Books > Taking Risks >

Excerpts

Chapter 1

Essence of Risk

A life without adventure is likely to be unsatisfying, but a life in which adventure is allowed to take whatever form it will, is likely to be short.
Bertrand Russell

INTRODUCTION

Some risks such as natural disasters are dramatically obvious, and they affect many people. Hurricanes, volcanoes, and earthquakes can destroy entire communities in minutes. Malaria, smallpox, and sleeping sickness can devastate populations. Other risks are more personal. Crossing busy streets, working in a polluted environment, borrowing money, or failing to contain our anger are everyday risks that most of us face.

Consider getting food from the supermarket. We jump in the car and run a risk of being hit by a careless or drunk driver, even if we are cautious ourselves. Once inside the store we are faced with goods that may have been tampered with or tomorrow may be declared carcinogenic or environmentally harmful. On a more minor level, we may discover that the meat we bought was on sale at a competing supermarket.

If everyday choices present us with a shower of risks, our major choices are immersed in a virtual downpour. Think of what can go wrong in getting married, changing jobs, or taking a trip to a foreign country. In many of these situations, our choices do not just affect us, they affect many other people. Business and government decisions can influence the risks faced by thousands or millions of others.

Risk is a pervasive part of all actions. While eventual death may be certain, every day we engage in activities -- even just crossing the road to mail a letter -- that carry a risk of death. Although death is the ultimate risk, the economic and social risks that we face can be more oppressive. Seemingly secure jobs may disappear in economic hard times. Seemingly stable marriages may shatter for many possible reasons.

Life requires choices; choices require risks. While you can choose to minimize the risks you face, you cannot avoid risks completely. Along with death and taxes, risk is one of the certainties of life.

Proust-like, we may decide to isolate ourselves in the safety of our homes to avoid risks. However, we still may fall down the stairs, have a plane crash into the roof, be electrocuted in the bath, be assaulted by a burglar, be blown away by a tornado, or be downwind of an unsafe nuclear power plant. We may move to an isolated mountain region to escape from the risks of a city's polluting factories, traffic congestion, and racial tension only to expose ourselves to the risk of inadequate emergency medical services in the event of a heart attack. That many risks are hidden and not foreseen may mean greater peace of mind, but it does not reduce our exposure to their possible negative effects.

Primitive man had little control over his environment. The daily activities of acquiring food and shelter were fraught with risk. While modern man has gained some control over his environment and may experience fewer risks in acquiring the basic necessities, a more complex environment has brought new risks. In this century, technology and collective action fill our lives with man-made hazards such as nuclear war or acid rain. Gaining control over some risks has led to different types of risks that may even be more dangerous than the ones that have been mitigated. In general, avoiding one kind of risk will introduce some other risk. By not building nuclear powered energy plants, we run risks of impairing our economic structure by running short of fuel, increasing coal-related disease, and so forth.

Some actions appear to be free of risk. We invest our life savings in a "risk-free" savings account insured by the federal government instead of buying mutual funds or we marry our childhood sweetheart whom we have known for twenty years instead of marrying the seductive stranger. Although these actions seem to have predictable outcomes, they have risks of their own. Several years after putting the money into the savings account, it provides little security because of rampant inflation. Investing in mutual funds would have provided better security against changes in the price level. After marrying the childhood sweetheart and building a family, the marriage crumbles because the partners no longer find each other stimulating.

Thus even apparently riskless actions have risks associated with them due to unforeseen events or changes in perspective. Nonetheless it is often a useful fiction to think of some particular action as riskless so that other risky actions can be judged against it.

We sometimes think that by not taking an action we can avoid risks. You are approached by the chief executive of a multinational firm who wants you to leave your current position to assume the presidency of a small subsidiary. You delay responding so that you can think about the offer and its implications for yourself and your family. A week later another executive is made president of the subsidiary which grows into a multi-million dollar enterprise. By not accepting the risks of changing jobs, you missed an opportunity.

