INVENTORY PLANNING AND CONTROL

Inventory Management: Inventory holding costs money – it includes cost of capital (e.g. interest charge etc) and
cost of storage, apart from those associated with obsolete/surplus/non-moving items. No company has access to unlimited
working capital and hence the quantity procured becomes a critical parameter in the materials management area.

Inventory is often defined as an idle resource of any kind, having an economic value in the sense, the raw material can be
converted into semi-finished goods and with additional value, becomes finished goods. In all these cases, the company’s
working capital is tied up and hence the finance manager is wary of servicing the idle working capital at, about say, 30%
per annum. On the contrary, if the item is not kept in the stores, there will be a stockout, if the demand arises. Thus larger
quantum of inventories do not necessarily lead to higher volume of output, while lack of inventories will hamper
production.

It is noted that inventories cost money to acquire as well as hold them. The cost of acquisition – including
communication, development etc varies considerably. This depends upon quantum purchased, imports, advertisement,
follow-up, travel, communication and other elements. The cost of acquisition is increased, as smaller quantities are
procured each time, but the decrease in inventories also decreases the inventory carrying charges. Thus the materials
manager has to balance the two opposing costs, in order to strike the optimum level of inventories, which will maximize
the total cost in an organisation.

Thus the basic objective of any inventory management is to determine the right quantity to be procured, in order to release
the capital for more productive use. The materials management department is accused of both stockouts as well as large
investments in inventories. The solution lies in exercising a selective inventory control and application of inventory
management principles. The optimum quantity should be sufficient to achieve maximum production, but it should not be
so excessive with the locking of working capital, as to restrict the ability of an organisation to earn a high rate of return.

Let us consider three items - U, V and W- accounting for an annual consumption of Rs, 60,000, Rs. 4,000 and Rs. 1,000
respectively. Table below shows two situations – the left half showing the case when uniform inventory control is applied
to all the items and the right half showing the case when selective inventory control is applied. The basis of the selective
inventory control was arrived at on the principles of ABC analysis (i.e. on the annual usage/value).

Uniform inventory control Selective inventory control

Average Average
No. No.
Annual .Value Inventory Annual Value Inventory
of of
Item Consumpti per (Rs.) – Item Consumpt per (Rs.) –
orders orders
No. on Value order half of No. ion Value order half of
per per
(Rs.) (Rs.) value per (Rs.) (Rs.) value per
year year
order order
U 60,000 4 15,000 7,500 U 60,000 8 7,500 3,750
V 4,000 4 1,000 500 V 4,000 3 1,333 667
W 1,000 4 250 125 W 1,000 1 1,000 500
Total Working Capital Commitment Total Working Capital Commitment
8,125 4,917
(Rs.) (Rs.)

The table shows that by selective control applications, the working capital commitment is reduced by about 50% for these
three items, with the total number of orders remaining at 12.

Economic order Quantity (EOQ): The economic ordering quantity helps the purchase executive to arrive at a solution
satisfying the four conflicting parameters – (a) not too much, (b) not too little, (c) at minimum total cost, and (d) for
increased profitability. Let us consider an item with an annual usage of 1,000 Nos. and price per unit as Rs.1.00. The
total carrying cost including the cost of capital, and storage charges is 30%. The acquisition cost is Rs. 600.00 per order.

Let us arrive at the EOQ by trial and error method as shown in the following table. The ordering cost per year is obtained
by multiplying the number of purchase orders per year by ordering cost. Other steps are explained in the table. The annual

1
acquisition cost of materials, being same for all order quantities, has not been taken into consideration for this tabular
calculation.

Number of
Order Average
purchase Inventory Carrying Total Cost per
quantity with Ordering Inventory Value of
orders per year Cost per year year
every Cost per [half of average
[Annual [30 % of value of [Ordering Cost +
purchase year order inventory
Consumption / average inventory] Carrying Cost]
order (Rs.) quantity] (Rs.)
Order (Rs.) (Rs.)
(Nos.) (Nos.)
Quantity]
600/- x 5 200 / 2 = 100 x 1/- 3,000/- + 30/- =
200 1,000/200 = 5 100/- x 0.3 = 30
= 3,000 100 = 100 3,030
500 2 1,200 250 250 75 1275
1,000 1 600 500 500 150 750
2,000 0.5 300 1,000 1,000 300 600
5,000 0.2 120 2,500 2,500 750 870

As clear from the table, the optimum order quantity that minimizes the total of ordering cost and carrying charges is
2,000.

