ISM-C1004 - Business Analytics 1 - Assignment 3 (Total 28 points)

Transhipment with fixed costs (6 pts)

By now you are familiar with Pulp & Paper Ltd. that produces cardboard at 3 mills in Finland, Sweden
and Belgium. From the mills the cardboard is transported to 4 warehouses located in Finland,
Lithuania, Czech Republic and the UK, and then the sheeted cardboard to 13 customer areas.

The Sales & Marketing department has just released its demand estimates for the following month and
unfortunately the total demand exceed the total production capacity. The mills have informed Sales &
operations of the extra capacity they could provide for the following month and also provided
estimates on the cost of making use of this extra capacity.

The Advanced Analytics team has formulated a MILP model that extends the transhipment model you
are familiar with to also optimize decisions on the use of the extra capacity. The model is as follows:

𝑀={0,1,2}: Mills
Indexes:

𝑊={0,…,3}: Warehouses
𝐶={0,…,12}: Customer areas

𝑑_𝑗: demand in customer area 𝑗

Parameters:

𝑝_𝑖: production capacity at mill 𝑖

𝑒_𝑖: extra production capacity at mill 𝑖
𝑐_𝑖𝑘^𝑎: tranportation costs per ton from mill 𝑖 to warehouse 𝑘
𝑐_𝑘𝑗^𝑏: tranportation costs per ton from warehouse 𝑘 customer area 𝑗
𝑐_𝑖^𝑒: cost of using the extra production capacity at mill 𝑖

𝑥_𝑖𝑘: Tons transported from mill 𝑖 to warehouse 𝑘

Decision variables:

𝑦_𝑘𝑗: Tons transported from warehouse 𝑘 customer area 𝑗

𝑧_𝑖: Use extra capacity at mill 𝑖

min⁡∑_(𝑖∈𝑀)▒ 〖∑ _(𝑘∈𝑊)▒ 〖𝑐 _𝑖𝑘^𝑎 𝑥_𝑖𝑘 〗 + 〗 ∑ _(𝑘∈𝑊)▒∑_(𝑗∈𝐶)▒ 〖𝑐 _𝑘𝑗^𝑏

𝑦_𝑘𝑗 〗 +∑_(𝑖∈𝑀)▒ 〖𝑧 _𝑖 𝑐_𝑖^𝑒 〗
∑_(𝑘∈𝑊)▒𝑥_𝑖𝑘 ≤𝑝_i+𝑧_𝑖 𝑒_𝑖, 𝑖∈𝑀 (1)
∑_(𝑖∈𝑀)▒𝑥_𝑖𝑘 =∑_(𝑗∈𝐶)▒𝑦_𝑘𝑗 , 𝑘∈𝑊 (2)
∑_(𝑘∈𝑊)▒𝑦_𝑘𝑗 =𝑑_𝑗, 𝑗∈𝐶 (3)
𝑥_𝑖𝑘≥0, 𝑖∈𝑀, 𝑘∈𝑊
𝑦_𝑘𝑗≥0, 𝑘∈𝑊, 𝑗 ∈𝐶
𝑧_𝑖∈{0,1},𝑖∈𝑀
Unfortunately the Python implementation of the model was not completely finished before the analyst
responsible for its implementation broke her leg in a wierd skiing acident in the Alps. It is your job to
finish the code and run the code to identify the optimal transportation routes as well as the mills in
which extra capacity should be utilized.

You can use any Python environment you want, for instance, https://jupyter.cs.aalto.fi/

a) Complete the source code ("transhipment.py" in MyCourses) and copy paste the lines of code you
added to the textbox on the right. For each line of Python code you also need to have comments
exlaining what it does in your own words. (3 pts)

b) Report the optimal total costs, i.e., optimal objective function value. (1 pt)

Answer here
You can use any Python environment you want, for instance, https://jupyter.cs.aalto.fi/

a) Complete the source code ("transhipment.py" in MyCourses) and copy paste the lines of code you
added to the textbox on the right. For each line of Python code you also need to have comments
exlaining what it does in your own words. (3 pts)

b) Report the optimal total costs, i.e., optimal objective function value. (1 pt)

Answer here

c) Report the mills in which it is optimal to utilize the extra capacity.(1 pt)

Answer here

d) Report the optimal transportation routes and volumes. Include only those routes in which some
non-zero amount is being transported) (1 pt)

Answer here
al 28 points)
ANSWER FOR a) [PASTE SOURCE CODE HERE]
from pulp import *

#Indices
Mills = list(range(3))
Warehouses = list(range(4))
Customer_areas = list(range(13))