Not only can risks not be totally avoided, but most individuals seek risks in at least some aspects of their lives. Uncertainty about outcomes of virtually all important activities provides the excitement that stimulates as well as creating the anxiety that worries. People engage in hazardous recreational activities such as hang gliding and rock climbing, they play the stock market, and they gamble partly because of the stimulation that accompanies the risk. Success itself increases risks as we discover whether we can handle the new opportunities that become available.

We must face risks in all aspects of our lives and in the many roles we play. The risks we confront as a business executive or community leader are not the same risks we deal with as spouse or parent. We face personal risks that are financial, physiological, medical, social, and so forth. Risks also affect our careers and the organizations that employ us. Most human endeavors bring major risks. To set high goals of success is to run high risks of failure.

We expose ourselves to personal financial risks in several ways. We can live beyond our means when we spend more than our income and wealth can support. We can hold too many of our assets in investments that have a chance of major losses. We can hold too great a share of these assets as highly levered investments that are subject to the control of creditors. In each case loss of credit, loss of an asset, or personal bankruptcy are possible outcomes. During hard economic times unemployment rises and savings dwindle. Individuals with prudent financial investment and expenditure strategies do not suffer the same wide swings in economic well-being as those who sought greater gains with their associated financial risks.

Risks to one's physical health can take many forms -- accidents, disease, violence, heredity, diet, exercise (or its absence), personal habits, and so forth. Some physical risks have only minor consequences such as temporary mild discomfort or inconvenience whereas others have more major consequences such as permanent physical disability, severe suffering, or even death.

If you are an average American male under age 55, the number of days you lose from your life expectancy has been estimated to be more than ten times as high from motor vehicle accidents as from fires (195 days to 14 days). While indulging in regular coffee drinking will only shorten your life by an expected six days, being 30% overweight will cut your life expectancy by more than three and a half years! Even worse than being overweight is smoking more than 20 cigarettes a day; the risks involved (primarily lung cancer) can be expected to shorten your life by almost seven years. The ultimate risk (among normal activities), though, seems to be the risk of remaining single. It has been estimated that being unmarried shortens a man's life expectancy by nine and a half years (Cohen and Lee, 1979)!

Setting high goals can also involve social risks which result in negative consequences if we fail to meet them. The social risk that perhaps affects us most personally is the loss of self-esteem that we sometimes experience in the face of repeated failures. Social risks occur when parent-child relationships within the family undergo change. The normal process of dating exposes the participants to social risks as they experiment with their interpersonal relationship. Joining and participating in any organization or group may lead to social rejection if one fails to be accepted as a contributing member. In its extreme form social rejection can take the form of imprisonment or death when a society imposes its harshest penalties on those convicted of taking antisocial criminal risks.

The choice of one's career has many inherent risks. A career as wife and mother has different financial, physical, and social risks than a career as banker or lawyer. The attainment of a chosen career depends upon satisfying its educational and apprenticeship requirements as well as overcoming any other barriers to entry that have been set up. Once a career is established, will it provide the envisioned opportunities and stimulation that are needed for maintaining interest in the career? Will one's performance on the job lead to success and career advancement? One of the greatest risks is changing careers in midlife when we may be less adaptable to change and when there is less time available for adequate career development before retirement.

Although everyone makes daily decisions involving risks, not all of us make decisions that result in risks for thousands or even millions of people. Managers of business firms or political leaders, however, continually face such risks. Government officials decide whether to commit us to military ventures and trade wars. They decide on traffic speed limits, drug testing requirements, disposal of hazardous wastes, and so forth. Business managers oversee the building of the cars, production of the drugs, and running of the factories that are the objects of such political decisions.

Not all managers, however, are fully accountable for their decisions. While politicians are subject to review by the electorate every few years, many of the decisions are made in a bureaucracy that is well buffered from detailed scrutiny. In business firms job security is not legislated, so bad outcomes for the firms are more likely to be followed by bad outcomes for the responsible managers.

