Modeling Risk and Realities Week 4 Session 3
Modeling Risk and Realities Week 4 Session 3
Session 1
Session 2
Session 3
Stargrove.xlsx
Sample of random variables from the normal distribution with mean 90 and
standard deviation of 25 and its descriptive statistics
Demand Realizations
Sample Mean
Sample St. Dev.
34.65997588
95.14134513
103.4140237
83.66079351
146.1676643
57.91029844
128.539838
159.5054041
76.73592927
116.6149982
100.235027
12.31380659
99.2776844
#N/A
38.9396755
1516.298328
-0.581372514
-0.101693197
124.8454282
34.65997588
159.5054041
1002.35027
10
27.85576579
100.235027
38.9396755
Sample of random variables from the normal distribution with mean 90 and
standard deviation of 25 and its descriptive statistics
Demand Realizations
Sample Mean
Sample St. Dev.
34.65997588
95.14134513
103.4140237
83.66079351
146.1676643
57.91029844
128.539838
159.5054041
76.73592927
116.6149982
100.235027
12.31380659
99.2776844
#N/A
38.9396755
1516.298328
-0.581372514
-0.101693197
124.8454282
34.65997588
159.5054041
1002.35027
10
27.85576579
100.235027
38.9396755
95% confidence level identifies the 95% confidence interval for the true
expected value of the simulated random variable: based on the results of this
simulation, we are 95% confident that the true expected value is in the interval =
sample mean +/- 95% confidence level 110.24 +/- 27.86 = [82.38, 138.10]
Sample of random variables from the normal distribution with mean 90 and
standard deviation of 25 and its descriptive statistics
Demand Realizations
Sample Mean
Sample St. Dev.
34.65997588
95.14134513
103.4140237
83.66079351
146.1676643
57.91029844
128.539838
159.5054041
76.73592927
116.6149982
100.235027
12.31380659
99.2776844
#N/A
38.9396755
1516.298328
-0.581372514
-0.101693197
124.8454282
34.65997588
159.5054041
1002.35027
10
27.85576579
100.235027
38.9396755
95% confidence level identifies the 95% confidence interval for the true
expected value of the simulated random variable: based on the results of this
simulation, we are 95% confident that the true expected value is in the interval =
sample mean +/- 95% confidence level 110.24 +/- 27.86 = [82.38, 138.10]
Sample of random variables from the normal distribution with mean 90 and
standard deviation of 25 and its descriptive statistics
Demand Realizations
Sample Mean
Sample St. Dev.
34.65997588
95.14134513
103.4140237
83.66079351
146.1676643
57.91029844
128.539838
159.5054041
76.73592927
116.6149982
100.235027
12.31380659
99.2776844
#N/A
38.9396755
1516.298328
-0.581372514
-0.101693197
124.8454282
34.65997588
159.5054041
1002.35027
10
27.85576579
100.235027
38.9396755
95% confidence level identifies the 95% confidence interval for the true
expected value of the simulated random variable: based on the results of this
simulation, we are 95% confident that the true expected value is in the interval =
sample mean +/- 95% confidence level 110.24 +/- 27.86 = [82.38, 138.10]
Sample of random variables from the normal distribution with mean 90 and
standard deviation of 25 and its descriptive statistics
Demand Realizations
Sample Mean
Sample St. Dev.
34.65997588
95.14134513
103.4140237
83.66079351
146.1676643
57.91029844
128.539838
159.5054041
76.73592927
116.6149982
100.235027
12.31380659
99.2776844
#N/A
38.9396755
1516.298328
-0.581372514
-0.101693197
124.8454282
34.65997588
159.5054041
1002.35027
10
27.85576579
100.235027
38.9396755
95% confidence level identifies the 95% confidence interval for the true
expected value of the simulated random variable: based on the results of this
simulation, we are 95% confident that the true expected value is in the interval =
sample mean +/- 95% confidence level 110.24 +/- 27.86 = [82.38, 138.10]
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
Reward $
Risk
50,778,195.32
0.2
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
Reward $
Risk
50,778,195.32
0.2
Profit
Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
50778195.32
2945866.396
54018185.95
#N/A
9315647.495
8.67813E+13
1.570184843
-1.419981677
28633616.9
30166383.1
58800000
507781953.2
10
6664012.769
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
Reward $
Risk
50,778,195.32
0.2
Profit
Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
50778195.32
2945866.396
54018185.95
#N/A
9315647.495
8.67813E+13
1.570184843
-1.419981677
28633616.9
30166383.1
58800000
507781953.2
10
6664012.769
Our estimate for the reward measure, $50,778,195 may also be quite removed
from the true expected profit value associated with the decision we consider
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
Reward $
Risk
50,778,195.32
0.2
Profit
Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
50778195.32
2945866.396
54018185.95
#N/A
9315647.495
8.67813E+13
1.570184843
-1.419981677
28633616.9
30166383.1
58800000
507781953.2
10
6664012.769
Our estimate for the reward measure, $50,778,195 may also be quite removed
from the true expected profit value associated with the decision we consider
With a simulation that samples the random input variables only 10 times, the
reliability of the estimates for the mean and the standard deviation for any
