Simulation

Question 1

Risk Management

Accepted Answer

Every business decision entails risk– Incorrect/missing information– Uncertainty of future outcomes (consequence) Risk Analysis– Identifying risk factors– Analysing impact & likelihood of individual risk factors Risk Management– Taking risk into account in decision-making– Contingency planning: anticipating and coping with risk factors– Risk reduction (diversification/hedging)

Question 2

Step 1: Structure the problem

Accepted Answer

1. What is the objective? (cannot be influences and needs actions) Maximize profit often, but generic Minimize risk Specific: Can identify that we are looking at a toy company Objective cannot be influence and needs Actions 2. What are the available actions? (Action serves the Objective) Buy the Airplane Not Buy the Airplane Harvest before storm or after? Rent SuperDoppler? Buy Botrytis spores? 3. What consequences (risks) should we watch out for? Economy Competition Mix of Chatered Flights vs. Scheduled flights Demand & Utilization Rate Oil Price Useful Life i.e. Demand, Mxix, Total Demand, Facity Risks, Delays,

Question 3

Common mistakes

Accepted Answer

pessmimistic - optimistic assumption depend on which is more profitable never hardcode numbers, always have input parameters in the cell Base= Opportunity values

Question 4

Analysis Guideline

Accepted Answer

Ignore uncertainty?– single-point forecasts: always wrong!– decisions based on them: dangerous! Analysis I: Scenario Analysis –> Range of Results Scenario Analysis very quick and easy to do “Scenarios are discrete, internally consistent views of how the world will lookin the future, which can be selected to bound the possible range of outcomesthat might occur.” (Michael Porter, Competitive Strategy) “Scenarios should present testing conditions for the business. The future willof course be different from all of these views/scenarios, but if the company isprepared to cope with any of them, it will be able to cope with the realworld.” Three Scenarios– Pessimistic  $ -936,657– Base Case  $ -64,782– Optimistic  $ 1,441,798 – Result: Range of Net Present Value will be in the range [$-936K - $1,441K] Analysis II: Sensitivity– observe changes in profit due to changes in assumptions– identify main uncertainty drivers --> Tornado chart With the Tornado Chart we identified that Load Factor is most sensitive Data Table - Two Way with color chart --> NPV vs. Load Factor of Scheduled Flights (Average people per flight for Xtra plane) Shape - Straight Line / Linear Slope of this Line - How much more can we make by one more person per flight (only one extra person over a year makes 250K $) Total Range is the most important Break-Even Point --> Point at which we have a positive NPV (67%, average 67 ppl on plane) Conclusion: The NPV is quite sensitive to the load By increased the Load Factor by x% then Analysis III: Monte-Carlo Simulation – evaluate possible outcomes– determine expected result, range of results, probability of results (e.g. probability of break-even), downside risk, etc.

Question 5

Step 5: Tornado Diagram

Accepted Answer

When there are many risk factors, can rank them in order of importance using a tornado diagram Shows impact of each parameter, sorted in magnitude, i.e. widest range: Ticket Price Load Factor of schedules flights Charter price/hours Popowrtion of scheduled flights Operating cost/hour --> Allows to focus on most Important Risks

Question 6

Step 4: Sensitivity Analysis

Accepted Answer

Do sensitivity analysis to understand how the total profit/cost changes as risk factors change Identify main uncertainty drivers, i.e. Operating cost/hour, Proportion of scheduled flights, Hours flown. Charter price/hour, Load factor of scheduled flights, Ticket price/hour With the Tornado Chart we identified that Load Factor is most sensitive Tools: 1 Way Data Table, Two Way Data Table Sensitivity Chart: i.e. NPV vs. Load Factor of Scheduled Flights a. Slope of this Line: How much more can we make by one more person per flight "Only one extra person over a year makes 250K $" b. Total Range is the most importantc. Break-Even Point: Point at which we have a positive NPV (67%, average 67 ppl on plane) d; Conclusion: The NPV is quite sensitive to the load: By increased the Load Factor by 1% then our NPV increases by x

