How to Set Performance Targets

in Inventory Control

Dr. Everette S. Gardner, Jr.

1
 How to Set Performance Targets
in Inventory Control
 Clean up the parts list
 Develop a basic forecasting system
 Use the forecasts to classify parts (ABC+)
 Decide what to stock
 Decide where to stock it
 Divide the inventory into control groups:
 JIT, MRP, EOQ, Annual buy
 Develop benchmark performance measures

2
Clean Up the Parts List
 Code substitute items
 Ensure historical demand recorded against primary items
 Delete obsolete items (no longer used in current
product line)
 Do the part numbers apply to other customers or
products?
 Review items with no recent demand
 Stocking rules usually depend on demand in the last 6
months or the last year
 Negotiate with vendors to return parts for credit

3
Water Filtration Company:
Status of 23,192 inventory items
7,526 with no hits 2,200 obsolete
in 12 months 9% 2,928 substitute
33% items
13%

4,202 with
inadequate
6,336 active items demand to stock
27% 18%

4
The Importance of Forecasting
 Forecasts determine production and inventory
quantities
 MRP: Master schedule
 EOQ: Order quantity, leadtime demand, safety stock
 JIT: Requirements to internal and external suppliers

5
The Importance of Forecasting (cont.)
 Better forecast accuracy cuts inventory
investment. Example:
 Forecast accuracy is measured by the standard
deviation of the forecast error.
 Safety stocks are usually set at 3 times the standard
deviation
 If the standard deviation is cut by $1, safety stocks
are cut by $3

6
Forecasting Tools for Inventory Control
 Simple exponential smoothing
 Weighted-moving-average technique for stable items
 Highly recommended for repair parts demand
 Trend-adjusted exponential smoothing
 Estimates and projects growth (or decline) in demand
 Types of growth
 Exponential
 Linear
 Damped
 Both models are easily modified to handle
seasonal demands

7
Origins of the Damped Trend
 Reference
 Gardner & McKenzie, Management Science, 1985
 Operational requirement
 Automatic forecasting system for military repair and
maintenance parts
 Theory
 Lewandowski, IJF, 1982 (M1-Competition)
Trend extrapolation should become more
conservative as the forecast horizon increases.

8
The Damped Trend

1) Error = Actual demand – Forecast

2) Level= Forecast + Weight1(Error)

3) Trend = (Previous trend) + Weight2(Error)

4) Forecast for t+1= Level + Trend

5) Forecast for t+2 = Level + Trend + 2 Trend

.
.
9
Automatic Forecasting with the Damped Trend

 Constant-level data
 Forecasts emulate simple smoothing
 Consistent trend
 Forecasts emulate Holt’s linear trend
 Erratic trend
 Forecasts are damped

10
 Automatic Forecasting with the Damped Trend
In constant-level data, the forecasts emulate simple
exponential smoothing:

36

35

34

33

32

31

30

29

28

27

26

11
 Automatic Forecasting with the Damped Trend
In data with a consistent trend and little noise, the
forecasts emulate Holt’s linear trend:

60

55

50

45

40

35

30

25

20

12
 Automatic Forecasting with the Damped Trend
When the trend is erratic, the forecasts are damped:

50 Saturation level

45

40

35

30

25

20

13
 Automatic Forecasting with the Damped Trend
The damping effect increases with the level of noise in
the data:

50

Saturation level
45

40

35

30

25

20

14
 11-Oz. Corn chips
Monthly Inventory and Sales

Actual Inventory

Sales

15
 Damped-trend performance
11-oz. Corn chips
$500,000

Outlier Actual
Forecast
$450,000

$400,000

$350,000

$300,000

$250,000

$200,000

16
Investment analysis: 11-oz. Corn chips
Forecast annual usage $4,138,770
Economic order quantity $318,367
Standard deviation of forecast errors $34,140

Nbr. shortages
per 1,000 Probability Safety Order Maximum
order cycles of shortage stock quantity investment
100.0000 0.1000 $43,758 $318,367 $362,125
50.0000 0.0500 $56,167 $318,367 $374,534
1.0000 0.0010 $105,510 $318,367 $423,877
0.0100 0.0000 $145,601 $318,367 $463,968
0.0001 0.0000 $177,496 $318,367 $495,863

17
 Safety stocks vs. shortages
$200,000

$180,000
Target
$160,000

$140,000
Safety stock

$120,000

$100,000

$80,000

$60,000

$40,000

$20,000

$0
0 10 20 30 40 50 60 70 80 90 100

Shortages per 1,000 order cycles

18
 Safety stocks vs. forecast errors
$200,000
Safety stock
$150,000

$100,000

$50,000

$0

($50,000) Forecast errors

($100,000)