random quantity involved may be limited
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
Reward $
Risk
50,778,195.32
0.2
Profit
Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
50778195.32
2945866.396
54018185.95
#N/A
9315647.495
8.67813E+13
1.570184843
-1.419981677
28633616.9
30166383.1
58800000
507781953.2
10
6664012.769
Based on the results of this simulation, we can be 95% confident that the true
expected profit under the decision we consider lies in the interval $50,778,195
+/- $6,664,013 = [$44,114,182, $57,442,208]
Stargrove_1000.xlsx, seed 123 for the B column and seed 1234 for the C column
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
57,920,800.05
35,017,702.31
45,448,387.76
40,417,837.90
56,460,689.30
56,113,851.70
55,434,315.43
58,302,165.39
53,191,844.26
56,471,197.37
Reward $
Risk
52,131,111.07
0.156
Profit
Mean
Standard Error
Median
Mode
52131111.07
205518.4113
54261174.8
58800000
Standard Deviation
6499062.808
Sample Variance
4.22378E+13
Kurtosis
Skewness
0.783172165
-1.151004166
Range
32537281.28
Minimum
26262718.72
Maximum
Sum
Count
Confidence Level(95.0%)
58800000
52131111066
1000
403297.2995
Based on the results of this longer simulation, we can now be 95% confident that
the true expected profit under the decision we consider lies in the interval
$52,131,111 +/- $403,297 = [$51,727,814, $52,534,408]
Stargrove_1000.xlsx
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Profit
30,166,383.10
51,368,084.19
56,668,287.72
50,966,937.91
58,503,486.59
40,442,284.55
57,108,800.62
58,800,000.00
46,992,738.24
56,764,950.24
57,920,800.05
35,017,702.31
45,448,387.76
40,417,837.90
56,460,689.30
56,113,851.70
55,434,315.43
58,302,165.39
53,191,844.26
56,471,197.37
Reward $
Risk
52,131,111.07
0.156
below Threshold?
Mean
Standard Error
Median
Mode
0.156
0.011480235
0
0
Standard Deviation
0.363036907
Sample Variance
0.131795796
Kurtosis
1.609118878
Skewness
1.898921967
Range
Minimum
Maximum
Sum
156
Count
1000
Confidence Level(95.0%)
0.022528141
Based on the results of this longer simulation, we can be 95% confident that the
true value of the risk measure under the decision we consider lies in the interval
0.156+/- 0.023 = [0.133, 0.179]
Suppose that Stargrove would like to compare the decision of building 12 regular
floors and 3 luxury floors (R=96 and L=12) with the decision of building 11 regular
floors and 4 luxury floors (R=88 and L=16)
We can use 1000 random values we have already generated for the demand for
regular apartments and 1000 random values we have already generated for the
demand for luxury apartments to estimate the reward and the risk associated with
the decision of R=88 and L=16
We can then compare reward and risk estimates for the two decisions
Stargrove_1000_TwoDecisions.xlsx
Profit for R=96, L=12
Mean
52131111 Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
205518.4
54261175
58800000
6499063
4.22E+13
0.783172
-1.151
32537281
26262719
58800000
5.21E+10
1000
403297.3
0.156
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
0.011480235
0
0
0.363036907
0.131795796
1.609118878
1.898921967
1
0
1
156
1000
0.022528141
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
0.159
0.011569
0
0
0.365859
0.133853
1.491811
1.867841
1
0
1
159
1000
0.022703
Stargrove_1000_TwoDecisions.xlsx
Profit for R=96, L=12
Mean
52131111 Mean
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
205518.4
54261175
58800000
6499063
4.22E+13
0.783172
-1.151
32537281
26262719
58800000
5.21E+10
1000
403297.3
0.156
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
0.011480235
0
0
0.363036907
0.131795796
1.609118878
1.898921967
1
0
1
156
1000
0.022528141
Standard Error
Median
Mode
Standard Deviation
Sample Variance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
Confidence Level(95.0%)
0.159
0.011569
0
0
0.365859
0.133853
1.491811
1.867841
1
0
1
159
1000
0.022703
Stargrove_1000_TwoDecisions.xlsx
Decision
R=96, L=12
R=88, L=16
[51.73, 52.53]
[50.31, 51.01]
[0.133, 0.179]
[0.136, 0.182]
Based on the results of the simulation with n=1000 runs, we are 95% confident
that the expected profit under the decision R=96, L=12 is higher than the
expected profit under the decision R=88, L=16
Stargrove_1000_TwoPolicies.xlsx
Decision
R=96, L=12
R=88, L=16
[51.73, 52.53]
[50.31, 51.01]
[0.133, 0.179]
[0.136, 0.182]
Based on the results of the simulation with n=1000 runs, we are 95% confident
that the expected profit under the decision R=96, L=12 is higher than the
expected profit under the decision R=88, L=16
The results of this simulation do not allow us to distinguish between the levels of
risk associated with those two decisions at the same level of confidence
Ultimately, the goals are to 1) limit the consideration set to decisions that result in
risk measures limited by the tolerance level of a decision maker, and 2) among
the decisions that satisfy constraint(s) on acceptable risk level(s), choose one
that generates highest reward, at the selected confidence level
Simulation provides imperfect estimates of reward and risk, but the notion of
confidence intervals enables a decision maker to compare alternatives even
using those imperfect estimates