Risk Management BUSS 5292 代写

1 Risk Management BUSS 5292 Topic 1 History of risk management thinking 1 Course Facilitator: Kesten Green Study Period 1, 2016 Today’s session • Introductions • About the Risk Management course • Administrative matters • History of risk management thinking 2 Risk Management - BUSS 5292 Course Objective • Introduce risk management concepts and tools, that you can use to help you to… • Structure and solve managerial decision problems. 3 2 Your objectives for the course • ? 4 How to do well in Risk Management • Make use of the Course Site • Course Outline provides key administrative information • Use the News Forum for general communications • Video clips and readings • Study Guide is not used • Add your photo and profile • Read topic materials before lecture • For assignments: Use good sources, and your own words • Think about applications • Ask clarification questions 5 Administration: Course Outline & LearnOnline Site • Refer to the Course Outline and course internet site (LearnOnline) for information on the administration of the course, including… – Contacts – Overview – Dates – Resources – Assessment – News – Other stuff… 6 3 Maths 7 Why lectures? • To help you understand the course material • To help you relate the material to practical problems • To test your understanding • To get tips on how to get good marks • To make useful contacts • ? 8 What do we mean by risk? q “If the uncertainty associated with an event can be quantified on the basis of empirical observations or causal knowledge (physical design), the uncertainty is called risk. q Relative frequencies and probabilities are ways to express risks. q Contrary to everyday use of the term, a risk need not be associated with harm; it can refer to a positive, neutral, or negative event. q The classical distinction between known risks (‘risk’) and unknown risks (‘uncertainty’) is attributed to the economist Frank Knight.” Gerd Gigerenzer, Glossary of “Risk savvy” 9 4 The risk management problem: Old(ish) version “We plan, God laughs” Yiddish proverb 10 “Nothing is certain but death and taxes.” From Benjamin Franklin’s 1789 letter to Jean-Baptiste Leroy 11 The risk management problem: New(er) version (Franklin’s Law) § Ignorance § Mistakes § Deceit § Conflict § Forecasting failure § Forecast uncertainty § ... 12 Why so little certainty? 5 13 Summary history of risk management thinking Date Key ideas 3200 B.C. Certainty, given signs from gods (Asipu) 400 B.C. Behaviour now risks soul in afterlife (Plato) 400 A.D. Dominance principle, 2 x 2 matrix (Arnobius) 1657 A.D. Probability theory introduced (Pascal) … development of mathematical probability theory, and concepts of causality, systematic observation, and the experimental method (Science)... 1792 A.D. Modern quantitative risk analysis (Laplace) Recorded history indicates risk management was an important service at least as far back as the 4 th Millenium  B . C . At the begining of history and in the absence of the scientific method... What would happen was assumed to be determined by the whim of gods. People nevertheless wanted advice to help make decisions about risky or uncertain situations... Demand was met with supply in the form of risk management consultants (a . k . a . priests or seers), who developed rituals they claimed divined the plans of the gods. 14 Risk management at the dawn of history Demand for seers persists to this day • In 1978, Scott Armstrong summarized evidence showing that in complex and uncertain situations, expert forecasts* were no more accurate than forecasts by people with little expertise. • People resisted this evidence, hence the “Seer-sucker Theory.” *Unaided by evidence-based principles on how to forecast 15 6Risk Management BUSS 5292 代写 Forecasts by experts • Unaided experts’ forecasts are of no value when the situation is complex and uncertain. • Does not help when judgments are expressed in complex mathematics. • Does not help when the experts get more data 16 Unaided expert forecasts: recent evidence • Tetlock’s Expert Political Judgment (2005) also found experts’ forecasts lacked value: – evaluated forecasts from 284 experts in politics and economics – who made about 82,361 forecasts – over two decades Possible states God exists God doesn’t exist Accept BIG win  Small loss Christianity Alternative decisions Remain pagan  BIG loss Small win 18 Dominance principle (c400  A . D .) 7 19 “Pascal’s Wager”: expected value, & utility maximisation Possible states God exists God doesn’t exist (p=.5? .00001?) Accept BIG win  Small loss Christianity  ∞? Alternative decisions Remain pagan  BIG loss Small win E (“wager for God”) = p x ∞ + (1-p) x “small loss” = ∞ 20 Risk Management BUSS 5292 代写 More on Pascal’s contribution: The division problem or problem of points Problem of dividing a prize when a series of games is interrupted before agreed winning total is reached (e.g. first to win 7 games) assuming equal chances of winning. Equivalent to determining the ratio of the probabilities that each player will win the remaining games he needs to win before the other does… applied to the value of the prize. In other words, each player gets his calculated expected value of the prize. 21 Key developments in math-prob. theory and scientific causality Date Key ideas 1657 A.D. Probability theory introduced (Pascal) 1662 A.D. Life expectancy tables (Graunt) 1692 A.D. Causal probabilities calculable (Arbuthnot) 1693 A.D. Life expectancy table and annuities (Halley) 1792 A.D. Modern quantitative risk analysis (Laplace) [Analysed the probability of death having had a smallpox vaccination and without having had one.] 8 22 Graunt’s life expectancy table: Empirical probabilities Births and deaths data collected c1554 at behest of London Merchants by Chancellor Thomas Cromwell Life expectancy table published in 1661 from an analysis of the data by “obscure haberdasher” John Graunt… and so pioneered modern statistics, including PDFs. (Table and chart from Thompson’s (Rice U) “John Graunt’s Life Table”) Modern use of experimentation in the 16 th & 17 th Centuries e.g. Galileo (1612) Bodies that stay atop water, or move in it Key changes: 1. Observation to correct theory (vs to support argument or established theory) 2. Experimentation, or active observation whereby the situation of interest is manipulated to see what happens 3. Control extraneous influences that might bias observations and lead to erroneous conclusions. 23 Developments in causal knowledge: Experimentation A causal relationship exists if the effect… 1. Follows the cause 2. Is connected to the cause in some way 3. Cannot be plausibly explained in any way other than by the cause* From J. S. Mill’s formulation Modern scientific experimentation is the best way to determine whether the three necessary conditions of a causal relationship exist. *Importance of testing multiple reasonable hypotheses. 24 Developments in causal knowledge: Establishing cause and effect 9 Why experimental findings? Meta-analyses of experimental evidence from tests of multiple reasonable hypotheses is the basis for scientific advances (Chamberlin, 1890) Infeasible to identify causality from analyses of nonexperimental data in uncertain complex situations. Illusions in regression analysis Directions of effects from nonexperimental studies often differ from those from experimental studies. Armstrong & Patnaik (2009). 25 26 Risk Management BUSS 5292 代写 Multiple hypotheses and knowledge Knowledge advances when multiple hypotheses are tested, especially if hypotheses challenge accepted wisdom, e.g.: • Anti-inflammatory drugs harm head injury patients • Duodenal ulcers are caused by bacteria, not spicy food • Market-share objectives harm profits. • Minimum-wage laws harm low-skilled workers • Regulation harms consumers • Pre-announced satisfaction surveys harm satisfaction

Risk Management BUSS 5292 代写