CHAPTER 13

FORECASTING

Review and Discussion Questions

1. What is the difference between dependent and independent demand?

Independent demand is forecasted item demand that occurs separately from the demand for
other items. Dependent demand is calculated from the demand for another item. The
difference lies in the way the two demands are determined.

6. What strategies are used by supermarkets, airlines, hospitals, banks, and cereal manufacturers
to influence demand?

Supermarkets—several sales items, free giveaway items (such as a pound of butter or a loaf
of bread), an occasional “midnight madness” sale where the store is open late or even all
night.

Airlines—excursion rates, age rates (senior citizens, children, youth fares), charter flights,
off-season rates, exceptionally good meals (or no meals for further reduced prices), more
flights, tie-in with hotels or auto rental, and tour agencies for “package tours,” free
champagne, free stop-over at a third point, or new terminals.

Hospitals—patients generally go to the hospital recommended by their physician. Therefore,

hospitals offer free office space, nursing assistance, lab equipment, staff positions, and patient
billing to physicians. Hospitals frequently advertise their occupancy rate and room rates,
which tend to influence demand. Also, they could become a preferred provider organization.

Banks—free gifts for new accounts, free checking, free safety deposit box, free financial
advice, “club memberships,” and free use of “executive lounges” for depositors in various
size account ranges, community rooms for club meetings.

Cereal manufacturers—TV advertising, sponsorship of some youth events, free prizes in

cereal boxed, using prime display space.

7. All forecasting methods using exponential smoothing, adaptive smoothing, and exponential
smoothing including trend require starting values to get the equations going. How would you
select the starting value for, say Ft-1?

Starting values can be simply an average of the early periods, or a guess. If the starting value
is taken some period back (as opposed to starting to use the equations on very recent data) the
equation will have a chance to adjust as it is carried forward to today.

8. From the choice of simple moving average, weighted moving average, exponential
smoothing, and linear regression analysis, which forecasting technique would you consider
the most accurate? Why?

Of these four choices, the weighted moving average is the most accurate, since specific
weights can be placed in accordance with their importance. The other methods make

148
Chapter 13

assumptions, such as an average, straight line, or exponential curve. The weighted average
may be modified to any form. If a long time span is taken, however, the weighted average
can be cumbersome to use. In addition, as time periods pass, the user most likely would like
to change the weights. This would add to the difficulty of using the techniques for a large
number of applications such as forecasting demand for inventory items.

Problems
1.

b. ∑(y i − Yi ) 2
728205.128 = 269.85
S xy = i =1
=
n−2 12 − 2

Therefore, 3 standard errors of the estimate would be 3(269.85) = 809.55 or 810

2.

a. FJuly = .60(15) + .30(16) + .10(12) = 15.0

b. FJuly = (15 + 16 + 12) / 3 = 14.3

c. FJuly = FJune + α (AJune – FJune) = 13 + .2(15-13) = 13.4

149
 Forecasting

d.

x y xy x2
1 12 12 1
2 11 22 4
3 15 45 9
4 12 48 16
5 16 80 25
6 15 90 36
Total 21 81 297 91

x = 3.5
y = 13.5
∑xy −n x y
b= = 0.77
∑x
2
2
−n x
a = y −b x = 10.8

Y = a + bx = 10.8 + .77x

e. FJuly, where July is the 7th month.

Y = a + bx = 10.8 + .77(7) = 16.2

3.

a. Ft+1 = Ft + α (At – Ft), α = .20

Month Demand Forecast Absolute Deviation

January 100 80 20
February 94 84 10
March 106 86 20
April 80 90 10
May 68 88 20
June 94 84 10

Total 90

b. MAD = 90/6 = 15

150
Chapter 13

CHAPTER 15
INVENTORY CONTROL

Review and Discussion Questions

1. Distinguish between dependent and independent demand in a McDonald’s, in an integrated
manufacturer of personal copiers, and in a pharmaceutical supply house.

The key to the answer here is to consider what must be forecasted (independent demand), and,
given the forecast, what demands are thereby created for items to meet the forecasts
(dependent demand).

In a McDonald’s, independent demand is the demand for various items offered for sale—Big
Macs, fries, etc. The demand for Egg McMuffins, for example, needs to be forecasted. Given
the forecast, then, the demand for the number of eggs, cheese, Canadian bacon, muffins, and
containers can then be computed based on the amount needed for each Egg McMuffin.

The manufacturer of copiers is integrated, i.e., the parts, components, etc. are produced
internally. The demand for the number of copiers is independent (must be forecasted). Given
the forecast, the Bill of Materials is exploded to determine the amounts of raw materials,
components, parts, etc. that are needed (more on the BOM in chapter 16).

The pharmaceutical supply company is an extreme case where only end items are carried and
nothing is produced internally. The bill of materials is the end item and, therefore, the
independent demand (forecasted from customers) is the same as the dependent demand. One
might attempt to consider that when the demand for items occurs together, that this is similar
to a bill of materials. However, this is not a bill of materials, but rather a causal relationship
making it easier to forecast.

2. Distinguish between in-process inventory, safety stock inventory, and seasonal inventory.

In-process inventory consists of those items of materials components and partially completed
units that are currently in the production process.

Safety-stock inventory is set so that inventory is maintained to satisfy some maximum level of
demand. It could be stated that safety stock is that level of inventory between the minimum
expected demand and the desired level of demand satisfaction.

Seasonal inventory is that inventory accumulated to meet some periodic increase in demand.

3. OMIT

4. Under which conditions would a plant manager elect to use a fixed-order quantity model as
opposed to a fixed-time period model? What are the disadvantages of using a fixed-time
period ordering system?