Mathematical Derivation: Assumptions:

• Demand for the product is constant and uniform through the period.
• Lead time is constant.
• Price per unit of product is constant.
• Inventory holding cost is based on average inventory.
• Ordering or setup costs are constant.
• No backorders are allowed i.e. all demands for the product will be satisfied.
Inventory on Hand

FIXED ORDER QUANTITY MODEL

R
Lead Time, L R

Time

Nomenclatures:

TC = Total Annual Cost in Rs. S = Setup Cost or Cost of Placing an order, Rs./Order
D = Annual Demand in Nos. R = Reorder Point, Nos.
C = Cost per unit, Rs./No. L = Lead Time, in time unit e.g. days or weeks etc.
Q = Quantity to be ordered in Nos.

2
H = Annual Holding and Storage Cost per unit of average inventory (often holding cost is taken as a percentage of the
cost of the item, such as H = i C, where i is the percentage carrying cost.). (also called annual carrying cost).
Relevant Costs

FIXED ORDER QUANTITY MODEL

Annual Product costs, based on size of the order.
Annual acquisition
cost of items,
Annual Total
D x Cost,
C TC

Annual
Ordering Cost,
(D/Q) x S
Annual Holding Cost, H

Qopt
..

Order Quantity Size, Q

Qopt = Q* = Order Quantity at which total cost is a minimum.

Now, Annual total cost = Annual acquisition cost + Annual ordering cost + Annual holding or carrying cost

TC = D x C + (D/Q) x S + (Q/2) x H

For minimization of the total annual cost, we first differentiate TC with respect to Q and then set it to zero. Thus we first
find the value of Q*. Then we again differentiate the first order differentiation w.r.t Q and put the Q* value in the second
order differentiation to check whether it gives the minimum value of TC.

Differentiating w.r.t. Q, and setting,d (TC) / dQ = 0, we get, 0 = 0 + [- (D x S) / Q2] + H/2

Thus, EOQ = Qopt = [(2D x S) / H]1/2

Performing the second order differentiation w.r.t Q, we get, d2 (TC) / d2 Q = 2 x (D x S) / Q3 + 0
At the value of Q = Q*, d2 (TC) / d2 Q = [+ ve]. This shows that at Q*, we the minimum value of TC.

3
We have assumed constant demand and lead time, hence the reorder level, R = [d1 x L], where, d1 = average daily
demand and L = Lead time in days.

Crossing of the solution of the previous problem by the mathematical formula for EOQ,

Given, D = 1,000 Nos.; S = 600/-; H = 0.3 x 1/- = 0.30/-;

Hence EOQ = Q* = [2 x 1,000 x 600 / 0.30] 1 /2 = 2,000 nos. (Checked)

4
Summary:-
Purpose of Inventory:

1. To maintain independence of operation.

2. To meet variation in product demand.
3. To allow flexibility in production scheduling.
4. To provide a safeguard for variation in raw material delivery time.
5. To make advantage of economic purchase order size.

Inventory Costs

1. Holding or Carrying costs.

2. Setup or Production change costs.
3. Ordering costs.
4. Shortage costs.

Common Inventory Systems:

1. Fixed-order Quantity models (or EOQ, or Q-models). These models are “event -triggered’. A
fixed-order quantity model initiates an order when the event of reaching a specified reorder
level occurs. This event can take place any time, depending on the demand for the item
considered. Thus a fixed-order quantity model is a perpetual system, which requires that
every time a withdrawal from inventory or an addition to inventory is made, records must be
updated to reflect whether the reorder point has been reached.

2. Fixed –time period models (or Periodic review systems, or P-models). These models are
“time – triggered’. The fixed-time period model is limited to placing orders at the end of a
predetermined time period; only the passage of time triggers the model. Here the counting
takes place only at the review period.

3. Hybrid systems that combine features of both EOQ i.e. Q-models and P-models.

5
 INVENTORY MODELS
Case1. Fixed order Quantity Model – Deterministic Uniform Demand with constant Lead-
time and no shortage: -

Inventory Level

Batch Size Q
(Q – a x t)

Reorder Level R

0
Time, t
Q/a Q / 2a Q / 3a

L L

Total Annual Cost

Annual
TC
Cost in Rs.