#Parameters
Transp_cost_A = [[54, 96, 137, 161],
[61, 121, 133, 141],
[211, 191, 90, 72]] #Transportation costs from mills to warehouses
Transp_cost_B = [[100, 120, 150, 160, 300, 310, 340, 490, 430,360, 280, 350, 200],
[200, 240, 280, 310, 280, 400, 440, 410, 380, 190, 80, 150, 90],
[280, 260, 320, 400, 140, 319, 290, 240, 230, 80, 60, 60, 160],
[320, 280, 200, 130, 130, 70, 100, 90, 100, 190, 370, 320, 390]] #Transportation costs from
warehouses to customers
Ext_cost = [300000,400000,450000] #Extra capacity costs
Capacity_limit=[9900, 2100, 4200] #Production capacities
Capacity_ext = [1100,1400,1500] #Extra capacities
Demands = [650,260,650,130,780,2500,910,3120,910,3640,1040,1650,1780] #Customer area demands

model = LpProblem("PP_MILP_model", LpMinimize) #Create a optimization model (minimization)

x=[[LpVariable("x_"+str(i)+str(k),0,None) for k in Warehouses] for i in Mills] #Create decision variables
x_ik
y=[[LpVariable("y_"+str(k)+str(j),0,None) for j in Customer_areas] for k in Warehouses] #Create decision
variables y_kj
z=[LpVariable("z_"+str(i),cat="Binary") for i in Mills ] #Create decision variables z_i

#Set a temporary objective function for testing purposes

model += (
lpSum(Transp_cost_A[i][k] * x[i][k] for i in Mills for k in Warehouses)
+ lpSum(Transp_cost_B[k][j] * y[k][j] for k in Warehouses for j in Customer_areas)
+ lpSum(Ext_cost[i] * z[i] for i in Mills),
"Total_Cost",
)

#TODO: add objective function and constraints here after back from Alps!:
#constraint 1: capacity contraint with extra capacity
for i in Mills:
model += (
lpSum(x[i][k] for k in Warehouses) <= Capacity_limit[i] + z[i] * Capacity_ext[i],
f"Capacity_at_Mill_{i}",
)
#constraint 2: balancing nodes
for k in Warehouses:
model += (
lpSum(x[i][k] for i in Mills) == lpSum(y[k][j] for j in Customer_areas),
f"Flow_Balance_at_Warehouse_{k}",
)
#constraint 3: balancing demand
for j in Customer_areas:
model += (
lpSum(y[k][j] for k in Warehouses) == Demands[j],
f"Demand_at_Customer_{j}",
)
model.solve() #Solve the optimal solution to model

#Print optimal objective function value:

#constraint 3: balancing demand
for j in Customer_areas:
model += (
lpSum(y[k][j] for k in Warehouses) == Demands[j],
f"Demand_at_Customer_{j}",
)

model.solve() #Solve the optimal solution to model

#Print optimal objective function value:

print("Optimal total cost "+str(model.objective.value())+" .")

#Print optimal decision variable values:

for i in Mills:
print("------")
if (z[i].varValue==1):
print("Purchase extra capacity for Mill "+str(i))
for k in Warehouses:
if (x[i][k].varValue>0):
print("Transport "+str(x[i][k].varValue)+" tons from Mill "+str(i)+" to Warehouse "+str(k)+".")

print("---------------")
for k in Warehouses:
for j in Customer_areas:
if (y[k][j].varValue>0):
print("Transport "+str(y[k][j].varValue)+" tons from Warehouse "+str(k)+" to Customer area
"+str(j)+".")
print("---")
tion costs from

stomer area demands

l (minimization)
e decision variables

uses] #Create decision

Grading
a) Is the implementation correct (+0-3pts)
1pt for correct objective function that includes comments
1pt for each set of constraints (1), (2) and (3) that includes comments
b)
s) 1pt for correct answers
0.5pts in an answer has small mistakes
c)
1pt for correct answers
0.5pts in an answer has small mistakes
d)
1pt for correct answers
0.5pts in an answer has small mistakes
ouse "+str(k)+".")

o Customer area
ISM-C1004 - Business Analytics 1 - Assignment 3 (Total 28 points)
Transportation problem with non-linear costs (6 pts)
P&T company produces canned peas. Peas are prepared at three canneries and shipped by truck to four warehouses.

The company has started a project that seeks to reduce transportation costs. The current monthly distribution of peas,
consumes production capacities of canneries (𝑠_𝑖), is presented in Table 1.

