Dr. S.

Venkataramanaiah (S Venkat)
OM & QT Area IIM Indore Prabhandh Shikhar Rau-Pithampur Road Indore- 453 331 Email: svenkat@iimidr.ac.in

3/22/2009

PGP1-OM 2 -S Venkat

Inventory Management- Outline

Importance Cost components Basic EOQ Model Q and P Models Safety stock determination J & G Distributors Case Conclusions
PGP1-OM 2 -S Venkat 2

3/22/2009

Objectives
To Understand The underlying concepts of inventory management and importance of analytical tools Role of relevant costs and models Issues in setting of safety levels Applicability of models and features

3/22/2009

PGP1-OM 2 -S Venkat

Logistics Costs for the US Economy

Avg cost of US inventory is around 30-35% of value (India may be around 40%)

Organized Sector Transportation Inventory Warehousing Packaging Losses Total Cost Logistics Cost ($b) GDP ($ billion)

US: 1990 50% 20% 24% 6% 11.4% 659 5,800

US: 2001 60%

US: 2002 63%

40%

37%

9.5% 957 10,080

8.7% 910 10,470

Source: CASS Information Systems, 2002

3/22/2009 PGP1-OM 2 -S Venkat 4

Introduction
Inventory costs (IC) are due to obsolescence, insurance, opportunity cost etc Reduction in inventory adds to bottom line efficiency Profit can be increased by reduction in inventory Types of inventory-RM, FG, components, supplies and WIP. In services-tangible goods and supplies necessary for administering the services

3/22/2009

PGP1-OM 2 -S Venkat

Introduction
Inventory analysis aims are specifying when and how much to order Long term relationship between vendor and supplier, diversity of operations and technology changed the inventory mgt issues When and how much to order changes to when and how much to deliver and where?

3/22/2009

PGP1-OM 2 -S Venkat

Inventory Definition
Inventory is the stock of any item or resource used in an organization and can include: raw materials, finished products, component parts, supplies, and work-in-process etc An inventory system is the set of policies and controls that monitor levels of inventory and determines what levels should be maintained, when stock should be replenished, and how large orders should be (size)
3/22/2009 PGP1-OM 2 -S Venkat 7

Purposes of Inventory
1. To maintain independence of operations 2. To meet variation in product demand 3. To provide a safeguard for variation in raw material delivery time 4. To allow flexibility in production scheduling 5. To take advantage of economic purchase-order size (EOQ) 6. Inventory is costly and large amounts are generally undesirable 7. Long cycle times are due to large amounts of inventory 3/22/2009 PGP1-OM 2 -S Venkat

Inventory Costs
Holding (or carrying) costs Costs for storage, handling, insurance, pilferage, breakage, depreciation, taxes, opportunity cost of capital etc Setup/changeover costs Costs for arranging specific equipment setups, change of SOP, paper work etc Ordering costs Costs of placing an order, tracking of orders, follow-up etc Shortage costs Costs of canceling an order, etc Objective is to minimise all the above costs
3/22/2009 PGP1-OM 2 -S Venkat 9

Costs in Inventory Planning

Carrying Cost
Interest for short-term borrowals for working capital Cost of stores and warehousing Administrative costs related to maintaining and accounting for inventory Insurance costs, cost of obsolescence, pilferage, damages and wastage All these costs are directly related to the level of inventory

3/22/2009

PGP1-OM 2 -S Venkat

10

Costs in Inventory Planning

Ordering Cost
Search and identification of appropriate sources of supply Price negotiation, contracting and purchase order generation Follow-up and receipt of material Eventual stocking in the stores after necessary accounting and verification A larger order quantity will require less number of orders to meet a known demand and vice versa
Cost of carrying and cost of ordering are fundamentally two opposing cost structures in inventory planning
3/22/2009 PGP1-OM 2 -S Venkat 11

Costs in Inventory Planning

Shortage Cost
Costs arising out of pushing the order back and rescheduling the production system to accommodate these changes Rush purchases, uneven utilisation of available resources and lower capacity utilisation Missed delivery schedules leading to customer dissatisfaction and loss of good will The effects of shortage are vastly intangible, it is indeed difficult to accurately estimate
3/22/2009 PGP1-OM 2 -S Venkat 12

A graphical representation
Sum of the two costs

EOQ Model

Cost of Inventory

Total cost of carrying

Minimum Cost Total cost of ordering

Economic Order Qty.