In 1976 John deLorean quit his high-paying job as vice-president of General Motors to start his own automobile company. The costs of the machinery and supplies necessary to compete in the auto industry are enormous. Even though no one had successfully created a major auto company since the 1930s, deLorean managed to obtain the financial backing of the British Government by building his plant in Northern Ireland. Unfortunately, the automobile market collapsed just as deLorean was getting his first cars to dealers. With slow sales, massive amounts of money were required to keep the factory going, and in late 1982 John deLorean was faced with the impending bankruptcy of his company. He was charged with being involved in a major drug deal that, if it had been successful, would have provided the funds to keep his company going. Although there was a videotape of a drug transaction, the jury decided that deLorean was entrapped and he was acquitted. John deLorean clearly fits our image of the entrepreneurial risk taker.

Most business decisions involve tradeoffs that lead to different risks for employees, stockholders, consumers, suppliers, and management. A decision to close an obsolete plant may save the stockholders money that in the long run can be used for creating jobs. In the short run, however, it throws people out of work. The decision to withhold a drug from the market for further testing may minimize the possible hazards of side-effects and prevent future lawsuits, but it denies many others treatment that could help them now.

While many operating decisions made by managers (e.g., which accounting method to use, what inventory level to maintain) seem to have few risks, the strategic decisions are fraught with peril. Consider the Tylenol tragedy in which several bottles of the pain remedy capsules sold in Chicago contained cyanide that caused the death of seven unsuspecting users. Prior to this tragedy Johnson & Johnson, along with many other major pharmaceutical companies, packaged their over-the-counter drugs without any special protective covering to prevent tampering.

There was little time to decide what to do as a nervous public demanded action to stop the mounting death toll and Tylenol sales plummeted. There was little information about what was happening and what was causing it. Was the problem a local Chicago problem or was it more widespread? Were only Tylenol products affected and were only the capsules (but not the solid pills) adulterated? Was the cause unintended and restricted to a malfunction in the production process? Was it the work of a madman, extortionist, or terrorist group that tampered with the bottles randomly in selected retail outlets? Johnson & Johnson had inadequate control of the situation as it examined its options to prevent more deaths, to reassure the public, to better understand the problem, to prevent additional tampering of its products, and to rebuild Tylenol sales.

If Johnson & Johnson did not recall Tylenol capsules in Chicago, more deaths might result. If the recall were not extended to the national market or to other Tylenol products, some other crazed individual or group might try to copy the tragic tampering to capture headlines or to extort financial gain. If recalled capsules were not immediately and systematically tested, management would not know the extent and source of the tampering until other deaths might have occurred.

Facing all these risks simultaneously, the company recalled and tested all Tylenol capsules in Chicago, tested selected lots of capsules from other marketing areas, and thoroughly reviewed their production, testing, and distribution procedures. While these actions were under way, they kept the public informed about the extent of the problem and the steps they had taken to prevent future deaths. Johnson & Johnson assured the public that the tampering had occurred at the retail level and it was restricted to the Chicago area.

Three months after the tragedy had begun Johnson & Johnson announced to the public that Tylenol products were being distributed in a new "tamper-proof" package that had three safety features: a foil seal on the top of the bottle, a plastic ring on the cap, and a cellophane wrap on the box. Despite losing most of its very large 35% share of the capsule pain remedy market due to the tragedy, within a year sales had bounced back to over 25 % of the market.

Rolls Royce is synonymous with prestige and success. The image of the luxury car, however, did not carry over to the management of the company in the 1960s. A company that relied on aircraft engine sales to provide over 80% of their revenue, Rolls Royce could foresee that their current aircraft engine sales would dry up within a decade. To assure survival, top management decided that they had to obtain a contract for the engines on one of the new generation of wide-bodied jets. After losing out on the Boeing 747, Rolls Royce made a special effort to make a deal with Lockheed. The engines that they proposed supplying for the Lockheed L1011 required a higher thrust than any aircraft engine that had previously been developed. In addition, the engines were to be extremely quiet and durable. New materials would be used and many other technological advances were promised. The contract involved delivery of engines at a fixed cost three years ahead. Considerable late delivery penalties were accepted. The financial commitment amounted to over 90% of Rolls Royce's total assets. Thus the technological and financial risks were enormous. Management believed they were justified in light of the potential gains and the consequences if the risks were not taken. Unfortunately, these difficulties escalated and Rolls Royce declared bankruptcy three years later in February 1971.

