Risk and Decision Making Flashcards
Risk
several possible outcomes but the probabilities can be quantified
Uncertainty
possible outcomes and probabilities are unknown
Techniques for handling uncertainty
~ maximum payback period
~ increasing discount rate
~ assessing best and worst to get a rnage of outcome
~ sensitivity analysis - measure the margin of safety
Sensitivity analysis
= 100% x NPV/ PV of cash flow subjects uncertainty
Sensitivity analysis: Advantages (3)
~ facilitate subjective judgements
~ identifies areas critical to success
~ straightforward
Sensitivity analysis: Disadvantages (4)
~ only on factor can be analysed
~ assumes changes to variables occur independently (they always link)
~ identifies how far a variable needs to change, but not the probability
~ doesn’t point to the correct decisions, just provides info
Linear regression: Advantages
~ simple to use
~ easy to explain
~ predict impact of expanding variables
Linear regressions: Disadvantages
~ not always a linear relationshisp betwn variables
~ doesn’t consider multiple variables
~ data collected may be inaccurate
Types of predictive analytics
~ linear regression
~ decision trees
~ simulation
~ prescriptive analysis
Linear regression
quantifies relationship between a dependent and independent variable
Decision trees: Advantages
~ easy to explain and use
~ analyse diff outcomes baced on multiple variables
Decision tree: Limitations
~ variables have to be simple
~ large decision trees can be difficult to understand
Simulation
allows the effect of more than one variable changin at the same time to be assessed
e.g. Monte Carlo simulation
Simulation: Disadvantages
~ provides info, not decision
~ expensive
~ relies on assumptions that might not be true (relating to probability distributions and relationships
Simulation: Disadvantages
~ provides info, not decision
~ expensive
~ relies on assumptions that might not be true (relating to probability distributions and relationships
Monte Carlo Simulation
the use of random numbers and probability statistics to investigate problems
Simulation results:
Higher average NPV = more risky
Lower average NPV = less risky
~ decision dependent on attitude risk
Prescriptive analytics
combining statistical tools used in predictive analysis with AI and alogrithms
Prescriptive analytics: Advantages
can identify the optimum decision by incorporating multiple variables
Prescriptive analytics: Disadvantages
~ creating models is complex and requires speicifc skills
~ relaibility of model depends on reliability fo data + past-present relationships
Expected value: Advantages
~ easy to understand
Expected value: Disadvantages
~ probabilities can be difficult to estimate
~ average may not correspond to any possible outcomec
~ gives no indication of the spread of possible events