As a part of the project, the company has collected data on the transportation distances between the canneries and the
at historical data on shipping costs, and noticed that the single truck cost per kilometer on a particular route increases i
route. It is suspected that P&T is such a large player that its use of the trucking capacity affects prices. Specifically, P&T

[𝑐𝑜𝑠𝑡 𝑜𝑓 𝑎 𝑠𝑖𝑛𝑔𝑙𝑒 𝑡𝑟𝑢𝑐𝑘 𝑝𝑒𝑟 𝑘𝑚 ]=𝑎×[# 𝑜𝑓 𝑡𝑟𝑢𝑐𝑘𝑠 𝑜𝑛 𝑡ℎ𝑒 𝑟𝑜𝑢𝑡𝑒]+𝑏

fits the historical data well when coefficients 𝑎 and 𝑏 have the values presented in Table 3.

𝑖∈{1,…,3}, 𝑗∈{1,…,4}, for the decision variables representing the transportation volumes (in number of trucks). For
a) Mathematically formulate an NLP model to minimize transportation costs so that demands are satisfied and capacitie

and 𝑏, the following symbols: 𝑟_𝑖𝑗 for distances, 𝑠_𝑖 for cannery capacities and 𝑑_𝑗 for warehouse demands. (2p)

Answer here min⁡𝑥𝑥𝑥

b) Implement a spreadsheet model (3p) [HINT: You can start with a standard transportation model implementation with
decision variables. Then add two additional 3x4 tables to compute cost/km and total cost on each route. With the curre
would be equal to $277141]

c) Solve the model using "GRG nonlinear"-algorithm with the current plan as a starting solution (i.e. as values in the var
optimal objective function value. (0.5 p)

Answer here 233658.1

d) Solve the model using "GRG-nonlinear"-algorithm with a starting solution of all zeros. Report optimal objective functi
the value in c) but can also be the same]
216226.6
Answer here

Table 1
Destination (Warehouse)
Current plan for distribution (trucks)
Sacramento Salt Lake City Rapid City Albuquerque
Bellingham 75 10
Source
(Cannery) Eugene 5 60 60
Albert Lea 15 85

Table 2
Destination (Warehouse)
Distance (km)
Sacramento Salt Lake City Rapid City Albuquerque
Bellingham 1339 1481 1887 2502
Source
(Cannery) Eugene 1016 1201 1991 2283
Albert Lea 2872 1968 1120 1977
 Table 3
Regression model coefficients
a 0.003
b 0.3465
 (Total 28 points)

ed by truck to four warehouses.

ent monthly distribution of peas, which meets warehouse demands (𝑑_𝑗) and

es between the canneries and the warehouses (Table 2). They have also looked
er on a particular route increases if a lot of peas are transported along that
ty affects prices. Specifically, P&T has found that the regression model

+𝑏 Spreadsheet implementation her

ble 3.

emands are satisfied and capacities are not exceeded. Use the notation 𝑥_𝑖𝑗,
volumes (in number of trucks). For the fixed parameters use, in addition to 𝑎
𝑗 for warehouse demands. (2p)

tation model implementation with two 3x4 tables representing distances and
cost on each route. With the current distribution plan the transporation costs

g solution (i.e. as values in the variable cells when opening solver). Report the

os. Report optimal objective function value (0.5 p) [HINT: this might differ from

Grading
a) Is the formulation re
Formulation is reasona
exist.) (2 pts)
Formulation is reasona
Formulation is reasona
Completely unreasonab

b) Is the spreadsheet im
The variables are clearl
function and constraint
Small errors or not clea
There are major errors
No spreadsheet implem

c) Resonable value, doe

(+0.5pt)

d) Resonable value, doe

(+0.5pt)
There are major errors
No spreadsheet implem

c) Resonable value, doe

(+0.5pt)
d) Resonable value, doe
(+0.5pt)
eet implementation here (b)

bellingham
source eugene
albert sea

a
b
cost per km

total cost

constraint capacity

Grading
a) Is the formulation reasonable? (+0-2pts)
Formulation is reasonable and correct. (Note that multiple equivalent formulations
exist.) (2 pts)
Formulation is reasonable, but there are minor mistakes (1pts)
Formulation is reasonable, but there are major mistakes. (0.5pts)
Completely unreasonable or no mathematical formulation given. (0 pts)

b) Is the spreadsheet implementation reasonable? (+0-3pts)