Does this story imply that large risks should not be taken? Not at all. Several years later, Boeing was making aircraft commitments (for the 757 and 767) that involved three times as much as their total assets at the time. Boeing not only survived, they established a position of strong technological leadership. Although the recession of the early 1980s hit them hard, they were in no worse shape than other aircraft companies.

RISKY SITUATIONS

Risk Components: Exposure to a Chance of Loss

The main definition of the verb "risk" in the Oxford English Dictionary is "to expose to the chance of injury or loss." Early references date back to the seventeenth century. The origin is thought to be Italian (risco) but is uncertain. It is worthwhile to reflect on the implications of various aspects of the definition. First, it is necessary that there be a potential loss of some amount (we will use "loss" as a general expression to include "injury"). Second, there must be a chance of loss. A sure loss is not a risk. Third, the notion "to expose" means that the decision maker can take actions that can increase (or decrease) the magnitude or chance of loss. Therefore "to risk" implies the availability of a choice. This exposure may be to the person making the risky decision or to other persons or groups in the environment.

The second definition in the dictionary, "to venture upon," suggests even more of an orientation toward action than the first definition. To the extent that a person has some influence on the state of affairs, he takes risks in order to bring about a more preferred outcome. In 1759 Samuel Johnson was explicit about this aspect when he advocated that we "risk the certainty of little for the chance of much."

Note on the Definition of Risk

There is uncertainty about the price of gold at the end of this year but we would not call it a risky situation for you unless there were some consequences to you of the different price levels. Correspondingly, if you knew that a stock that you owned and could not sell was sure to lose half its value over the next month, this situation would not be risky because there was no uncertainty. Hence uncertainty is a necessary condition for riskiness but it is not sufficient. Similarly, potential loss is a necessary condition for riskiness but it is not sufficient. Together, however, uncertainty and potential loss are sufficient for a situation to be risky.

In general "loss" is a relative loss. A particular payoff is a "loss" if it is considered worse than some particular reference level of payoff. The reference level may be a zero payoff, the current status quo, a target or aspiration level, or the best payoff available in a situation. Hence even a payoff distribution consisting entirely of monetary gains is considered risky when each outcome is compared to the best possible outcome. While losses are commonly thought of in monetary terms, they may be anything that is of concern to a particular decision maker.

Two main forms of potential loss should be considered: (1) an outcome that will make us worse off than some reference status quo position, or (2) an outcome that is not as good as some other outcome that might have been obtained. The first case is easily perceived as a real loss; the second case, an opportunity loss, may not be perceived as easily. Opportunity losses can sometimes turn apparently risk-free situations into risky ones when unforeseen events occur. We shall use the terminology magnitude of loss to refer to either of these forms of loss.

For simplicity, in the discussion above we did not distinguish between single events and multiple events, or between single types of loss and multiple types of loss. If outcomes depend on only one event, say weather, the situation is less risky than if they depend on many events (e.g., weather, economic conditions, demographic trends, actions of competitors). Similarly, situations in which choices entail possible losses on only a single dimension, say financial, are less risky that those involving losses on many dimensions (e.g., financial, health, social, career).

In summary, then, there are three components of risk -- the magnitude of loss, the chance of loss, and the exposure to loss. To reduce riskiness, it is necessary to reduce at least one of these components. Presumably if all of these elements were trivial, then we would hardly call the situation risky. The degree of risk can be thought of as being directly proportional to the chances and size of the loss and to the degree of exposure of the decision maker to the chance of loss.