3/22/2009

Level of Inventory
PGP1-OM 2 -S Venkat 13

Independent vs. Dependent Demand

Independent Demand (Demand for the final endproduct or demand not related to other items)

Finished product

E(1)

Component parts
3/22/2009 PGP1-OM 2 -S Venkat

Dependent Demand (Derived demand items for component parts, subassemblies, raw materials, etc)
14

Inventory systems
Period/ frequency Lead time Multiple High value Dependent Demand Newspaper, crackers, shoes, and uniforms etc
3/22/2009

Single Components parts low value

(0, prob, det)

Uniform, high volume, commodity, Garments etc

PGP1-OM 2 -S Venkat

Independent
15

Basic Fixed-Order Quantity Model and Reorder Point Behavior

1. You receive an order quantity Q. 4. The cycle then repeats.

Number of units on hand

Q R

2. You start using them up over time.

L
Time

R = Reorder point Q = Economic order quantity L = Lead time

3/22/2009

3. When you reach down to a level of inventory of R, you place next order for Q
16

PGP1-OM 2 -S Venkat

Issues in using EOQ Model

Model assumptions
1. The demand is known with certainty 2. Demand is continuous over time 3. There is an instantaneous replenishment of items 4. The items are sourced from an outside supplier 5. Assumptions about order quantity a) There are no restrictions in the quantity that we can order b) There are no preferred order quantities for the items c) No price discount is offered when the order size is large
3/22/2009 PGP1-OM 2 -S Venkat 17

Issues in using EOQ Model

Model assumptions
Despite this, the EOQ model could be applied with suitable modifications because it is robust Assumptions 3, 4 and 5 can be addressed with required modifications Relaxing assumption 1 will result in shortages due to difficulty in estimating demand

3/22/2009

PGP1-OM 2 -S Venkat

18

Basic Fixed-Order Quantity (EOQ) Model Formula

Total Annual = Cost Annual Annual Annual Purchase + Ordering + Holding Cost Cost Cost
TC=Total annual cost D =Demand C =Cost per unit Q =Order quantity S =Cost of placing an order or setup cost R =Reorder point L =Lead time H=Annual holding and storage cost per unit of inventory
19

D Q TC = DC + S + H Q 2
3/22/2009 PGP1-OM 2 -S Venkat

Deriving the EOQ

Using calculus, we take the first derivative of the total cost function with respect to Q, and set the derivative (slope) equal to zero, solving for the optimized (cost minimized) value of Qopt
QOPT = QOPT = 2DS 2(Annual Demand)(Order or Setup Cost) 2DS 2(Annual Demand)(Order or Setup Cost) = = H Annual Holding Cost H Annual Holding Cost
Reorder p oint, R = d L Reorder p oint, R = d L
_
__

We also need a reorder point to tell us when to place an order

3/22/2009

d = average daily demand (constant) L = Lead time (constant)

PGP1-OM 2 -S Venkat 20

10

EOQ Example 1
Given the information below, what are the EOQ and reorder point?

Annual Demand = 1,000 units Number of demand days per year considered as 365 Cost to place an order = $10 Holding cost per unit per year = $2.50 Lead time = 7 days Cost per unit = $15

3/22/2009

PGP1-OM 2 -S Venkat

21

EOQ-Example 1 Solution
Q OPT =
d =

2DS = H

2(1,000 )( 1 0 ) = 89.443 un its or 90 units 2.50

1,000 units / year = 2.74 units / day 365 days / year

_

Reorder point, R = d L = 2.74units / day (7days ) = 19.18 or 2 0 u n i t s

In summary, you place an optimal order of 90 units. In the course of consumption of the units to meet demand and when there are 20 units left, place the next order of 90 units.

Any observatio ns?