($150,000)

($200,000)

19
 11-Oz. Corn chips
Target vs. actual packaging inventory
$2,500,000
Actual Inventory
from subjective
Actual Inventory
$2,000,000 forecasts
from subjective
forecasts

$1,500,000

$1,000,000

$500,000

$0
Target maximum
inventory based on
damped trend Month
Monthly Usage

20
How to forecast regional demand
 Forecast total units with the damped
trend
 Forecast regional percentages with simple
exponential smoothing

21
Regional sales percentages: Corn chips
50%

East
40%

30% South

North
20%

West
10%

0%
Mar Jun Sep Dec Mar Jun Sep Dec

22
Target Inventory Analysis

 Actual inventory based on subjective decisions:

$182.6 million
 Target inventory based on the damped trend and
EOQ/Safety stocks:
$135.0 million
 Projected savings:
$47.6 million

23
 Water Filtration Company:
Inventory Classification
Active Items (some demand in past year)

Total nbr. Percent of Total sales Percent of

Class Class Limits of items items forecast forecast
A > $36,000 364 4% $11,743,610 60%

B $600 - $35,999 4,232 46% $7,316,999 38%

C < $600 4,668 50% $365,605 2%

Totals 9,264 100% $19,426,214 100%

24
 Water Filtration Company:
Inventory Classification

Number
of Percent
Class Definition items of items
Active Items A Active items: Sales > $36,000 364 1.51%
B Active items: Sales $600-$35,999 4,232 17.6%
C Active items: Sales < $600 4,668 19.42%
Slow Movers Z Zero forecast items: no hits in 6 mon. 1,548 6.44%
D Disposal items: no hits in 12 mon. 11,526 47.95%
Buys not based X One-time buys 146 0.61%
on demand N New items: established in last 12 mon. 968 4.03%
Miscellaneous F Free (no-cost) items 27 0.11%
P Problem items (missing data) 560 2.33%

Totals 24,039 100.0%

25
What to Stock?
 Compare costs
 Cost to stock = (Avg inventory balance x holding rate)
+ (number of stock orders x transportation cost)
 Cost to not stock = Number of customer orders x
transportation cost

 Transportation costs for not stocking may be both in-

and out bound, depending on whether we choose to
drop-ship from the vendor

26
What to Stock? (cont.)
 Simplify decisions using “hit” rules
 A “hit” is one customer order for any number of units.
Cost comparisons usually result in minimum number of
hits that must occur before it is economical to stock.
 Example:
 Class A 6 hits in 6 months
 Class B 4 hits in 6 months
 Class C 3 hits in 12 months

27
 Water Filtration Company:
Inventory Status as of July, 2009

Nbr
Minimum Excess Nbr items items not
Class Definition hits to stock stock qty. stocked stocked Total
Active items: Sales >
A 6 in 6 mon. > 6 mon. 290 74 364
$36,000
Active Items: Sales $600 -
B 3 in 6 mon. > 6 mon. 3,108 1,124 4,232
$35,999

C Active items: Sales < $600 4 in 12 mon. > 12 mon. 2,938 1,730 4,668

Zero forecast items: no hits

Z - All - 1,548 1,548
in 6 mon.
Disposal items: no hits in 12
D - All - 11,526 11,526
mon.

X One-time buys - All - 146 146

New items: established in Same as

N - 170 628 968
last 12 mon. ABC

Totals 6,506 16,776 23,452

28
Where to Stock?
 Centralized order entry is mandatory
 Apply the hit rules by location
 This automatically tailors the range of stock to the
customer base at each location
 Must designate who suppliers whom when a hit occurs at
a non-stocking location

 Recognize that consolidating stocks makes

dramatic reductions in total inventory investment

29
Who Supplies Whom?

Stocking
LA forecast FL forecast CA/OR forecast
Warehouses

1 LA + FL LA + CA/OR FL only 0

2 LA + CA/OR LA + FL 0 CA/OR only

3 FL + CA/OR 0 LA + FL CA/OR only

30
 Effects of Consolidating Inventories:
Manufacturer of Communication Systems

6.5
Investment
(millions)

5.5

4.5

3.5

2.5

1.5
1 2 3 4
Number of Warehouses

31
 Monthly Item Migration Processing
Original Class Result of monthly review Action
N (new items) Enough hits in last 6 months to Generate forecast, change N
stock. to A,B, or C
At least 1 hit in last 6 months, Generate forecast, change N
but not enough to stock to A, B, or C
No hits in last 6 months, but Set forecast = zero, change
hits occurred 7-12 months ago N to Z (zero forecast)
No hits in last year Change N to D (disposal)