The variables are clearly named and the cell values correspond to the objective
function and constraint coefficients. (3pt)
Small errors or not clearly names (2pt)
There are major errors in the spreadsheet implementation. (1pt)
No spreadsheet implementation given. (0 pts)

c) Resonable value, does not have to be correct as non-linear solver is not that good
(+0.5pt)

d) Resonable value, does not have to be correct as non-linear solver is not that good
(+0.5pt)
There are major errors in the spreadsheet implementation. (1pt)
No spreadsheet implementation given. (0 pts)

c) Resonable value, does not have to be correct as non-linear solver is not that good
(+0.5pt)
=
d) Resonable value, does not have to be correct as non-linear solver is not that good
(+0.5pt)
destination
sacramento salt lake city rapid city albuquerque
37 0 0 48
43 60 22 0
0 0 53 47
1339 1481 1887 2502
1016 1201 1991 2283
2872 1968 1120 1977
0.003
0.3465
0.4569706645 0.3465 0.3465 0.491029335497
0.4760293355 0.5265 0.411971 0.3465
0.3465 0.3465 0.506029 0.486970664503
22531.733707 0 0 59187.43173647 233658.1
20882.106573 37939.59 17900.41 0
0 0 30137.9 45078.95617906
demand
1 1 1 1 85 <= 85
1 1 1 1 125 <= 125
1 1 1 1 100 <= 100
80 60 75 95
= = = =
80 60 75 95
ISM-C1004 - Business Analytics 1 - Assignment 3 (Total 28 points)
Non-linear location problem (4 pts)
Whale Oil Ltd. operates six offshore oilrigs. It is planning to build a service platform from
which maintenance helicopters could serve the rigs, and needs to decide what would be the
best location for the service platform to minimize annual flight costs. The locations of the
rigs are given in the Table 1 and illustrated on the attached map.

The cost of one kilometer helicopter flight is about 5 euros. It is estimated that there will be
250 flights annually to each rig, except for Heidrun (175 flights) and Midgard (600 flights).
Furthermore, the service platform has to be within 110 km range from one of the two land
bases to allow building fiber optic communications cables.

a) Mathematically formulate an NLP problem that identifies the cost minimizing location for
the service platform. Remember to include the return flight to the distance travelled per
flight. (2p) [HINT: You can use the notations in Table 1 to make your formulation more
compact]

Answer here min⁡…

b) Implement the NLP model using spreadsheets (1p)

c) Solve the NLP model with "GRG nonlinear"-algorithm, report the optimal objective
function value and draw the optimal solution to the map (0.5p)

Answer here

d) Solve the NLP model with "Evolutionary"-algorithm, report the optimal objective
function value and draw the optimal solution to the map (0.5p) [HINT: answer might differ
form that of c)]

Answer here

Table 1
Oil rigs
Land base 1 Land base 2 Norne Heidrun Midgard Smorbukk Draugen Ormen Lange

i 1 2 3 4 5 6 7 8
xi 280 200 234 243 224 189 220 128

yi 70 25 322 186 220 185 123 72

al 28 points)

Spreadsheet implementation her

350

Norne
300

250

Midgard
200
Smorbukk Heidrun

150

Draugen
100

Ormen Lange Land base 1

50

Land base 2

0
120 140 160 180 200 220 240 260 280 300

Grading

a) Is the formulation reasonable? (+0-2pts)

Formulation is reasonable and correct. (Note that multiple
equivalent formulations exist.) (2 pts)
Formulation is reasonable, but there are minor mistakes
(1pts)
Formulation is reasonable, but there are major mistakes.
(0.5pts)
Completely unreasonable or no mathematical formulation
given. (0 pts)
b) Is the spreadsheet implementation reasonable? (+0-1pts)
The variables are clearly named and the cell values correspond
to the objective function and constraint coefficients. (2pt)
Small errors or not clearly names (0.5pt)
There are major errors in the spreadsheet implementation.
(0pt)
No spreadsheet implementation given. (0 pts)

c) Resonable value, does not have to be exactly correct as non-

linear solver is not that good (+0.5pt)

d) Resonable value, does not have to be exactly correct as

non-linear solver is not that good (+0.5pt)
eet implementation here (b)
ISM-C1004 - Business Analytics 1 - Assignment 3 (Total 28 points)
Non-linear resource allocation in marketing (5 pts)
GearSledge Ltd. is an online retailer of outdoors-equipment. They spend 150,000 euros
annually on marketing. The analysts at GearSledge have looked at historical data on
marketing expenditures and the impact they have had on obtaining new customers.

in euros (𝑥_𝑖) and the resulting increase in new customers (𝐶_𝑖 (𝑥_𝑖 )) is
Based on this analysis they found that the relationship between marketing expenditure

chraractarized by the non-linear regression model

𝐶_𝑖 (𝑥_𝑖 )=𝑎_𝑖−𝑎_𝑖 𝑒^(−(𝑥_𝑖/𝑏_𝑖 )^(𝑐_𝑖 ) ) ,

where model coefficients 𝑎_𝑖, 𝑏_𝑖 and 𝑐_𝑖 depend on the type of media (see Table 1).