3/22/2009 PGP1-OM 2 -S Venkat 22

11

EOQ-Example 2
Determine the economic order quantity and the reorder point for the following Annual Demand = 10,000 units Demand days per year considered as 365 Cost to place an order = $10 Holding cost per unit per year = 10% of cost per unit Lead time = 10 days Cost per unit = $15

3/22/2009

PGP1-OM 2 -S Venkat

23

EOQ-Example 2 Solution
Q OPT = d=
_

2DS = H

2(10,000 ) (10) = 3 6 5 . 1 4 8 u nits, o r 366 units 1.50

10,000 units / year = 27.397 units / day 365 days / year

R = d L = 2 7 . 3 9 7 u n its / d a y ( 1 0 d a ys) = 2 7 3 . 9 7 o r 2 7 4 u n i t s

Place an order for 366 units and start consuming the units, when on hand inventory is 274 units, place the next order of 366 units.
Any observations between example 1 and 2
3/22/2009 PGP1-OM 2 -S Venkat 24

12

Inventory Systems
Multi-Period Inventory Models- re-ordering systemsensures availability throughout the period, multiple orders placed.

Fixed-Order Quantity (Q) Models

Event triggered, EOQ system, event may take place any time depending on demand, orders are placed when inventory reaches ROL, perpetual system, has no review period, low avg inv, more time consuming, requires quick response, Q and R needs to be determined, variable order cycle time and fixed order quantity etc. Ex. running out of stock, spares, high priced, critical, important items)
3/22/2009 PGP1-OM 2 -S Venkat 25

Common Demand Assumptions

Characterized by two moments: mean and standard deviation Demand is independent in non-overlapping time increments Demand over t time units has mean t and standard deviation t

3/22/2009

PGP1-OM 2 -S Venkat

26

13

Q R Model
Continuous review Fixed order quantity Q Re-order when inventory on-hand & onorder drops to or below R Assume fixed replenishment time L

3/22/2009

PGP1-OM 2 -S Venkat

27

Continuous Review (Q) System

An illustration
Q Inventory Position Physical Inventory

Inventory Level

ROP

Mean Demand during LT SS Safety Stock

3/22/2009

Time PGP1-OM 2 -S Venkat

28

14

How to set R?
R should cover the lead-time demand with high probability Lead-time demand has mean L and standard deviation L

R = L + z L = Expected lead-time demand + safety stock

3/22/2009 PGP1-OM 2 -S Venkat 29

Service Measures
Type 1 probability of stock out during a replenishment event Type 2 fill rate, equal to percent of all demand met from stock

3/22/2009

PGP1-OM 2 -S Venkat

30

15

How to set R?
z is safety factor Suppose demand is normally distributed For Type I service (from normal table):
z = 1.64 provides 0.95 coverage probability z = 2 provides 0.98 coverage probability z = 3 provides 0.999 coverage probability

3/22/2009

PGP1-OM 2 -S Venkat

31

How to set R?
Suppose demand is normally distributed For Type II service:
Fill rate = Expected Backorders per Cycle Q

Expected Backorders per Cycle z = 1.64 z=2 z=3

3/22/2009

0.0211 0.0085 0.0004

PGP1-OM 2 -S Venkat

L L L
32

16

P- Model with Safety Stock

Fixed-Time Period (P) Models-Time triggered (Ex. Monthly sales call by sales representative
q = Average demand + Safety stock Inventory currently on hand
q = d (T + L) + Z T + L - I Where : q = quantitiy to be ordered T = the number of days between reviews L = lead time in days d = forecast average daily demand z = the number of standard deviations for a specified service probabilit y T + L = standard deviation of demand over the review and lead time I = current inventory level (includes items on order)
3/22/2009 PGP1-OM 2 -S Venkat 33

P Model: Determining the Value of sT+L

T+ L = T+ L =

((
TT + L +L i == 1 i 1

ddi i

))

22

S iin ccee eeaacch d aay iiss iin d eep een d een tt aan d d iiss cco n ssttaan tt,, S n h d y n d p nd n nd d on n TT++LL = = (T + L ))dd22 (T + L

The standard deviation of a sequence of random events equals the square root of the sum of the variances
3/22/2009 PGP1-OM 2 -S Venkat 34

17

Periodic Review (P) System

An illustration
Q 2R Q 3R Inventory Position Physical Inventory Order Up to Level QR S