Annual Holding
Cost,( Q / 2 ) x H

Annual Cost of Items

(D x C)
Annual Ordering Cost, (D / Q) x S

Q* Order Quantity Size, Q in units

6
Where, TC = Total annual Cost; D = Annual Demand; C = Cost per Unit; Q= Quantity to be Ordered;
S = Setup Cost or Cost of Placing an Order; R = Reorder Level; L = Lead time; H = Annual Holding
and Storage cost per unit of average inventory, usually taken as a percentage of the cost of the item.

Total Annual Cost = Annual purchase Cost + Annual Ordering Cost + Annual Holding Cost.

TC = D x C + ( D / Q) x S + ( Q / 2 ) x H

Differentiating w.r.t. Q, dTC/dQ = 0 + ( - D x S / Q2 ) + H / 2 = 0

Economic Ordered Quantity = EOQ = Q* = [2 x D x S / H] ½

and, the Reorder Level, R = d- x L = Average Daily Demand X Lead Time in Days.

Say, Cost item = Rs. 10/- per unit; a = Demand Rate = 8,000 units per month; Ordering cost = Rs.
12,000/- per order; H = Holding Cost Rs. 0.30/- per unit per Month.

EOQ = [ 2x 8,000 x 12,000 / 0.30]1/2 = 25,298 units.

Reorder Time = t* = (Q* / a) = [ 25,296 / 8,000] = 3.20 months.

Example to explain the effect of quantity discount: A publishing house purchases 2,000 units of a
particular item per year at a unit cost of Rs. 20. The ordering cost per order is Rs. 50 and the
inventory carrying cost is 25 % of the average inventory. Find the optimal order quantity. If 3 %
discount is offered by the supplier for the purchase in lots of 1,000 or more, should the publishing
house accept the offer?

Ans: (a) Economic / Optimal Order Quantity = EOQ = [(2 x S x Co) / ( i x C1) ] 1/2
Where, S = Annual consumption / usage = 2,000 ; Co = Ordering cost per order = Rs. 50 ; i =
Carrying cost percentage = 0.25 ; C1 = Unit cost of the item = Rs. 20.

Thus, EOQ = [(2 x 2,000 x 50) / (0.25 x 20)] 1/2 = 200 units.

(b) When 3 % discount is offered by the supplier for the purchase in lots of 1,000 or more:

TC1 = Total Annual Cost (when order size is 1,000 units) = [Material Cost + Ordering Cost +
Carrying Cost]
= 2,000 x 20 x (1 - 0.03) + (2,000/1,000) x Rs. 50 + ½ x 1,000 x 0.25 x Rs. 20 x (1 - 0.03)
= 38,800 + 100 + 2,425. = Rs. 41,325.

TC2 = Total Annual Cost (when order size is 200 units) = [Material Cost + Ordering Cost + Carrying
Cost]
= 2,000 x 20 + (2,000/200) x Rs. 50 + ½ x 200 x 0.25 x Rs. 20
= 40,000 + 500 + 500 = Rs. 41,000.

Since the total cost under “order size = 200” is lower, the publishing house should not accept the
offer.

7
 Case 2. Fixed Order Quantity Model – Deterministic and Uniform Demand with Constant
Lead-time and Planned Shortage: -

Inventory Level
(W - a x t)
Batch Size, Q

W
Sizde

Time t
(W / a)

Maximum Shortage (Q - W)
(Q / a)
Cycle Time

Let,
c = production or procurement cost per unit, say Rs 10/- per unit.
p = shortage cost per unit per unit time short, say Rs 1.10/- per unit per month
h = holding cost per unit per unit time, say Rs. 0.30/- per unit per month
K = set up or ordering cost, say Rs. 12,000/- per set-up
a = demand rate, say 8,000 Nos. per month
W= Inventory level just after a batch of Q nos. is added.
Q = Order Quantity or Production Batch size in nos.

We are interested to know the economic ordering quantity Q*, optimum level of inventory W* and
optimum cycle time period t*. There are two decision variables, W and Q.

Production or Ordering Cost per Cycle = Ordering Cost + Material procurement per cycle
= K + c Q ……………………………………………..(i)

Now, average inventory level during each cycle = [(W + 0)/ 2] = W / 2, units.

Also, during each cycle, the inventory is positive for a time period = [W / a].

