Decision analysis lecture 1
Some definitions
Decision-making can be regarded as the cognitive process resulting in the selection of an action
among several alternative possibilities. It is the study of identifying and choosing alternatives based
on the values and preferences of the decision maker.
Decision analysis (DA) is the discipline to address important decisions in a formal manner. Decision
analysis includes many procedures, methods and tools for prescribing a recommended course of
action by applying the maximum expected utility action axiom.
Decision making is a way of facing opportunities, challenges and uncertainties of life.
Simple decisions; complete insight in outcomes, exception
Hard decisions; complex due to conflict and uncertainty, no obvious and single solution
Most people dislike to be confronted with hard decisions which often result in bad decision making.
Ingredients to resolve decisions:
1. Clear definition of decision problem
2. Specification of objectives
3. Definition of alternatives (your decision can be no better than your best alternative)
4. Evaluation of possible consequences
5. Balancing trade-offs
6. Clarification uncertainties (uncertainty of the outcomes reflecting your inner beliefs)
7. Consideration risk tolerance level
Personal judgements (subjective evaluations) about uncertainties and outcome-values are important
inputs within the decision making process.
A good decision does not guarantee a good outcome, because a good decision is not the same as
luck. A good decision is consistent with the decision makers believes about uncertainties and the
decision makers’ preferences for possible risky consequences.
Risk and uncertainty
The definition of risk and uncertainty vary in literature. Uncertainty refers to imperfect knowledge. It
is value-free; you do not have a certain preference in outcome possibilities.
A risk refers to uncertain consequences, particularly to unfavourable consequences. It is not value-
free. A risk is the combined effect of the probability of an outcome and the consequences (risk = P *
consequence). To take a risk is to expose yourself to a significant chance of injury, harm or loss.
The general tendency is to perceive risks in negative terms, but without a risk, you cannot have a
gain.
Risk management to avoid high losses and to maximise opportunities.
Type and sources of risk:
o Business risk (risks facing the firm independently of the way in which it is financed)
دProduction risk (e.g. unpredictable nature of weather, uncertainty about
performance of crops or livestock due to for example diseases)
دPrice or market risk (e.g. demand fluctuation, time preference, exchange rates)
دInstitutional risk
political risks; unfavourable policy changes
sovereign risks; caused by actions of foreign governments
contractual risks; risks inherent in the dealings between business partners
and other trading organisations
دHuman or personal risk (e.g. key persons, knowledge/expertise)
, o Financial risk (has no influence on the business risks and result from the method of financing
the firm)
Most people dislike risk; risk aversion (=the willingness to forgo some expected return for a reduction
in risk).
Risk management to identify, analyse, assess, treat and monitor risk to maximise opportunities. So
to maximise the utility (wealth) and not to minimise risk. After all, decision makers are prepared to
take some risks if they expect to earn profit; they try to avoid or diminish those risks judged too great
to be born. Risk management is a continuous and adaptive process, consisting of seven cyclical steps.
1. Establish the context: setting the scene and identifying the parameters within which risky
choice is to be analysed. And establish the relationship between the person doing the
analysis and the DM (because decision analysis is fundamentally a personal approach to
rationalising risky choice).
2. Identify important risky decision problems: it usually considers what can go wrong, how likely
that eventuality is and how serious the consequences might be. The risky decision for
analysis may be either how to deal with threats to the business or whether to take up some
new but risky investment opportunities that improve business performance. So this step is
mainly about listening and prioritising.
3. Structure the problem: identifies stakeholders (a.o.)
4. Analyse options and consequences: identify choice options and uncertain states.
5. Evaluate and decide: evaluating the consequences of each possible action given each
possible state.
6. Implement and manage: simply doing what has been decided upon.
7. Monitor and review: necessary to establish whether the plans decided upon are working and
to identify aspects where further decisions need to be made. If adjustments are needed, the
steps in risk management may have to be revisited to deal with the problem appropriately.
Decision analysis relates to step 4 and 5. Decision analysis may be defined as the philosophy, theory,
methods and practices necessary to systematically address important risky decisions. It includes
methods and tools for identifying, representing and assessing important aspects of a risky decision,
leading to a recommended course of action consistent with careful consideration of the possible
consequences of the alternative choices, the associated probabilities of those consequences, and the
relative preference for possible outcomes. It is the prescriptive theory of choice.
, Assessing beliefs and preferences
It is assumed that important risky decisions are best dealt with by breaking the choice down into
separate judgements about the uncertainty affecting the (possible) consequences of the decision and
about the preferences of the DM for different consequences. Combine the results of the parts to
determine the ‘best’ choice for a DM.