GearSledge has hired you to build prescriptive analytics tools to support marketing allocation
decisions. GearSledge's new policy is that more should be spent on internet marketing
(Podcasts, Social media) than on traditional media (TV, radio). Moreover, to meet prior
contractual commitments at least 30 keuros must be spent on Radio.

a) Develop a single figure visualization that shows the new customers obtained as a function of
marketing expenditure for each media. Clearly label axes and graphs. An appropriate choice for
the range of the horizontal axis would be from zero euros to the annual budget. (1p)

b) Mathematically formulate an NLP problem that identifies the allocation of marketing

contractual constraints are satisfied. (2p) [HINT: You can use the symbols 𝑎_𝑖,𝑏_𝑖, 𝑐_𝑖, 𝑥_𝑖 to
expenditures so that the number of new customers is maximized, while budget, policy and

make the formulation more compact. WARNING: 𝐶(𝑥_𝑖/1000)≠𝐶(𝑥_𝑖 )/1000 ]

Answer here max⁡…

c) Implement the NLP model using spreadsheets, solve it, and report the optimal allocation.
(2p) [HINT: Give a feasible starting solution for the solver]

Answer here

Table 1 i ai bi ci
TV 1 1500 20000 1
Podcasts 2 1300 10000 0.5
Radio 3 2000 70000 4
Social media 4 2000 60000 2
Total 28 points)

000 euros
ata on
omers.
penditure

1).

llocation
ng
or Spreadsheet implementation here (b)

unction of
choice for

ting

𝑐_𝑖, 𝑥_𝑖 to
cy and

ocation.

Grading
a) Is the visualization appropriate (0-1pts)

b) Is the formulation reasonable? (+0-2pts)

Formulation is reasonable and correct. (Note that multiple equivalent formulations exist.)
(2 pts)
Formulation is reasonable, but there are minor mistakes (1pts)
Formulation is reasonable, but there are major mistakes. (0.5pts)
Completely unreasonable or no mathematical formulation given. (0 pts)

c) Does the spreadsheet implementation work? (+0-2pts)

Solver set up. The variables are clearly named and the cell values correspond to the
objective function and constraint coefficients. (2pt)
Small errors (1pt)
There are major errors in the spreadsheet implementation. (0.5pt)
No spreadsheet implementation given. (0 pts)
ons exist.)

the
ISM-C1004 - Business Analytics 1 - Assignment 3 (Total 28 points)
Non-linear portfolio optimization problem (7 pts)
Wealthy industrialist H.E. Pennypacker is restructuring her investment portfolio and has decided to allocate capital acros

1. NoDur: Consumer NonDurables -- Food, Tobacco, Textiles, Apparel, Leather, Toys

2. Durbl: Consumer Durables -- Cars, TV's, Furniture, Household Appliances
3. Manuf: Manufacturing -- Machinery, Trucks, Planes, Off Furn, Paper, Com Printing
4. Enrgy: Oil, Gas, and Coal Extraction and Products
5. Chems: Chemicals and Allied Products
6. BusEq: Business Equipment -- Computers, Software, and Electronic Equipment
7. Telcm: Telephone and Television Transmission
8. Utils: Utilities
9. Shops: Wholesale, Retail, and Some Services (Laundries, Repair Shops)
10. Hlth: Healthcare, Medical Equipment, and Drugs
11. Money: Finance
12. Other: Mines, Constr, BldMt, Trans, Hotels, Bus Serv, Entertainment

She has hired you for support the decision-making, and asks you to construct a Markowitz model to identify an allocatio
standard deviation of monthly returns), and a monthly expected return of at least 1%. Furthermore, at most 2% of the c
fund and at most 1% to Enrgy fund, since H.E. Pennypacker does not want to appear as a supporter of the Tobacco and O
5% of the capital should be invested in Health fund. The planning horizon is one month and the model should use data o
industry funds (Table 1).

a) Build the NLP model using spreadsheets and solve the optimal allocation. Shorting of the funds is not allowed, i.e., fun
(3p) [HINT: To check the correctness of your implementation, make sure that if you invest in only one fund, then the exp
model match those in the historical data].

b) To which funds should H.E. Pennypacker allocate capital and how much? What is the expected return and standard de
allocation? (1p)