Inventory Level
SS Safety Stock

R L

2R

3R

3/22/2009

PGP1-OM 2 -STime Venkat

35

P Model- Example
Given the information below, how many units should be ordered? Average daily demand for a product is 20 units. The review period is 30 days, and lead time is 10 days. Management has set a policy of satisfying 96 percent of demand from items in stock. At the beginning of the review period there are 200 units in inventory. The daily demand standard deviation is 4 units.
3/22/2009 PGP1-OM 2 -S Venkat 36

18

P Model: Solution
T+ L = (T + L) d 2 =

(30 + 10 )( 4) 2 = 25.298

The value for z is found by using the Excel NORMSINV function, or from statistical tables.
From statistical tables, the z value at probability of 0.96 is 1.75
q = d (T + L) q = 20(30 q = 800 + Z
T +L

- I 298) - 200 or 645 units

+ 10) + 44.272

+ (1.75)(25. - 200

= 644.272,

So, to satisfy 96 percent of the demand, you should place an order of 645 units at this review period
3/22/2009 PGP1-OM 2 -S Venkat 37

P System- Example
P System Using the time between orders derived from the EOQ model as the basis for review period Review period, R = 2 weeks Mean demand during (L + R), ( L+R ) = 200*(2 + 2) = 800 Standard deviation of demand during (L + R), (L+R ) = 2 + 2 * 40 = 80 For a service level of 95%, SS =Z *(L +R)= 1.645*80 = 131.6 132 Order up to level, S = ( L+R )+Z *(L +R)= 800 + 132 = 932

The P system can be designed as follows: The inventory level in the system is reviewed every two weeks and an order is placed to restore the inventory level back to 932 units. This will ensure a service level of 95%.

3/22/2009

PGP1-OM 2 -S Venkat

38

19

Base Stock Model

Periodic review with r = review period Review period often dictated by logistics or transportation considerations, by production scheduling cycle etc Fixed time between orders Size order to bring on-hand & on-order inventory up to base stock (or order-up-to point). Assume fixed replenishment time L
3/22/2009 PGP1-OM 2 -S Venkat 39

Comparison
Q R model Fixed lot Variable time between orders A items Large setup costs Lot size dictated by process considerations Less inventory

Base stock model Variable lot Fixed time between orders B/C items Shared setup costs Review period dictated by process considerations

3/22/2009

PGP1-OM 2 -S Venkat

40

20

Whats the impact on inventory & service?

Improving the forecast? Reducing the lead-time? Increasing the frequency of replenishments? Reducing the lot size? Better and more timely information? Reducing variability?

3/22/2009

PGP1-OM 2 -S Venkat

41

Computing safety stock-Using Normal

Distribution
Let the demand during lead time follow a Normal distribution Mean demand during lead-time = ( L ) Standard deviation of demand during lead-time = (L ) Desired service level = (1 ) The probability of a stock out = Z * Standard normal variate corresponding to an area of (1 ) covered on the left side of the normal curve = Z
L

Safety stock (SS) is given by SS = Z * ( L )

, 3/22/2009 PGP1-OM 2 -S Venkat 42

21

Periodic & Continuous Review Systems: A comparison

Criterion How Much to Order Continuous Review (Q) System Fixed Order Qty: Q Periodic Review (P) System S = ( L + R ) + Z * ( L+ R) QR = S - IR When to Order Safety Stock Salient Aspects ROP = (L ) + Z * ( L ) SS = Z * ( L) Implemented using two bin system Suited for medium and low value items Every R Periods SS = Z * ( L+R ) More safety stock More responsive to demand Ease of implementation

3/22/2009

PGP1-OM 2 -S Venkat

43

Price-Break Model
Based on the same assumptions as the EOQ model, the price-break model has a similar Qopt formula:

Q OPT =

2DS 2(Annual Demand)(Order or Setup Cost) = iC Annual Holding Cost

i = percentage of unit cost attributed to carrying inventory C = cost per unit Since C changes for each price-break, the formula above will have to be used with each price-break cost value
3/22/2009 PGP1-OM 2 -S Venkat 44

22

Price-Break Example
A company has a chance to reduce their inventory ordering costs by placing larger quantity orders using the price-break order quantity schedule below. What should their optimal order quantity be if this company purchases this single inventory item with an e-mail ordering cost of $4, a carrying cost rate of 2% of the inventory cost of the item, and an annual demand of 10,000 units?
Order Quantity(units) Price/unit($) 0 to 2,499 $1.20 2,500 to 3,999 1.00 4,000 or more .98
3/22/2009 PGP1-OM 2 -S Venkat 45

Price-Break Example (Contd..)