Fixed-order quantity models–when holding costs are high (usually expensive items or high
deprecation rates), or when items are ordered from different sources.

151
 Forecasting

Fixed-time period models—when holding costs are low (i.e., associated with low-cost items,
low-cost storage), or when several items are ordered from the same source (saves on order
placement and delivery charges).

The main disadvantage of a fixed-time period inventory system is that inventory levels must
be higher to offer the same protection against stockout as a fixed-order quantity system. It
also requires a periodic count and closer surveillance than a fixed-order quantity system. A
fixed-order quantity system can operate with a perpetual count (keeping a running log of every
time a unit is withdrawn or replaced) or through a simple two-bin or flag arrangement wherein
a reorder is placed when the safety stock is reached. This latter method requires very little
attention.

5. OMIT

6. OMIT

7. Discuss the assumptions that are inherent in production setup cost, ordering cost, and carrying
cost. How valid are they?

Investigation of ordering and production setup cost will likely show that a single, unique cost
does not exist for each product, nor is it linearly related to the number of order (as implied in
the equations or inventory models). In the purchasing department, for example, an employee
is paid either a salary or an hourly rate for a normal work week. The cost for that employee is
sometimes divided among the number of items or orders for which he has responsibility,
resulting in an averaged or allocated cost for each order he places. However, when we
consider an inventory ordering cost based on the number or orders per year (as is done in most
inventory models), reducing the number of orders the individual places does not necessarily
decrease the net cost to the firm since his weekly pay remains the same. What happens is
really an increase in the ordering cost for each of the remaining items within his responsibility.

Nonlinearity of costs also occurs in production setups. Consider the time for making a setup
in preparation for a production run. Setup time is roughly based on an expected frequency of
making this particular product run. However, as the frequency increases, familiarity with the
setup allows some shaving of the setup time. Moreover, if the setup is repeated often, an
investment in specialized equipment or the construction of jigs may become warranted,
reducing the setup time even more.

The terms carrying or holding costs for maintaining goods in inventory include a multitude of
cost elements. To determine the nature and amounts of these costs can be a challenging feat.
Fortunately, total inventory cost curves tend to be dish shaped and can, therefore, tolerate
some error. The holding costs associated with insurance, obsolescence, and personnel who are
handling materials are extremely difficult to ascertain on an item-by-item basis, yet each
requires realistic analysis. Warehouse storage costs of an item, for example, may be based on
a ratio of its required square footage and the entire available warehouse space, but this may
not be an accurate representation since it is an allocation of cost rather than true cost. Take
the warehouse that is too large, or is used to stock products in an off season or depressed
period. Allocation based on a share of total warehouse cost will result in a high cost for
storage, when, in fact, excess storage space should create pressure for higher—not lower—
order quantities.

152
Chapter 13

In the simple inventory model, holding costs are based on the average inventory on hand.
“Average” inventory presumes that, as stock is depleted, other product lines will be moved in
to occupy the space. It may be that costs should be based on maximum inventory, especially
if these is an excess of space, or if the needs of an item are so specialized that no other
products can use the space (for example, due to environmental requirements). Each remaining
cost may be similarly challenged. Breakage, pilferage, deterioration, and insurance costs are
not constant but, rather, vary with inventory size. As the value of inventory increases,
insurance rates are lower, more refined handling procedures can be installed to reduce
breakage, some environmental control and maintenance can be used to reduce deterioration,
and better security procedures can reduce theft.

These challenges to determining true costs are not intended to discourage the use of inventory
models. The intent, rather, is to prevent the use of any model without clear knowledge of its
requirements and assumptions. Indeed, each application must consider the operating
conditions and needs of the firm. An appropriate model can then be developed in a fashion
similar to those covered in this chapter.

PROBLEMS:

3.
2 DS 2(1000 )25
Qopt = =
H 100 = 22.36 → 23

4. Service level P = .95, D = 5000, d = 5000/365, T = 14 days, L = 10 days, σ = 5 per day,

and I =150.
q = d (T + L ) + zσT +L − I

σ T +L = (T + L)σ 2 = (14 +10 )( 5) 2 = 24.495

From Standard normal distribution, z = 1.64

5000
q= (14 + 10 ) + 1.64 ( 24 .495 ) − 150 = 218.94 → 219
365

5. Service level P = .98, d = 150, T = 4 weeks, L = 3 weeks, σ = 30 per week, and I =500
pounds.
q = d (T + L ) + zσT +L − I

σ T +L = (T + L)σ 2 = ( 4 + 3)( 30 ) 2 = 79.4

From Standard normal distribution, z = 2.05

q = 150(4+3) + 2.05(79.4) – 500 = 712.77 → 713pounds

153
 Forecasting

6.

14.
Quantity range Cost (C) EOQ Feasible
Less than 100 pounds $20 per pound 219 pounds No
100 to 999 pounds $19 per pound 225 pounds Yes
1,000 or more pounds $18 per pound 231 pounds No
2 DS
Note: EOQ =
iC

Therefore, calculate total cost at Q=225, C=$19, and at Q=1000, C=$18

D Q 3000 225
TC Q =225 ,C =19 = DC + S + iC = 3000 (19 ) + 40 + (. 25 )19 = $58,068
Q 2 225 2

D Q 3000 1000
TC Q =1000 ,C =18 = DC + S + iC = 3000 (18 ) + 40 + (. 25 )18 = $56,370
Q 2 1000 2
The best order size is 1,000 units at a cost of $18 per pound.

154