Answer here

c) After seeing the results H.E. Pennypacker wants you to analyze the tradeoff between risk and expected returns more c
of different values for the expected return requirement and produce a graph showing standard deviation of returns as a
Range of five different values suffices, i.e., no use of macros is required. Make sure that this range includes the maximum
portfolios. (3p)

Table 1
mean 0.257407 0.739753 0.95321 1.187037 0.790617 1.004444 0.189012 0.931481
std 5.987369 10.52106 7.662813 14.60619 7.380956 7.737281 7.858871 4.560736

Monthly return (%)

Date
NoDur Durbl Manuf Enrgy Chems BusEq Telcm Utils
201801 1 2.31 0.54 1.06 1.26 -0.73 4.47 0.71 -3.82
201802 2 -4.17 -6.07 -5.52 -10.38 -4.34 -1.34 -4.97 -5.18
201803 3 0 -2.23 1.14 0.91 2.44 0.58 -3.14 5.13
201804 4 0.79 -2.57 -0.78 9.71 -1.16 0 -1.36 3.47
201805 5 2.45 6.54 5.49 7.05 2.98 6.6 -2.14 1.77
201806 6 6.8 0.94 0 7.66 0.08 0.74 6.63 1.95
201807 7 -0.94 1.25 3.28 -2.05 1.82 -0.23 -1.23 1.36
201808 8 2.47 3.09 0.53 -4.72 -0.83 7.58 7.29 0.9
201809 9 0.16 -3.88 -0.53 0.53 -2.23 -2.53 2.19 -0.08
201810 10 -4.58 -13.97 -12.65 -15.34 -11.05 -10.97 -4.91 -1.55
201811 11 0.19 -0.88 0.36 -11.99 0.96 0.66 -0.77 4.34
201812 12 -12.39 -12.93 -13.51 -22.1 -11.27 -9.68 -13.74 -4.67
201901 13 10.63 16.9 13.97 18.04 16.67 15.71 12.75 4.97
201902 14 3.37 7.04 5.71 0.74 5.38 7.12 2.49 3.26
201903 15 -0.68 -6.73 -3.83 1.13 -4.34 -0.6 -3.49 3.24
201904 16 2.23 4.26 3.66 0.06 5.14 4.71 3.71 1.07
201905 17 -9.18 -11.28 -9.47 -20.71 -10.09 -8.11 -6.38 -1.48
201906 18 5.19 10.97 8.58 3.83 7.02 5.08 2.45 4.46
201907 19 1.44 0.65 -1.65 -9.41 2.53 1.12 -0.12 -1.31
201908 20 -2.91 -6.35 -7.9 -16.38 -5.69 -5.32 -8.13 0.49
201909 21 2.61 6.15 5.79 5.05 7.87 0.49 0.16 3.75
201910 22 -1.48 0.42 1.52 -9.84 -2.62 0.47 -3.16 -1.71
201911 23 2.06 2.44 1.56 -3.52 1.01 4.88 3.14 -2.66
201912 24 2.63 1.84 3.65 19.58 2.19 3.38 4.37 5.48
202001 25 -2.41 -3.81 -2.66 -18.7 -6.47 1.46 -1.6 0.79
202002 26 -9.88 -10.97 -8.27 -18.58 -11.38 -7.53 -4.48 -10.1
202003 27 -18.07 -27.7 -24.55 -46.59 -21.16 -17.99 -23.39 -10.75
202004 28 13.27 28.15 16.58 67.45 16.49 21.39 10.11 10.04
202005 29 4.25 10.4 6.15 -4.31 9.32 10.67 3.61 3.84
202006 30 3.92 21.52 10.51 3.54 6.51 8.26 12.54 -4.35
202007 31 5.1 9.04 4.5 -1.17 6.98 10.87 0.18 3.36
202008 32 4.25 11.6 5.42 -1.81 7.69 4.01 3.44 -0.85
202009 33 -1.93 -0.97 -2.09 -13.69 0.51 -2.66 -5.91 -2.13
202010 34 2.52 3.88 2.25 -4.93 2.06 0.8 -4.34 4.06
202011 35 17.74 30.5 23.7 35.93 18.98 21.99 25.69 6.41
202012 36 6.2 8.92 8.78 15.14 9.51 12.52 3.41 1.38
202101 37 7.13 30.93 8.76 22.18 7.39 8.76 14.96 -0.11
202102 38 6.74 7.57 10.98 29.23 9.25 7.24 8.15 -1.51
202103 39 4.01 0.71 4.55 -4.8 5.93 -1.53 3.17 9.81
