Inventory Simulation With Lost Sales

Author
Crystal Ball
Summary
The two basic inventory decisions that managers face are: (1) how much additional inventory to order or produce, and (2)
when to order or produce it. Although it is possible to consider these two decisions separately, they are so closely related
that a simultaneous solution is usually necessary. Given variable (uncertain) demand over a 52-week period, this model
determines an optimal order quantity and reorder point that results in the lowest possible total annual costs.
Adapted from James R. Evans and David L. Olson. Introduction to Simulation and Risk Analysis. New York: Prentice-Hall,
1998.
Keywords: inventory, position, reorder point, order quantity, safety stock, lost sales, costs, lead time, optimization
Discussion
Typically, the objective is to minimize total inventory costs. Total inventory costs typically include holding, ordering, shortage,
and purchasing costs. In a continuous review system, managers continuously monitor the inventory position. Whenever the
inventory position falls at or below a level R, called the reorder point, the manager orders Q units, called the order quantity.
(Note that the reorder decision is based on the inventory position, including orders not yet received, and not the inventory
level, including only inventory on-hand. If managers used the inventory level, they would place orders continuously as the
inventory level fell below R until they received the order.) When you receive the order after the lead-time, the inventory level
jumps from zero to Q, and the cycle repeats.
In inventory systems, demand is usually uncertain, and the leadtime can also vary. To avoid shortages, managers often
maintain a safety stock. In such situations, it is not clear what order quantities and reorder points will minimize expected
total inventory cost. Simulation models can address this question. In this example, demand is uncertain and is Poissondistributed with a mean of 100 units per week. Thus, the expected annual demand is 5,200 units.
Note: For large values of the rate parameter, the Poisson distribution is approximately normal. Thus, this assumption is the
same as saying that the demand is normally distributed with a mean of 100 and standard deviation of 15. The Poisson is
discrete, thus eliminating the need to round off normally distributed random variables.
Additional relationships that hold for the inventory system are:
- Each order costs $50 and the holding cost is $0.20 per unit per week ($10.40 for one year).
- Every unfilled demand is lost and costs the firm $100 in lost profit.
- The time between placing an order and receiving the order is 2 weeks. Therefore, the expected demand during lead-time
is 200 units. Orders are placed at the end of the week, and received at the beginning of the week.
The traditional economic order quantity (EOQ) model suggests an order quantity:
Q = Sqrt ((2 * 5200 * 50) / 10.4) = 224
For the EOQ policy, the reorder point should equal the lead-time demand. That is, if the lead-time demand is exactly 200
units, an order should be placed when the inventory position falls to 200 units. Then, the order will arrive when the inventory
level reaches zero.
However, if demand fluctuates about a mean of 200 units, shortages will occur approximately half the time. Because of the
high shortage costs, the manager would use either a larger reorder point, a larger order quantity, or both. In either case, the
manager will carry more inventory on average, which will result in a lower total shortage cost but a higher total holding cost.
A higher order quantity lets the manager order less frequently, thus incurring lower total ordering costs. However, the
appropriate choice is not clear. Simulation can test various reorder point/order quantity policies.

Before examining the spreadsheet simulation model, step through the logic of how this inventory system operates. Assume
that no orders are outstanding initially and that the initial inventory level is equal to the order quantity, Q. Therefore, the
beginning inventory position will be the same as the inventory level. At the beginning of the week, if any outstanding orders
have arrived, the manager adds the order quantity to the current inventory level.
Next, determine the weekly demand and check if sufficient inventory is on hand to meet this demand. If not, then the
number of lost sales is the demand minus the current inventory. Subtract the current inventory level from the inventory
position, set current inventory to zero, and compute the lost sales cost. If sufficient inventory is available, satisfy all demand
from stock and reduce both the inventory level and inventory position by the amount of demand.
The next step is to check if the inventory position is at or below the reorder point. If so, place an order for Q units and
compute the order cost. The inventory position is increased by Q, but the inventory level remains the same. Schedule a
receipt of Q units to arrive after the lead-time. Finally, compute the holding cost based on the inventory level at the end of
the week (after demand is satisfied) and the total cost.
Using Crystal Ball
Crystal Ball enhances your Excel model by allowing you to create probability distributions that describe the uncertainty
surrounding specific input variables. This model includes 52 probability distributions, referred to in Crystal Ball as
"assumptions." Each assumption describes the variation in the weekly demand. Each assumption cell is colored green and
is marked by an Excel note (mouse over the cell to view the note). To view the details of an assumption, highlight the cell
and either select Define Assumption from the Define menu or click on the Define Assumption button on the Crystal Ball
toolbar.

