Chapter 11, Part B Inventory Models: Probabilistic Demand

Lecture Outline Single-Period Inventory Model with Probabilistic Demand Single Order-Quantity, Reorder-Point Model with Probabilistic OrderReorderDemand Periodic-Review Model with Probabilistic Demand Periodic-

Slide 1

Slide 2

Probabilistic Models
In many cases demand (or some other factor) is not known with a high degree of certainty and a probabilistic inventory model should actually be used. These models tend to be more complex than deterministic models. The probabilistic models covered in this chapter are: single-period order quantity single reorder-point quantity reorder periodic-review order quantity periodic-

Slide 3

Slide 4

Harry Potter and the Order of the Phoenix by JK Rowling 5th

This book sold over a million copies in 1 day through Amazon alone. Ten million copies were sold in the first 3 days (English version alone). long queues of millions waiting for the opening of bookstores. If they cant buy it first week can they dont want to buy it . don The publishers were under pressure to produce the required amount of books

Slide 5

Newsboy problem: Single-Period Order Quantity SingleA single-period order quantity model (sometimes singlecalled the newsboy problem) deals with a situation in which only one order is placed for the item and the demand is probabilistic. probabilistic. D > Q If the period's demand exceeds the order quantity, the demand is not backordered and revenue (profit) will be lost. lost. D < Q If demand is less than the order quantity, the surplus stock is sold at the end of the period (usually for less than the original purchase price).

Slide 6

Single-Period Order Quantity

When To use Single-period inventory model with SingleProbabilistic Demand. Seasonal or perishable items. Cant be carried in the inventory. Can Cant be sold in the future Can Ex: News Papers Fashions Computer books Gadgets Harry Potter Books

Slide 7

Single-Period Order Quantity

One question only is in place How much of the products should we order There is no question such as when we should reorder.

Slide 8

Single-Period Order Quantity

Assumptions

Period demand follows a known probability

distribution (historical Data): normal: mean is , standard deviation is uniform: minimum is a, maximum is b Cost of overestimating demand: co Cost of underestimating demand: cu Shortages are not backordered. Period-end stock is sold for salvage (not held in Periodinventory).

Slide 9

Shoe Company
Mens shoe shop Men Summer season The shoe cost 40 40 If sold before 31 July price= 60 60 After 31 July (SALES) 30 30 D demand is between 350 to 650 pairs The probability distribution is uniform distribution

Slide 10

Cu=60-40=20. =60- 40=20 Loss because we lose the opportunity to win

Co =30-40=10. =30- 40=10 Loss because we sell in cheaper price

The big question is if we assume are selling 500 Is it worth to sell 501, what are our expected losses
Slide 11

What are our losses, in the following options

Increase our order quantity by 1 Q=501 Our losses will be in our inability to sell the additional unit ( we over estimated our demand) we sell it in cheaper price Co =10. =10 P(D<501)=P(D<500)=150/300 P(D<501)=P(D< EL(Q=501) = P* Co = .5 *10= Stay at 500 order Q=500 Our Losses will be in our inability to fulfil our demand losing the opportunity to win. Cu=20 =20 P(D>500)=150/300 EL( Q=500) = P* Cu = .5 *20=

10

Slide 12

What are our losses, in the following options

Increase our order quantity by 1 Q=502 Our losses will be in our inability to sell the additional unit ( we over estimated our demand) we sell it in cheaper price Co =10. =10 P(D<502)=P(D<501)=151/300 P(D<502)=P(D< EL(Q=501) = P* Co = .5033 *10= Stay at 501 order Q=501 Our Losses will be in our inability to fulfil our demand losing the opportunity to win. Cu=20 =20 P(D>501)=149/300 EL( Q=500) = P* Cu = .4966 *20=

5.033

9,933

Slide 13

What are our losses, in the following options

Increase our order quantity by 1 Q=601 Our losses will be in our inability to sell the additional unit ( we over estimated our demand) we sell it in cheaper price Co =10. =10 P(D<601)=P(D<600)=250/300 P(D<601)=P(D< EL(Q=501) = P* Co = .8333 *10= Stay at 500 order Q=600 Our Losses will be in our inability to fulfil our demand losing the opportunity to win. Cu=20 =20 P(D>600)=50/300 EL( Q=500) = P* Cu = .1667 *20=

8,33

3.33

Slide 14

Slide 15

Optimal solution would be when we stop increasing the level of our order and our losses form increasing will be more from status co. That will be when the expected losses would be the same EL (Q* +1)=EL (Q*) Co P(demand < Q *) = Cu P(demand > Q *) P(demand > Q *) + P(demand < Q *) =1 P(demand < Q *) = cu/(cu+co) /(c
Slide 16