202104 40 4.04 1.19 1.46 -1.46 3.78 2.65 2.59 3.23
202105 41 3.63 2.81 3.5 15.92 2 -0.45 6.36 0.08
202106 42 0.71 0.82 0.48 12.38 -1.03 7.83 3.35 1.2
202107 43 -2.93 -6.4 -2.39 -12.04 -2.94 -2.74 -2.48 1.81
202108 44 0.13 -4.27 0.65 0.9 1.03 1.67 4.14 2.38
202109 45 -4.7 -3.59 -4.63 21.51 -2.63 -4.89 -1.24 -5.23
202110 46 2.33 1.42 3.5 7.56 7.95 4.81 -3.45 4.13
202111 47 -3.13 -4.29 -2.27 -11.83 -2.96 -6.35 -8.6 -2.3
202112 48 2.78 -1.03 1.64 1.31 2.33 -1.2 -0.25 8.46
202201 49 -6.83 -9.89 -6.59 10.6 -7.14 -12.18 -2.81 -2.82
202202 50 0.1 -2.81 1.45 14.46 2.58 -0.66 0.39 0.61
202203 51 -1.62 -1.59 1.63 17.23 3.44 0.38 1.25 9.16
202204 52 -4.94 -11.19 -8.99 -1.05 -6.72 -13.9 -11.38 -4.16
202205 53 -2.36 1.16 -0.32 12.92 0.24 -4.02 5.41 6.75
202206 54 -7.26 -8.77 -10.15 -19.05 -10.67 -9.41 -11.9 -6.17
202207 55 4.98 9.69 10.88 11.02 9.09 11.03 1.42 7.88
202208 56 -3.14 -1.28 -2.1 5.98 1.27 -2.24 2.93 0.93
202209 57 -15.54 -15.91 -13.91 -12 -14.61 -12.04 -17.23 -11.79
202210 58 6.84 8.76 11.22 24.88 8.02 6.28 7.05 7.29
202211 59 1.74 2.84 3.32 0.31 4.3 -0.61 -5.66 5.17
202212 60 -6.97 -11.35 -6.03 -8.19 -7.68 -7.77 -9.14 -2.82
202301 61 13 30.82 16.55 3.28 13.57 19.03 24.24 2.52
202302 62 -5.03 -5.62 -0.89 -7.02 -3.84 -3.41 -8.11 -3.72
202303 63 -4.35 -8.21 -5.72 -6.7 -5.68 -2.44 -8.82 1.21
202304 64 -1.48 -6.13 -3.5 -2.4 -2.46 -6.42 -0.51 -0.1
202305 65 -3.59 -3.55 -3.5 -6.67 -2.76 7.61 -8.53 -3.16
202306 66 3.31 9.09 12.23 10.1 10.05 5.81 8.46 2.29
202307 67 2.86 7.96 7.11 11.59 5.03 5.89 2.96 2.66
202308 68 -5.05 -9.47 -5.58 -1.23 -7.5 -7.82 -2.64 -5.17
202309 69 -7 -7.56 -6.18 3.39 -7.72 -9.12 -6.71 -5.72
202310 70 -7.66 -14.45 -7.86 -4.44 -7.02 -10.07 -4.81 -0.59
202311 71 8.33 2.02 6.39 -4.67 8.21 11.64 5.55 5.85
202312 72 6.42 10.14 11.52 1.5 10.55 11.3 6.79 3.78
202401 73 -3.92 -8.12 -5.51 -5.19 -7.19 -3.35 -4.15 -4.88
202402 74 1.93 2.92 5.81 2.3 4.82 5.36 -5.81 2.37
202403 75 0.98 0.21 4.96 7.39 7.7 1.03 -4.44 5.41
202404 76 -4.81 -10.2 -5.18 0 -6.56 -7.3 -3.08 0.44
202405 77 7.16 17.77 5.36 3.58 4.33 2.62 11.92 5.6
202406 78 -4.33 -7.26 -4.68 -5.42 -8.07 0.01 1.89 -3.09
202407 79 4.81 4.06 6.84 4.35 4.3 4.34 13.58 8.93
202408 80 -1.51 -4.01 -2.67 -6.36 -2.26 -3.25 -3.85 0.3
202409 81 1.08 -2.72 4.29 -0.59 3.54 0.64 2.74 3.66
al 28 points)

Spreadsheet implementation here (b)

to allocate capital across 12 industry funds:

Grading
a) Is the implementation reasonable? (+0-3pts)
Implementation is reasonable and correct. (Not
Implementation is reasonable, but there are m
Implementation is reasonable, but there are m
Completely unreasonable or no mathematical fo
b) Everyting correct 1p
c) 3 points if the graph has been done. -1p if axe
min/max not included etc.

