Ch. 5 Risk Flashcards
(24 cards)
Simulation
Technique helpful in analyzing models in which the value to be assumed by one or more independent(x) variables is uncertain. It’s a computer model that imitates real life situation. (technique that describes the behaviour of a bottom line perf. measure)
Benefit of simulation
enables managers to answer “what-if” questions without changing a physical system
Categories of Simulation
-Monte Carlo-repeated samples from prob. dist. of model input to characterize the output
-Discrete Event- models dynamics and behaviour of interacting elements of a system. associated with physical sim.
Single period inventory Problems
Need random numbers to drive the whole process.
-(Newsvendor Problem) Vendor who sells newspapers everyday(orders newspaper w/o knowing the actual dem.) also used for one time decision for product to order
-Excess inventory- too many news. with 0 value. - material cost(purchase price), handling, holding cost, spoilage with potential salvage
-Inventory Shortage: too few opp. to make profit is lost. (implicit cost). Lost sales& prof. , loss of goodwill, cost of rush order to replenish
Sim. using @risk
-Distribution cell: input that’s uncertain(green)
-Parameter/chang.: a controllable input that we vary. (yellow)
-Output cell: what you want @Risk to monitor(blue)
Why is Add in software good
It is much easier because it has in built functions and automatically calculates various statistics and provides histograms
@RISK IS PART OF WHAT
Palisade’s Decision Tools Suite that allows you to run simulation in Excel Models. NEEDS TO BE STARTED BEFORE EXCEL IS OPENED
Use of distribution
needed as a source of randomness
-Some are continuous (Normal distribution) and others are discrete with integer values (Poisson and Binomial)
Recalc button use
In Random Mode,it generates new
values every time you re-calculate
the worksheet. F9 to generate random no.
Output Cell
-To statistically monitor during the sim.
-Can have more than 1 cell as output but must be identified indiv.
Number of Trials
1000 trials as its large enough to give good results
Interpretation of Mean
if sim. is run again output values will be the same. The values will be similar when a diff. seed no. is selected because the underlying distr. is the same
-State the avg expected cost is $
Avg. expected cost is from
The result is from a sample , not the result from the population μ
Central Limit Theorem
CLT states that sample means are
distributed normally. when the mean are normally distributed = conf. intervals
Normal curve
Properties of the Normal Distribution That Help Us Understand Confidence Intervals
Determining conf. intervals
-normally distributed
-sample mean
-sample sd or standard error
-conf. interval
Tradeoff between conf. interval
The tradeoff between the 3 different
confidence intervals is the certainty versus the
size of the interval
Sample proportion
The proportion of observations in the sample
that lie above or below the cut-off value of
interest
Pop. proportion
The (unknown) prop. of observations in the pop. that lie above or below the cut-off
value
Election as a pop.
The support percentage (population value) is
known only after all votes are counted after the election.
→ sampling risk is the possibility that the poll (sample) is not representative of the true numbers
In one of the most famous headlines in
American newspaper history, the Chicago
Daily Tribune’s front page read “Dewey
Defeats Truman” in the 1948 US Presidential
Election. (Picture shows the true winner,
Truman, holding the newspaper).
Choose Line chart
-Cost decision: curve must decrease followed by an incr. (pick lowest val)
-Profit decision: curve must have an incr. followed by a decr. (pick highest val. )
Advantages of Sim.
-Straightforward and flexible.
Can analyze large and complex situations.
Facilitates what-if analysis.
Does not interfere with the system
→ E.g. hospital and transportation
Can examine interactive effects.
Inclusion of scenarios that other models may not permit.
Disadvantages of Sim.
-Simulation can be very expensive and a long,
complicated process to develop.
Simulations do not generate optimal solutions to problems
→ we choose the best from the parameters selected
Managers must generate all the conditions and constraints.
Each model is unique and generally not
transferable to other problems.
Infos
Any uncertainty in the input cells flows through the spreadsheet model to create a related uncertainty in the value of the output cell(s).
The variability and distribution of the sample
values for the dependent variable(s) can then be analyzed to gain insight into the possible
outcomes that might occur.