Question 7

Simulation Analysis - Traingular Distributions vs. Normal Distribution

Accepted Answer

Load Factor = Triangular Distirbution “…a delicate point was the occupancy (load factor) … which was currently around 60% ... the load factor may turn out to be different and range from lows of 50% to highs of 70%.” Use it when 3 numbers and you need the min, max and most likely point RiskTriang(50% , 60% , 70%) Modeling Uncertainty = Normal Distribution “Operating costs were fairly easy to predict. Keith’s best estimate for operating costs was $1200. However this could be wrong by $50 in either direction, mainly depending on fuel prices” RiskNormal($1200 , $25)

Question 8

Flaw of Averages

Accepted Answer

Basing Decisions on average results (base case model) is dangerous. The Base case result may not equal the average of the Simulation Model i.e. Profit based on average numbers does not equal the true average profit / cost (Transatlantic Airlines) We need the full distribution of Outcomes to make a decision -> Simulation Analysis. Example: - River: The Average does not tell you how deep it can be - Drunk person on certer of highway: On average state of the drunk person is a line, but almost surely the person will eventually die

Question 9

Two Way Data Table

Accepted Answer

Does not take into Account the likelihood of all events happening. We can see which is more impactful

Question 10

RISKCUMUL Distribution

Accepted Answer

Total Demand min: 50000 25% - 125000 expected / mean = 150000 75% - 190000 max = 300000 +RISKCUMUL(min,max,(25%value,50%,75%),(25%,50%,75%)

Question 11

Visualize Stochastic Dominance

Accepted Answer

They do not cross: By looking at the Cumulative Distribution For every level, the chance of making a level, there is a higher chance of making this amount of money for the Boss than with the Sales Person If they cross: We cannot say who is stocastically domdinant

Question 12

Who uses (Monte Carlo) Simulation?

Accepted Answer

Investment Banking: Risk management (including estimating Value-At-Risk) Pricing complex derivatives Setting fees and testing different business models Oil and Gas Allocating exploration budget (Shell) Pharmaceutical Aiding investment decisions (Merk) Optimizing licensing deals (CAT) Retail Queue management (Tesco’s) Film industry Selecting films to finance (Hollywood)

Question 13

Step 3: Scenario Analysis

Accepted Answer

When there’s uncertainty, do scenario analysis to get a range of possible outcomes Result: Range Very quick and easy to do Can stop analysis here if the range of possible outcomes is small Do more analysis only if necessary! (ANALYSIS PARALYSIS)

Question 14

Definition: Scenario

Accepted Answer

Scenarios are discrete, internally consistent views of how the world will look in the future, which can be selected to bound the possible range of outcomes that might occur.” (Michael Porter, Competitive Strategy) “Scenarios should present testing conditions for the business. The future will of course be different from all of these views/scenarios, but if the company is prepared to cope with any of them, it will be able to cope with the real world.”

Question 15

Sensitivity Chart:

Accepted Answer

Slope of this Line: How much more can we make by one more person per flight "Only one extra person over a year makes 250K $" Total Range is the most importantBreak-Even Point: Point at which we have a positive NPV (67%, average 67 ppl on plane) Conclusion: The NPV is quite sensitive to the load: By increased the Load Factor by 1% then our NPV increases by x

Question 16

Step 6: Simulation Analysis

Accepted Answer

Do a simulation analysis to get the full picture of possible outcomes: you get the true expected profit/cost and risk We are able to see how the outputs vary as a function of the varying inputs. Computer model that imitates a real-life situation Explicitly incorporates uncertainty in one or more input variables Enter probability distributions for the uncertain quantities and seeing how the NPV varies as the uncertain quantities vary Allow these random input variables to take on various values, and you keep track of any resulting output variables of interest.

Question 17

Advantages of Simulation Models vs Deterministic Models

Accepted Answer

Deterministic (nonsimulation) approach: Best guesses are incomplete and misleading No guarantee that the NPV based on best-guess inputs is representative of true output (Flaw of averages) Inapprorpriate Deterministic model Best guesses for uncertain inputs -> Dlaw of averages -> Best guesses for important outputs Simulation Model: Gives entire distribution of results, not only single bottom-line result. Probability Distribution for Uncerain inputs -> Simulation model -> Probability Distribution for important outputs

Question 18

Most widely used measures to Quantify Risk

Accepted Answer

Standard deviation Probability of loss Total range

4. Decision & Risk Analysis (Fach) / Simulation (Lektion)