Single-Period Order Quantity

Formulas Optimal probability of no shortage: P(demand < Q *) = cu/(cu+co) /(c P(demand < Q *) =20/(10+20) = 2/3 Q*= 550

Slide 17

Slide 18

Single-Period Order Quantity

Formulas Optimal probability of no shortage: P(demand < Q *) = cu/(cu+co) /(c

Optimal order quantity, based on demand distribution: normal: Q * = + z

uniform: Q * = a + P(demand < Q *)(b-a) *)(b

Slide 19

Example: McHardee Press

Single-Period Order Quantity Single McHardee Press publishes the Fast Food Menu Book and wishes to determine how many copies to print.

the incremental profit per copy is

$0.45.

Any unsold copies of the book can

be sold at salvage at a $.55 loss.

Slide 20

10

Example: McHardee Press

Single-Period Order Quantity SingleSales for this edition are estimated to be normally distributed. The most likely sales volume is 12,000 copies , with standard deviation 4848. How many copies should be printed?

Slide 21

Example: McHardee Press

Single-Period Order Quantity Single Using incremental analysis with Co = .55 and Cu = .45, P (demand < Q *) =(Cu/(Cu+Co)) = .45/(.45+.55) = =(C /(C .45 What is Q* for a P of 0.45 Q* = +z : = 12,000. z=? =4848 +z z: Find Q * such that P(D < Q *) = .45. The P(D probability of 0.45 corresponds to z = -.12.

Q * = +z= 12,000 - .12(4848) = +z

Therefore, (20,000 - 12,000) = 1.65 or = 4848

11,418 books

Slide 22

11

Example: McHardee Press

Single-Period Order Quantity (revised) SingleIf any unsold copies can be sold at salvage at a $.65 loss, how many copies should be printed? Co = .65, (Cu/(Cu + Co)) = .45/(.45 + .65) = .4091 (C /(C Find Q * such that P(D < Q *) = .4091. z = -.23 P(D gives this probability. Thus, Q * = 12,000 - .23(4848) = 10,885 books

Slide 23

If There is a fixed cost of $5,000 to produce the book

Since 10,885 books are less than the breakeven volume of 11,111 books (= 5000/.45), no copies should be printed because if the company produced only 10,885 copies it will not recoup its $5,000 fixed cost.

Slide 24

12

If we dont know the standard deviation

The most likely sales volume is 12,000 copies and they believe there is a 5% chance that sales will exceed 20,000. : note that z = 1.65 corresponds to a 5% tail probability. x = +z 20,000=12,000+1.65 +z =4848

Slide 25

Reorder Point Quantity Model

A firm's inventory position consists of the on-hand inventory onplus on-order inventory (all amounts previously ordered but onnot yet received). An inventory item is reordered when the item's inventory position reaches a predetermined value, referred to as the reorder point. point.

Slide 26

13

The reorder point represents the quantity available to meet demand during lead time. time. Lead time is the time span starting when the replenishment order is placed and ending when the order arrives.

Slide 27

Reorder Point Quantity

Under deterministic conditions, when both demand and lead time are constant, the reorder point associated with EOQ-based models is set equal to EOQlead time demand. Under probabilistic conditions, when demand and/or lead time varies, the reorder point often includes safety stock. Safety stock is the amount by which the reorder point exceeds the expected (average) lead time demand.

Slide 28

14

Slide 29

Safety Stock and Service Level

The amount of safety stock in a reorder point determines the chance of a stockout during lead time. The complement of this chance is called the service level. Service level, in this context, is defined as the level, probability of not incurring a stockout during any one lead time. Service level, in this context, also is the long-run longproportion of lead times in which no stockouts occur.

Slide 30

15

Reorder Point Model

Assumptions Lead-time demand is normally distributed Leadwith mean and standard deviation . Approximate optimal order quantity: EOQ Service level is defined in terms of the probability of no stockouts during lead time and is reflected in z. Shortages are not backordered. Inventory position is reviewed continuously.

Slide 31

Example: Roberts Drug

Reorder Point Model Robert's Drugs is a drug wholesaler supplying 55 independent drug stores. Roberts wishes to determine an optimal inventory policy for Comfort brand headache remedy.

Sales of Comfort are relatively constant as the past 10 weeks of data (on next slide) indicate.

Slide 32

16

Example: Roberts Drug

Reorder Point Model Average=120, Sd=7.45 Sd=7.45 Week 1 2 3 4 5 Sales (cases) 110 115 125 120 125 Week 6 7 8 9 10 Sales (cases) 120 130 115 110 130

Slide 33

Example: Roberts Drug

Each case of Comfort costs Roberts $10 Roberts uses a 14% annual holding cost rate for its inventory. The cost to prepare a purchase order for Comfort is $12. What is Roberts optimal order quantity? Roberts

Slide 34

17

Example: Roberts Drug

Optimal Order Quantity The average weekly sales over the 10 week period is 120 cases. Hence D = 120 X 52 = 6,240 cases per year; Ch = (.14)(10) = 1.40; Co = 12.