First, plug data into formula for each price-break value of C
Annual Demand (D)= 10,000 units Cost to place an order (S)= $4 Carrying cost % of total cost (i)= 2% Cost per unit (C) = $1.20, $1.00, $0.98

Next, determine if the computed Qopt values are feasible or not

Interval from 0 to 2499, the Qopt value is feasible Interval from 2500-3999, the Qopt value is not feasible Interval from 4000 & more, the Qopt value is not feasible

Q OPT = Q OPT = Q OPT =

2DS = iC 2DS = iC 2DS = iC

2(10,000)( 4) = 1,826 units 0.02(1.20) 2(10,000)( 4) = 2,000 units 0.02(1.00) 2(10,000)( 4) = 2,020 units 0.02(0.98)
46

3/22/2009

PGP1-OM 2 -S Venkat

23

Price-Break Example (Contd..)

Since the feasible solution occurred in the first pricebreak, it means that all the other true Qopt values occur at the beginnings of each price-break interval. Why? Because the total annual cost function is a u shaped function So the candidates for the price-breaks are 1826, 2500, and 4000 units

Total annual costs

0
3/22/2009

1826

2500

4000

Order Quantity
47

PGP1-OM 2 -S Venkat

Price-Break Example (Contd..)

Use true Qopt values into the total annual cost function to determine the total cost under each price-break

TC = DC +

D Q S + iC Q 2

TC(0-2499)=(10000*1.20)+(10000/1826)*4+(1826/2)(0.02*1.20) = $12,043.82 TC(2500-3999)= $10,041

TC(4000&more)= $9,949.20 Finally, we select the least costly Qopt, which is this problem occurs in the 4000 & more interval. In summary, optimal order quantity is 4000 units
3/22/2009 PGP1-OM 2 -S Venkat 48

24

Inventory Systems
Single-Period Inventory Model One time purchasing decision (Ex. vendor selling tshirts at a game/event, crackers, NY greetings, News paper boy, fashion items, wedding cards, banners, election related materials, mkt campaigns, R & D projects, turn key projects, etc.) Seeks to balance the costs of inventory overstock and under stock- i.e., loss due to unsold items Vs opportunity loss due to excess demand Non-repetitive, long lead time, limited shelf life, Used in yield mgt cases-airlines, hotels, etc. Used in manf and service contexts- uncertainty in demand estimation is high, cost of slack resources is high (perishability).
3/22/2009 PGP1-OM 2 -S Venkat 49

Single-Period Inventory Model

P (Co ) (1 P )(C u ) P Cu Co + Cu
This model states that we should continue to increase the size of the inventory/ order as long as the probability of selling what is ordered is less than or equal to Cu/(Co+Cu)

Where : Co = Cost per unit of demand over estimated

Cu = Cost per unit of demand under estimated P = Probability that the unit will be sold
3/22/2009 PGP1-OM 2 -S Venkat 50

25

Single Period Model-Example 3

College basketball team is playing in a tournament. Based on past experience one can sell on average 2,400 shirts with a standard deviation of 350. In this process the seller can make $10 on every shirt, but lose $5 on every shirt not sold. How many shirts should the seller make for the game? Cu=$10 and Co = $5; P =$10/($10+$5)=0.667 Z.667 = .432 (use NORMSDIST(.667) or Appendix E) therefore the seller need 2,400 + .432(350) = 2,551 shirts
3/22/2009 PGP1-OM 2 -S Venkat 51

Single Period Model- Example 3

P 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 0.97 Z -0.67 -0.52 -0.39 -0.25 -0.13 0 0.126 0.253 0.385 0.524 0.674 0.842 1.036 1.282 1.645 1.881 avg sd 2400 350 OQ=avg+z*sd % change in OQ 2164 -9.836 2216 -7.648 2265 -5.619 2311 -3.695 2356 -1.833 2400 0 2444 1.8326 2489 3.6946 2535 5.6193 2584 7.6475 2636 9.8363 2695 12.274 2763 15.115 2849 18.689 2976 23.987 3058 27.428
Rel bet OQ and P % change in OQ 30 25
% Change in OQ