A, B, or C No hits in last 6 months Set forecast = zero,

(active items) Change A, B, or C to Z (zero
forecast)
Z (zero forecast) No hits in last 12 months Change Z to D (disposal)

1 or more hits Change Z to A, B, or C

D (disposal items) 1 or more hits Generate forecast,

Change D to A, B, or C

32
 Control Inventory Production Lead-time
System Class Schedule Behavior

JIT A, B Level Certain

MRP A, B Variable Reliable

EOQ/Safety
A, B Variable Variable
Stock

Annual buy C Any Any

33
The Economic Order Quantity (EOQ)
 Controls
 Cost per order
 Holding rate (% of inventory value)
 The EOQ
 Increases with the order cost
 Decreases with the holding rate
 Do not treat order cost and holding rate as fixed
values. Instead, do what-if analysis to hit target
values for
 Inventory investment on the balance sheet
 Stock replenishment workload

34
Inventory tradeoffs for Class B items
Number of Class B active items 3,600
Sum of annual demand forecasts $14,337,666
Sum of monthly demand forecasts $1,194,806
Inventory carrying cost 20.00%

EOQ in weeks of
Order Maximum Investment Number of annual buys
stock
Cost
Total Avg. per item Total Avg. per item Min Max
$5 $1,428,356 $397 28,567 7.9 2.6 14.9
10 2,020,000 $561 20,200 5.6 3.6 21.0
15 2,473,985 $687 16,493 4.6 4.5 25.8
20 2,856,711 $794 14,282 4.0 5.2 29.8

35
 Communications Systems Manufacturer:
Target inventory values for fiscal 2009

Control Number of Maximum Number of

Class
System Items Investment Annual Buys

A JIT 1,412 $4,945,000 16,944

B EOQ 3,999 $10,820,000 15,996

C Annual Buy 8,688 $2,004,000 8,688

Totals 14,099 $17,769,000 41,628

36
 Water Filtration Company: Inventory
Values
Class Definition Nbr items On-hand Excess On-order
A,B,C Active items: On-hand ok 5,754 $3,884,064 $0 $1,958,932

A,B,C Active items: On-hand excessive 584 $887,048 $491,960 $23,220

Not enough hits to stock or zero
A,B,C,Z 3,878 $943,950 $902,774 $27,398
forecast
Disposal items: no hits in 12
D 2,336 $458,968 $458,160 $121
mon.
X One-time buys 12 $5,128 $5,128 $0
New items: established in last 12
N 404 $127,124 $121,534 $93,362
mon.

Totals 12,968 $6,307,626 $1,979,556 $2,103,033

37
 Annual Purchasing Workload Estimates
New System: Old System:
Stock buys Stock buys
Annual Annual Annual Annual
Nbr items reorders reorders Nbr items reorders reorders per
Class stocked per item per class Class stocked per item class
A (JIT) 290 12 3,480 A(JIT) 290 12 3,480
B(EOQ) 3,108 4 12,432 B(EOQ) 3,108 12 37,296
C(Ann.) 2,938 1 2,938 C(Ann.) 2,938 12 35,256
Total 18,850 Total 76,032

New system: Worst case estimate of Summary:

number of drop-ship buys Workload comparisons
Max# hits Number Old system New system
Nbr items for no- of drop-
Class not stocked stock item ship buys Stock buys 76,032 18,850
A (JIT) 74 11 814 Drop-ship buys ? 11,624
B(EOQ) 1124 5 5,620 New items, one time buys 600 600
C(Ann.) 1730 3 5,190 Redistribution actions ? 1,600
Total 11,624 Total 76,632 32,674

38
 Monthly Inventory Performance Report
Lead-times
Lead-time exceptions 4.5%
Average lead-time 3.20
Stock Levels
Target $19,900,878
Actual $23,422,012
Excess $3,521,134
Stock Status
Items in stock 92.20%
Items out of stock 7.80%
Items Out of Stock
With open order 82.2%
No open order 17.8%
Customer Orders
Percent shipped complete 94.3%
Warehouse Refusals
Percent occurrence 2.8%

39
Conclusions
 Forecasting drives any inventory control
system
 Standard ABC classification doesn’t go far
enough
 Decision rules for what/where to stock must be
established early

40
Conclusions (cont.)
 Performance measurement is essential to:
 Justify a new system
 Tailor the system to the inventory

 Track progress

41
Conclusions (cont.)
 The best inventory system is likely to be a
hybrid of:
 JIT
 MRP

 EOQ

 Annual buy

42