» In a risky world (in reality) there is no way of knowing ahead of time what will be the correct
choice to make. In most cases, we even do not know all the possible consequences of
decisions.
» The approach suggested is based on an assumption about how most people will wish to
approach the task of making important risky decisions. Decision analysis is a prescriptive
model of choice. This is a logical derivation from some axioms or supposedly self-evident
truths about how a rational person would wish to act in making important risky decisions.
» Not everyone will want to address every important decision in this way.
» Many people find emphasis on subjectivity in decision analysis disturbing if not
unacceptable. But the aim in decision analysis is to make the analysis as objective as possible,
and ‘objectivity in science is a myth, in life an impossibility and in decision making an
irrelevance’.
A decision tree is a diagram that shows the decisions and events of the problem and their
chronological relationships. The time scale runs from left to right; usually with the first choice at the
extreme left and the eventual consequences at the extreme right.
A square reflects a decision point (decision node). At such a point there are branches
‘sprouting’ form this node representing alternative choices. These are decisions that you can
take.
o A circle reflects an uncertain process (chance node), so the outcome is determined by chance
or luck (and thus unknown beforehand). At such a point there are branches ‘sprouting’ from
this node representing alternative events or states.
The values given at the right-hand end of each act-event sequence are the payoffs or consequences
for the DM.
Analysing the decision tree using certainty equivalents
A decision with risky consequences is called a risky prospect. For every decision maker (DM) faced
with a decision with risky payoffs, there is a sum of money for certain that would make the DM
indifferent in facing the risky prospect or accepting the sure sum = certainty equivalent (CE). This
sure sum is the lowest sure price for which the DM would be willing to sell a desirable risky prospect,
or the highest sure payment the DM would make to avoid (get rid of) an undesirable risky prospect.
Certainty equivalent reflects a guaranteed return that makes a DM indifferent in taking a chance on
an uncertain return.
CEs vary between people, because people differ in attitudes towards risk and they have subjective
probabilities for uncertain events.
So the question is for what amount of money you are willing to get rid of the risky outcome. In other
words; to insure yourself against this risky outcome.
Sure sum is equivalent to the risky prospect in the judgement of DM (covers beliefs and preferences).
The certainty equivalent replaces the risky prospect and then the decision can be made.
Some definitions
Decision-making can be regarded as the cognitive process resulting in the selection of an action
among several alternative possibilities. It is the study of identifying and choosing alternatives based
on the values and preferences of the decision maker.
Decision analysis (DA) is the discipline to address important decisions in a formal manner. Decision
analysis includes many procedures, methods and tools for prescribing a recommended course of
action by applying the maximum expected utility action axiom.
Decision making is a way of facing opportunities, challenges and uncertainties of life.
Simple decisions; complete insight in outcomes, exception
Hard decisions; complex due to conflict and uncertainty, no obvious and single solution
Most people dislike to be confronted with hard decisions which often result in bad decision making.
Ingredients to resolve decisions:
1. Clear definition of decision problem
2. Specification of objectives
3. Definition of alternatives (your decision can be no better than your best alternative)
4. Evaluation of possible consequences
5. Balancing trade-offs
6. Clarification uncertainties (uncertainty of the outcomes reflecting your inner beliefs)
7. Consideration risk tolerance level
Personal judgements (subjective evaluations) about uncertainties and outcome-values are important
inputs within the decision making process.
A good decision does not guarantee a good outcome, because a good decision is not the same as
luck. A good decision is consistent with the decision makers believes about uncertainties and the
decision makers’ preferences for possible risky consequences.
Risk and uncertainty
The definition of risk and uncertainty vary in literature. Uncertainty refers to imperfect knowledge. It
is value-free; you do not have a certain preference in outcome possibilities.
A risk refers to uncertain consequences, particularly to unfavourable consequences. It is not value-
free. A risk is the combined effect of the probability of an outcome and the consequences (risk = P *
consequence). To take a risk is to expose yourself to a significant chance of injury, harm or loss.
The general tendency is to perceive risks in negative terms, but without a risk, you cannot have a
gain.
Risk management to avoid high losses and to maximise opportunities.
Type and sources of risk:
o Business risk (risks facing the firm independently of the way in which it is financed)
دProduction risk (e.g. unpredictable nature of weather, uncertainty about
performance of crops or livestock due to for example diseases)
دPrice or market risk (e.g. demand fluctuation, time preference, exchange rates)
دInstitutional risk
political risks; unfavourable policy changes
sovereign risks; caused by actions of foreign governments
contractual risks; risks inherent in the dealings between business partners
and other trading organisations
دHuman or personal risk (e.g. key persons, knowledge/expertise)
, o Financial risk (has no influence on the business risks and result from the method of financing
the firm)
Most people dislike risk; risk aversion (=the willingness to forgo some expected return for a reduction
in risk).