Note on the Risk Components

In some situations risky alternatives have a formalized assessment of the magnitude and chances of loss. For example, a bet on a single number at an American roulette table has a 97.4% chance (i.e., 37/38) of losing the amount bet. Similarly for other casino games, lottery tickets, and the like, we can calculate the chances of loss directly. The amount of loss is, of course, determined by the size of our bet. We also can calculate the amount of the potential gain.

In other gambling situations, such as the race track, while we know our potential loss, we have no objective estimate of the chances of loss or of the potential gain. Immediately after the betting on a race closes, and before the race is run, we can compute our potential gain, but we still must rely on our subjective estimate of the chances. Presumably the reason we bet is that our assessment of the chances of our horse winning was sufficiently higher than the assessment made by other bettors.

Adopting a Bayesian point of view, we assert that in virtually all situations, a person has some information from which to estimate the chances of the potential gains and losses if they are not given. Hence it does not appear to be a gross oversimplification to consider risky actions that have objectively specified chances of winning and losing. Even when objectively specified chances are given in a risky situation, the decision maker may modify these quantities because he perceives the situation differently from its objective statement.

Basic Risk Paradigm

When research and development offer the possibility of improving an existing product, a manager must decide whether to stick with the existing product or to make the improvements. If the manager stays with the current product, the profits will remain at their current levels. Suppose, for example, that an existing product has a profit level of $20 million. If the product is modified, the profits will depend upon the acceptance by consumers. Suppose market research indicates that there is a 75% chance of strong acceptance, resulting in an increase in profits of $4 million, but there is a 25% chance of weak acceptance resulting in a drop in profits of $12 million. Should the manager go ahead?

This decision illustrates the basic elements in risky situations. There are two actions. One action (called the "sure action") is the status quo while the other action (called the "risky action") has two possible outcomes, a gain or a loss. If we knew that the gain outcome was going to occur, we would select the risky action; if we knew that the loss outcome was going to occur, we would select the sure action. The problem is that we do not know for sure which of these two outcomes will occur. The outcome that occurs depends on an uncertain event for which we have only probabilistic knowledge. This prototypical risky situation will be called the "Basic Risk Paradigm." It provides a foundation for studying risk.

It is helpful to visualize the Basic Risk Paradigm in the form of a decision tree. Figure 1.1a shows the decision tree for the decision about modifying the product. The two actions, "modify" and "do not modify" are shown coming out of a square box, which depicts a choice that must be made by the decision maker. The action "do not modify" leads to the current profit of $20 million. The outcome of the action "modify" is dependent on whether consumers like the new product. The event that they do, labeled as "strong consumer acceptance," leads to a rise in profits to $24 million. If they do not, labeled as "weak consumer acceptance," profits will fall to $8 million. Since the outcome depends on an uncertain event (consumer acceptance), the two outcomes are shown coming out of a circle. The chances, 75% for "strong consumer acceptance" and 25% for "weak consumer acceptance" are noted on the corresponding branches. Thus a square box signifies a choice that the decision maker makes, while the circle signifies an event that is outside the control of the decision maker.

Note that the expected profit for the risky action is $20 million. That is, a 75% chance of a $24 million profit and a 25% chance of an $8 million profit would yield an expected profit of $20 million if the uncertain outcome were repeated a large number of times. Since one action yields a sure profit of $20 million, while the other action yields an expected profit of $20 million, this average return cannot be a basis for choosing between the actions. A person's preference or aversion to the "modify" action must therefore depend on his preference for or aversion to the risks that this action entails.

The Basic Risk Paradigm is depicted in an abstract form in Figure 1.1b. For most purposes it will be embellished, but the basic structure will always be present. Although most risky situations will have more alternatives, more uncertain events, and more outcomes, the basic form captures the central elements. The Basic Risk Paradigm will provide the foundation for assessing risk propensity in this book.

