions. Decision science—which only started to hit its stride in the 80s—is closely related to the fields of artificial intelligence and game theory. Whereas game theory concerns decision made by more than one ‘player’ in a competitive environment, decision sci- ence centers on the decisions made under uncertainty by a single player. It’s concerned with identifying the issues material to a given decision, including the deci- sion maker’s marginal utility function or any other preferences. Some consider decision theory to be a fusion of probability and utility theory. Within the general category of decision-theoretic models are belief (Bayesian) networks. Such networks (which can only be displayed graphically)6 aggregate a set of beliefs about reality. In a true decision-support 3 In the intelligence community, structured analytic methods are used to identify drivers and dependencies in a manner not subject to a number of cognitive biases. 4 For an example of how AIG’s downfall might have been identified as a highly plausible outcome, refer to Extreme Risk Management, Christina Ray, McGraw Hill, 2010 (pages 147-148) 5 Causality: Models, Reasoning, and Inference, Judea Pearl, Cambridge University Press, 2000 6 A belief network consists of the following elements: (a) A set of variables (portrayed as nodes in the graph); (b) a graphical structure that represents the interdependencies between the variables (portrayed as links between the nodes); and (c) a set of