20 15 10 5 0
0. 3 0.3 5 0. 4 0.4 5 0. 5 0.5 5 0. 6 0.6 5 0. 7 0.7 5 0. 8 0.8 5 0. 9 0.9 5 0.9 7 0.2 5

-5

-10 Prob (P) Rel bet P and Z Z

2.5 2 1.5
Z value

1 0.5 0
0.4 0. 45 0.5 0. 55 0.6 0. 65 0.7 0. 75 0.2 5 0.3 0.3 5 0.8 0.8 5

-0.5 -1

3/22/2009

PGP1-OM 2 -S Venkat

Prob (P)

0.9 0.9 5 0.9 7

52

26

Single Period Model-Example 4

vYield management-Hotel reservations-Number of overbookings vThe last minute cancellations: Avg 5 and SD: 3 vAvg room rate: $80, vWhen the hotel is overbooked, policy is to find a room in a nearby hotel and to pay for the customer and it costs $200 vHow many rooms that hotel should overbook? vCost of underestimate of cancellations: $80 and overestimation of cancellations: $200

3/22/2009

PGP1-OM 2 -S Venkat

53

Single Period Model-Example 4

Cu = $80 and Co = $200; P = $80 /($200 + $80) = 0.2857 Z.2857 = -0.56599 (use NORMSDIST (.2857) /Appendix E) -ve sign denotes overbooking should be less than avg -0.56599*3= 1.698 (appr. 2) hence 2 less than Avg (5) should be overbooked (3 over bookings)
Rel bet No of overbookings and TC
1400 1200 1000
Total cost

Total cost

800 600 400 200 0 0 1 2 3 4 5 6 7 8 9 10 No of over bookings

3/22/2009

PGP1-OM 2 -S Venkat

54

27

Miscellaneous Systems: Optional Replenishment System

Maximum Inventory Level, M

q=M-I Actual Inventory Level, I

M I

Q = minimum acceptable order quantity

If q > Q, order q, otherwise do not order any.

3/22/2009 PGP1-OM 2 -S Venkat 55

Miscellaneous Systems: Bin Systems

Two-Bin System

Full

Empty

Order One Bin of Inventory

One-Bin System

Order Enough to Refill Bin

3/22/2009

Periodic Check

PGP1-OM 2 -S Venkat

56

28

Inventory Accuracy and Cycle Counting

Inventory accuracy refers to how well the inventory records agree with physical count Cycle Counting is a physical inventorytaking technique in which inventory is counted on a frequent basis rather than once or twice a year

3/22/2009

PGP1-OM 2 -S Venkat

57

Alternative Classification Schemes

ABC Classification (on the basis of consumption value) XYZ Classification (on the basis of unit cost of the item)
High Unit cost (X Class item) Medium Unit cost (Y Class item) Low unit cost (Z Class item)

Selective Control of Inventories-

FSN Classification (on the basis of movement of inventory)

Fast Moving Slow Moving Non-moving

VED Classification (on the basis of criticality of items)

Vital Essential Desirable

On the basis of sources of supply

Imported Indigenous (National Suppliers) Indigenous (Local Suppliers)

3/22/2009

PGP1-OM 2 -S Venkat

58

29

ABC Classification System

Items kept in inventory are not of equal importance in terms of:

dollars invested profit potential sales or usage volume stock-out penalties

% of $ Value 30
0

60

A B C

% of Use

30 60

So, identify inventory items based on percentage of total dollar value, where A items are roughly top 15 %, B items as next 35 %, and the lower 65% are the C items
3/22/2009 PGP1-OM 2 -S Venkat 59

A graphical illustration
100% 90%
Class C

ABC Classification

80%

Consummption value (%)

Class B

70% 60%
Class A

50% 40% 30% 20% 10% 0%

0% 10 0% 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 %

No. of items (%)

3/22/2009 PGP1-OM 2 -S Venkat 60

30

Summary
Inventory mgt requires consideration of lot

of issues and trade-offs Enough care should be taken while taking assumptions Computational intensive Model selection is very critical Minor savings in inventory contributes significant savings in overall costs

3/22/2009

PGP1-OM 2 -S Venkat

61

31