el to identify an allocation with minimal risk (in terms of

ore, at most 2% of the capital can be allocated to NoDur
rter of the Tobacco and Oil & Gas industries. At least
model should use data on monthly historical returns of the

ds is not allowed, i.e., fund weights must be non-negative.

y one fund, then the expected return and risk given by your

d return and standard deviation of returns of the optimal

expected returns more closely. Solve the model for a range

deviation of returns as a function of the expected returns.
ge includes the maximum return and minimum risk

1.161481 0.144568 0.882469 0.629136

8.143928 8.71688 6.205352 7.01277

Shops Hlth Money Other

1.72 6.87 2.46 3.01
-4.66 -3.53 -2.6 -5.92
1.67 0.25 1.71 0.64
1.86 -0.2 0.82 1.56
3.93 7.68 2.93 4.64
3.59 -2.35 -0.24 0.71
0.41 -1.53 1.21 1.46
5.61 5.8 2.02 1.21
 -2.66 -1.15 -2.41 -2.3
-8.49 -14.51 -7.53 -9.26
-1.99 -0.4 1.45 -0.43
-13.18 -18.42 -10.66 -12.19
11.72 18.4 8.83 14.57
3.85 6.53 5.48 3.57
-3.78 3.09 -4.36 -1.16
2.15 -2.44 4.92 3.63
-11.25 -8.3 -5.38 -6.81
5.72 4.5 5.65 5.65
-1.07 -5.29 1.42 0.77
-3.68 -4.88 -4.54 -3.82
6.71 -3.44 4.16 2
0.49 -0.4 2.03 1.61
4 8.68 3.82 0.96
2.11 9.83 2.46 3.78
-5.74 4.92 -3.65 -2.42
-9.01 -0.86 -8.31 -9.14
-25.89 -17.43 -23.24 -28.12
23.13 21.74 9.01 17
10.23 13.91 0.74 10.97
6.69 4.69 4.57 6.99
5.85 3.69 -1.79 4.4
12.03 1.32 3.88 8.94
-2.13 -1.24 -3.74 -2.73
2.05 -2.73 7.02 -0.7
22.23 18.48 15.9 22.1
5.34 9.84 8.21 10.17
19.71 19.08 2.29 8.2
7.78 6.32 11.11 12.14
8.27 -3.57 6.87 5.7
3.47 -2.15 3.07 2.15
5.91 -3.31 3.57 1.07
1.85 5.33 -1.94 -0.45
-3.45 -8.83 -0.78 -3.15
0.46 5.11 3.4 1.96
-5.01 -5.71 -0.13 -2.13
3.18 -3.7 4.57 2.76
-2.71 -9.83 -2.8 -3.37
2.03 -5.69 3 1.16
-9.04 -15.42 -1.93 -8.1
-0.04 -4.69 0.12 0.23
-0.3 3.17 -1.17 3.01
-6.74 -19.88 -7.69 -8.81
-3.45 -7.54 1.8 -2.11
-9.36 2.2 -7.4 -9.16
9.07 6.96 6.27 7.14
0.57 2.74 -2.01 -0.77
-12.58 -10.55 -7.41 -7.67
10.73 1.18 8.82 5.55
4.48 -3.66 3.1 0.31
-9.58 -4.13 -4.14 -4.75
16.55 19.87 7.29 12.64
-3.94 -6.69 -0.19 -3.29
-5.68 -7.91 -12.5 -3.95
-1.98 1.88 -3.43 -0.64
-3.4 1.61 -3.38 -2.15
10.22 1.2 7.29 5.57
5.74 0.39 10.07 2.67
-6.52 -8.81 -5.61 -5.49
-7.7 -9.06 -4.22 -6.8
-6.24 -10.76 -3.63 -6.87
7.65 8.28 10 8.63
11.86 17.48 13.96 6.76
-5.23 -0.12 -2.84 -4.58
5.61 12.38 -0.62 4.69
2.55 1.68 2.96 3.73
-7.29 -10.97 -5.73 -5.7
15.73 1.66 4.53 1.65
-3.14 -6.88 0.55 -3.08
3.85 8.34 13.31 5.14
-2.63 -4.12 0.37 -3.63
3.26 -2.29 0.46 -0.59
tation reasonable? (+0-3pts)
reasonable and correct. (Note that multiple equivalent formulations exist.) (3 pts)
reasonable, but there are minor mistakes (2pts)
reasonable, but there are major mistakes. (1pts)
onable or no mathematical formulation given. (0 pts)
t 1p
aph has been done. -1p if axes not labeled, the frontier is at the wrong place,
ded etc.