Q * = 2 DC o /C h = (2(6240)(12))/1.40 = 327

Slide 35

Example: Roberts Drug

The lead time for a delivery of Comfort has averaged four working days. (week = 6 working days) Lead time demand has therefore been estimated as having a normal distribution with a mean of 80 cases and a standard deviation of 10 cases. (see later) Roberts wants at most a 2% probability of selling out of Comfort during this lead time. What reorder point should Roberts use?

Slide 36

18

Example: Roberts Drug

Optimal Reorder Point Hence Roberts should reorder Comfort when supply reaches r = + z Lead time demand is normally distributed with = m = 80, = 10 cases Z: Since Roberts wants at most a 2% probability of selling out of Comfort, the corresponding z value is 2.06. That is, Comfort, P (z > 2.06) = .0197 (about .02). Hence Roberts should reorder Comfort when supply reaches r = + z = 80 + 2.06(10) = 101 cases. The safety stock is z = 21 cases.

Slide 37

Slide 38

19

Reorder Point

Formulas Reorder point: Safety stock: Average inventory: Total annual cost: r = + z z ( Q ) + z [( )Q *Ch] + [z Ch] + [DCo/Q *] (hold.(normal) + hold.(safety) + ordering)

Slide 39

Example: Roberts Drug

Total Annual Inventory Cost Ordering: (DCo/Q *) = ((6240)(12)/327) = $229 (DC Holding-Normal: (1/2)Q *Co = (1/2)(327)(1.40) = 229 Holding(1/2)Q Holding-Safety Stock: Ch(21) = (1.40)(21) = 29 HoldingTOTAL = $487

Slide 40

20

Periodic Review Model

A periodic review system is one in which the inventory level is checked and reordering is done only at specified points in time (at fixed intervals usually). Assuming the demand rate varies, the order quantity varies, will vary from one review period to another. At the time the order quantity is being decided, the concern is that the on-hand inventory and the onquantity being ordered is enough to satisfy demand from the time the order is placed until the next order is received (not placed).

Slide 41

Slide 42

21

Periodic Review Order Quantity

Assumptions Inventory position is reviewed at constant intervals. Demand during review period plus lead time period is normally distributed with mean and standard deviation . Service level is defined in terms of the probability of no stockouts during a review period plus lead time period and is reflected in z. On-hand inventory at ordering time: H On Shortages are not backordered. Lead time is less than the review period length. length.

Slide 43

Periodic Review Order Quantity

Formulas Replenishment level: Order quantity: M = + z Q=MH

H: inventory on hand at the review period

Slide 44

22

Example: Ace Brush

Periodic Review Order Quantity Model Joe Walsh is a salesman for the Ace Brush Company. Every three weeks he contacts Dollar Department Store so that they may place an order to replenish their stock.

Weekly demand for Ace brushes at Dollar

approximately follows a normal distribution with a mean of 60 brushes and a standard deviation of 9 brushes per week.

Slide 45

Example: Ace Brush

Periodic Review Order Quantity Model Once Joe submits an order, the lead time until Dollar receives the brushes is one week. week. Dollar would like at most a 2% chance of running out of stock during any replenishment period. If Dollar has 75 brushes in stock when Joe contacts them, how many should they order?

Slide 46

23

Satisfying 3+1=4 weeks of demand

Slide 47

Example: Ace Brush

Demand During Uncertainty Period The review period plus the following lead time totals 4 weeks. This is the amount of time that will elapse before the next shipment shipment of brushes will arrive. Weekly demand is normally distributed with: Mean weekly demand, = 60 = 9 Weekly standard deviation, Weekly variance, 2 = 81 Demand for 4 weeks is normally distributed with: Mean demand over 4 weeks, = 4 x 60 = 240 2 =n 2 Variance of demand over 4 weeks, 2 = 4 x 81 = 324 Standard deviation over 4 weeks, = (324)1/2 = 18

Slide 48

24

Example: Ace Brush

Replenishment Level M = + z z: where z is determined by the desired stockout probability. For a 2% stockout probability (2% tail area), z = 2.05. , determined in the previous slide M = 240 + 2.05(18) = 277 brushes As the store currently has 75 brushes in stock, Dollar should order: 277 - 75 = 202 brushes The safety stock is: z = (2.05)(18) = 37 brushes

Slide 49

25