Note on the Basic Risk Paradigm

The expected value (i.e., arithmetic mean) of a risky action can be calculated as (1 - p) x G + p x L where G is the gain outcome, L is the loss outcome, p is the chance of loss, and 1 - p is the chance of gain. We will sometimes refer to the expected value of a risky action as its expected payoff or average payoff.

While the Basic Risk Paradigm is, as the name suggests, the most basic situation for studying risk, in most instances it will be necessary to elaborate upon it. These extensions can take many forms:

1. The sure action does not necessarily have to be the status quo. It can be any action with a sure outcome between the gain and loss outcomes of the risky action.
2. Both actions may be risky, although one of them may be riskier than the other.
3. There may be more than two actions.
4. The risky actions may have more than two possible outcomes.

These extensions, as well as other ones to be discussed, can provide a richer context for studying risk than the Basic Risk Paradigm itself. Even in these cases it is insightful to recognize the basic structure.

When potential loss is higher, the risks are greater. Introducing the modified product if a $60 million loss can occur is riskier than if a $12 million loss can occur. A $20,000 investment in commodity futures is riskier than a $2,000 investment. Opening a business in a politically unstable foreign country poses a greater magnitude of potential loss than operating the business in North America.

When the chances of loss are higher, the risks are greater. Introducing the modified product when there is a 50% chance of weak market acceptance is riskier than if there is a 25% chance of weak market acceptance. Investing in commodity futures involves a greater chance of loss than investing in government bonds. In stable countries we have few worries about our plant being looted and burned; in unstable countries, the chance of losing our investment is higher.

When the exposure to chance of loss is higher, the risks are greater. Even if the chances and magnitude of loss cannot be restricted, we may be able to affect the risk by reducing our exposure. Exposure to loss for three distinct entities should be considered: (1) the individual who makes the decision, (2) the decision maker's immediate social unit (usually the family or the firm), and (3) society at large. In the modified product example, outcomes to the firm were discussed because we were primarily concerned with the firm's exposure to risk. The uncertain market acceptance of a modified product could also be translated into potential losses for the manager (e.g., loss of reputation, incentive bonus, or job) or consumers (e.g., unavailability of preferred standard product).

In Ford Motor Company's decision on the location of the gas tank in the Ford Pinto, the exposure to risk was very different for the executives who approved the design, Ford which had to pay the lawsuits, and the consumers who were injured or died. Sharing or spreading risk is the most common way to reduce exposure to the individual, family, or firm. This method does not reduce the total loss that is incurred; it just redistributes the exposure to loss among a number of individuals.

Risk Determinants:
Lack of Control, Lack of Information, Lack of Time

Inherent in all risky situations are three identifiable determinants: lack of control, lack of information, and lack of time. If we had complete control over the situation, we could determine the best outcome and there would be no risk. If we had complete information about which event would occur, we could select the best alternative based on this knowledge and again there would be no risk. If we had unlimited time in which to decide which alternative to choose, we could wait until the outcome of the uncertain event was resolved and then choose the best alternative after the fact. This scenario also involves no risk.

Events are uncontrollable for a variety of reasons. The pure uncontrollable events are determined by nature -- earthquakes, oil formations, weather, and so forth. Next come events determined by other people, including large aggregates of people, such as the actions of competitors, demand for a firm's product, election results, etc. While there may be a distinction between the controllability of next year's weather versus next year's economic climate, they both seem equally uncontrollable to most of us. When the events are determined by specific people, for example the price of a competing product, we must consider the unpredictability of human behavior.

Another reason for lack of control is lack of suitable resources. You could control the price of silver next month if you had the resources of the Hunt brothers (and lack of control exerted by the SEC and others). You could even affect the weather with suitable amounts of chemicals and airplanes for seeding. Control requires not only potentially controllable events and the appropriate resources, it also requires suitable opportunities to intervene. In order to control a risky situation, we need information on which to base our control actions. In order to know how to control something by pushing the right button, we have to know which button does what. Since control requires the deployment of resources, we have to know the effect of using our resources in one fa