This model also includes one Crystal Ball forecast, shown in light blue. Forecasts are equations, or outputs, that you want
to analyze after a simulation. During a simulation, Crystal Ball saves the values in the forecast cells and displays them in a
forecast chart, which is a histogram of the simulated values. In this example, you want to analyze the Total Annual Costs
that result from a combination of hold, order, and short costs. To view a forecast with Crystal Ball, highlight the cell and
either select Define Forecast from the Define menu or click on the Define Forecast button on the toolbar.
When you run a simulation, Crystal Ball generates a random number for each assumption (based on how the assumption
has been defined) and places that new value in the cell. Excel then recalculates the model. You can test this by selecting
Single Step from the Run menu or clicking on the Single Step button on the toolbar.
After you run a simulation, you will see the forecast chart for Total Annual Costs. How realistic was your original, or base
case, calculation of $7,090? What is the certainty that you will incur this cost or more? What is the mean Total Annual Cost
value? What happens if you change the Order Quantity and Reorder Point values and re-run the simulation?
To view which of the assumptions had the greatest impact on the forecast, use the Sensitivity Chart. Do any of the weekly
demands have a greater effect on the forecast than the others?
Using OptQuest
Now that you have run at least one simulation, you can begin to address optimization using OptQuest. In this model, you
want to determine an optimal order quantity and reorder point that results in the lowest possible total annual costs, even
while you still have variability in the weekly demand.
OptQuest requires decision variables, which are model variables over which you have control. The two decision variables
defined in this model are Order Quantity and Reorder Point. Each decision variable is colored yellow and is marked by an
Excel note (mouse over the cell to view the note). To view the details of a decision variable, highlight the cell and either
select Define Decision from the Define menu or click on the Define Decision button on the Crystal Ball toolbar.

Start OptQuest from the Run menu and use the OptQuest Wizard to view the settings for the optimization. The problem has
no constraints and one objective: to minimize Total Annual Costs.
Run the optimization. For each optimization, OptQuest selects a new value within the defined range of each decision
variable (for example, an Order Quantity of 315) and runs a Crystal Ball simulation (for example, 2000 trials). OptQuest
then saves the mean Total Profit value. OptQuest then runs another simulation on a new set of decision variables.
OptQuest repeats this process, constantly searching for the lowest Total Annual Cost value until it either works through
every possible solution or reaches the end of the set running time.
As OptQuest runs, it uses multiple metaheuristic methods and techniques to analyze past results and improve the quality
and speed of its process. You can watch OptQuest's progress through the Performance Graph, which shows a flattened
line as it converges to an optimal result.
What is the best combination of order quantity and reorder point that results in the lowest mean total annual cost? Once
OptQuest is finished, you can copy the optimal results back to your spreadsheet with the Copy Best Solution to
Spreadsheet option in the Edit menu. Your spreadsheet now displays the optimal solution, and Crystal Ball displays the
Forecast Chart for the simulation from the best optimization. You can use OptQuest's Solution Analysis tool to review the
other quantities of products that resulted in low total annual cost.
Because this optimization used a step size of 5, you can fine-tune the solution by searching more closely around the best
solution using a smaller step size while also increasing the number of trials for better precision. This is a good practice,
since choosing too small a step size initially consumes a lot of time or, if time is restricted, OptQuest might not find a good
solution. Thus, as the number of decision variables and range of search increases, use larger step sizes and fewer trials
initially. Later, refine the search around good candidates.
Copyright Information
Copyright 2004, 2010, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names
may be trademarks of their respective owners.