Risk management to identify, analyse, assess, treat and monitor risk to maximise opportunities. So
to maximise the utility (wealth) and not to minimise risk. After all, decision makers are prepared to
take some risks if they expect to earn profit; they try to avoid or diminish those risks judged too great
to be born. Risk management is a continuous and adaptive process, consisting of seven cyclical steps.
1. Establish the context: setting the scene and identifying the parameters within which risky
choice is to be analysed. And establish the relationship between the person doing the
analysis and the DM (because decision analysis is fundamentally a personal approach to
rationalising risky choice).
2. Identify important risky decision problems: it usually considers what can go wrong, how likely
that eventuality is and how serious the consequences might be. The risky decision for
analysis may be either how to deal with threats to the business or whether to take up some
new but risky investment opportunities that improve business performance. So this step is
mainly about listening and prioritising.
3. Structure the problem: identifies stakeholders (a.o.)
4. Analyse options and consequences: identify choice options and uncertain states.
5. Evaluate and decide: evaluating the consequences of each possible action given each
possible state.
6. Implement and manage: simply doing what has been decided upon.
7. Monitor and review: necessary to establish whether the plans decided upon are working and
to identify aspects where further decisions need to be made. If adjustments are needed, the
steps in risk management may have to be revisited to deal with the problem appropriately.
Decision analysis relates to step 4 and 5. Decision analysis may be defined as the philosophy, theory,
methods and practices necessary to systematically address important risky decisions. It includes
methods and tools for identifying, representing and assessing important aspects of a risky decision,
leading to a recommended course of action consistent with careful consideration of the possible
consequences of the alternative choices, the associated probabilities of those consequences, and the
relative preference for possible outcomes. It is the prescriptive theory of choice.
, Assessing beliefs and preferences
It is assumed that important risky decisions are best dealt with by breaking the choice down into
separate judgements about the uncertainty affecting the (possible) consequences of the decision and
about the preferences of the DM for different consequences. Combine the results of the parts to
determine the ‘best’ choice for a DM.
» In a risky world (in reality) there is no way of knowing ahead of time what will be the correct
choice to make. In most cases, we even do not know all the possible consequences of
decisions.
» The approach suggested is based on an assumption about how most people will wish to
approach the task of making important risky decisions. Decision analysis is a prescriptive
model of choice. This is a logical derivation from some axioms or supposedly self-evident
truths about how a rational person would wish to act in making important risky decisions.
» Not everyone will want to address every important decision in this way.
» Many people find emphasis on subjectivity in decision analysis disturbing if not
unacceptable. But the aim in decision analysis is to make the analysis as objective as possible,
and ‘objectivity in science is a myth, in life an impossibility and in decision making an
irrelevance’.
A decision tree is a diagram that shows the decisions and events of the problem and their
chronological relationships. The time scale runs from left to right; usually with the first choice at the
extreme left and the eventual consequences at the extreme right.
A square reflects a decision point (decision node). At such a point there are branches
‘sprouting’ form this node representing alternative choices. These are decisions that you can
take.
o A circle reflects an uncertain process (chance node), so the outcome is determined by chance
or luck (and thus unknown beforehand). At such a point there are branches ‘sprouting’ from
this node representing alternative events or states.
The values given at the right-hand end of each act-event sequence are the payoffs or consequences
for the DM.
Analysing the decision tree using certainty equivalents
A decision with risky consequences is called a risky prospect. For every decision maker (DM) faced
with a decision with risky payoffs, there is a sum of money for certain that would make the DM
indifferent in facing the risky prospect or accepting the sure sum = certainty equivalent (CE). This
sure sum is the lowest sure price for which the DM would be willing to sell a desirable risky prospect,
or the highest sure payment the DM would make to avoid (get rid of) an undesirable risky prospect.
Certainty equivalent reflects a guaranteed return that makes a DM indifferent in taking a chance on
an uncertain return.
CEs vary between people, because people differ in attitudes towards risk and they have subjective
probabilities for uncertain events.
So the question is for what amount of money you are willing to get rid of the risky outcome. In other
words; to insure yourself against this risky outcome.
Sure sum is equivalent to the risky prospect in the judgement of DM (covers beliefs and preferences).
The certainty equivalent replaces the risky prospect and then the decision can be made.