Model

Inventory Simulation With Lost Sales

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

Target Inventory
Period
Initial Inventory
Lead time

Week
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

Beg
Inv Beg Order
Pos Inv Rec'd
385 385
670 285
570 185
0
470
85
1
755 370
0
655 270
0
555 170
1
455 455
0
740 355
0
640 255
0
540 155
1
440 440
0
725 340
0
625 240
0
525 140
1
425 425
0
710 325
0
610 225
0
510 125
1
410 410
0
695 310
0
595 210
0

385 units
400 units
385 units
2 weeks

Order Cost
Holding Cost
Lost Sales Cost

Units
End Lost
Order
Rec'd Dmd Inv Sales Placed?
0
100 285
0
1
0
100 185
0
0
0
100
85
0
0
385
100 370
0
1
0
100 270
0
0
0
100 170
0
0
385
100 455
0
0
0
100 355
0
1
0
100 255
0
0
0
100 155
0
0
385
100 440
0
0
0
100 340
0
1
0
100 240
0
0
0
100 140
0
0
385
100 425
0
0
0
100 325
0
1
0
100 225
0
0
0
100 125
0
0
385
100 410
0
0
0
100 310
0
1
0
100 210
0
0
0
100 110
0
0

$ 50
$0.20
$ 100

Optimize order quantity and reorder point to

minimize costs

Ending
Inv
Week
Pos
Due
670
4
570
0
470
0
755
7
655
0
555
0
455
0
740 11
640
0
540
0
440
0
725 15
625
0
525
0
425
0
710 19
610
0
510
0
410
0
695 23
595
0
495
0

Page 4

2,932

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

Hold
Cost
57.00
37.00
17.00
74.00
54.00
34.00
91.00
71.00
51.00
31.00
88.00
68.00
48.00
28.00
85.00
65.00
45.00
25.00
82.00
62.00
42.00
22.00

Total Annual Costs

$
700 $
-

$ 3,632

Order
Cost
$
50
$
$
$
50
$
$
$
$
50
$
$
$
$
50
$
$
$
$
50
$
$
$
$
50
$
$
-

Total
Cost
$ 107
$
37
$
17
$ 124
$
54
$
34
$
91
$ 121
$
51
$
31
$
88
$ 118
$
48
$
28
$
85
$ 115
$
45
$
25
$
82
$ 112
$
42
$
22

Short
Cost
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

Model

33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

A
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

B
495
780
680
580
480
765
665
565
465
750
650
550
450
735
635
535
435
720
620
520
420
705
605
505
405
690
590
490
775
675

C
110
395
295
195
480
380
280
180
465
365
265
165
450
350
250
150
435
335
235
135
420
320
220
120
405
305
205
105
390
290

D
1
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0

E
385
0
0
385
0
0
0
385
0
0
0
385
0
0
0
385
0
0
0
385
0
0
0
385
0
0
0
385
0
0

F
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100

G
395
295
195
480
380
280
180
465
365
265
165
450
350
250
150
435
335
235
135
420
320
220
120
405
305
205
105
390
290
190

H
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

I
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
1
0
0

Page 5

J
780
680
580
480
765
665
565
465
750
650
550
450
735
635
535
435
720
620
520
420
705
605
505
405
690
590
490
775
675
575

K
26
0
0
0
30
0
0
0
34
0
0
0
38
0
0
0
42
0
0
0
46
0
0
0
50
0
0
53
0
0

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

L
79.00
59.00
39.00
96.00
76.00
56.00
36.00
93.00
73.00
53.00
33.00
90.00
70.00
50.00
30.00
87.00
67.00
47.00
27.00
84.00
64.00
44.00
24.00
81.00
61.00
41.00
21.00
78.00
58.00
38.00

M
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

N
50
50
50
50
50
50
50
50
-

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

O
129
59
39
96
126
56
36
93
123
53
33
90
120
50
30
87
117
47
27
84
114
44
24
81
111
41
21
128
58
38

Model

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

order cost
Optquest

Page 6

Model

33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 7