Therefore,
Holding Cost per Cycle = Holding cost per unit time x Cycle time of inventory holding.
= [h x W /2] x [ w / a] = [h W2 / 2 x a ], Rs. per cycle……………….(ii)

Similarly, during each cycle shortage occurs for a time period = [(Q – W) / a], hence
Average amount of shortage during this time of [(Q - W) / a] = [{0 + (Q - W)} / 2] = [(Q - W)/2]
Nos.
Corresponding shortage cost per unit time = p x [(Q - W) / 2], Rs per unit time. Therefore,

Shortage cost per cycle = Average amount of shortage during the cycle x Shortage time in a cycle
8
 = [ p x {(Q-W}/2] x [(Q-W) / 2] = [ p x (Q – W)2 / 2 x a], Rs per cycle…..(iii)

Thus, total cost per cycle = T.C. = [Production or Ordering Cost + Holding Cost + Shortage Cost]
= [K + c Q] + [h W2 / 2a] + [p (Q – W)2 / 2a]
Hence,

Total cost per unit time = T = {T.C. / [Q / a]}

Hence, T = [{a x K /Q} + a x c ] + [ h W2 / 2 Q ] + [ { p (Q- S)2 / 2 Q ]…………………………(iv)

There are two decision variables, W and Q. For optimal values of W* and Q*, the partial derivatives
of the total cost per unit time T with respect to W and Q should be set equal to zero.

Thus, [(h x w) / Q ] - [ p (Q – W) / Q ] = 0 and ………………………………………………. …(v).

[- (a x K) / Q2 - (h x W2 / 2 Q2) + [ p (Q - W) / Q ] – [ p x (Q -W )2 / 2 Q2 ] = 0……..…..…(vi)

Solving, (v) and (vi), we get,

W* = [2aK / h ]1/2 x [p / (p + h)]1/2 = [2 x 8,000 x 12,000 / 0.30]1/2 x [1.1 /(1.1+0.3)]1/2

= 22, 424 Nos.

Q* = [2aK / h ]1/2x [( p + h ) / p]1/2 = [2 x 8,000 x 12,000 / 0.30]1/2 x [ (1.1 + 0.3) / 1.1] 1/2
= 28, 540 Nos.

Optimal Cycle length, t* =[Q* / a] = 28, 540 / 8,000 = 3.6 months.

Maximum Shortage = (Q* - W*) = 6,116 Nos.

Fraction of time that no shortage exists = [(W*/a) / (Q*/a)] = [W* / Q*] = [p / (p + h)]
= [1.1/ (1.1+0.3)] = 0.7857 (this is independent of K)

Remark: The normal formula for EOQ has been multiplied by a factor of [(p + h) / p]1/2.

9
Case 3. Fixed Order Quantity Model with deterministic and uniform demand rate R with no
lead time or shortage, where stock replenishment is made uniformly (rather than
instantaneously as are in earlier cases) at a rate k until the order quantity Q is
fulfilled: -

Inventory
Production & Consumption only @ R
Consumption @ (k - R)
N
(Q – R t1)

O Time
B
Prod.
time t1

Cycle time t

Run sizes are assumed constant and equal to Q. A new production run will be started whenever the
inventory becomes zero.
Let,

R = Demand or consumption rate = 10,000 Nos. bearings per day;

k = Production Rate per day = 25,000 Nos. bearings per day. (Obviously k must be more than R).
Q = Number of pieces produced per run; we have to find the optimum value of run size i.e. Q*.
t = time interval between production runs = Inventory cycle time = (Q / R)., Clearly, t* = (Q*/R)
t1 = production run time = (Q / k). (Obviously t will be more than t1).
C1 = Holding cost in Rupees per number per day. = Rs. 0.02/- per bearing per year
= (0.02 / 365) = Rs. 0.000055/- per bearing per day.
C2 = Set up or ordering cost in Rupees per production Run or Set up. = Rs 18.00/- per set up.

Inventory per cycle = Area of the triangle ONB = [½ x t x (Q – R x t1)]

= [½ x t x {Q – R x (Q / k)}] = [½ x t x Q x (1 – R / k)].

Total Cost per cycle = T.C. = Holding Cost per cycle + Set up Cost
= Inventory per cycle x C1 + C2
= ½ x C1 x t x Q x (1 – R/k) + C2

Therefore, Total cost per day = T = (T.C. / t) = [1/2 x C1 x Q x (1 – R /k) + (C2 / t) ]

10
Putting, (1/ t ) = ( R /Q), we get,

Total cost per day = T = [½ x C1 x Q x (1 - R/k) + (C2 x R /Q)]

For minimum cost per day, (dT / dQ) = 0 i.e. ½ x C1 x (1- R/ k) - C2 x R / Q2 = 0

Q* = [2 x C2 x R / {C1 x (1 – R/k)}] 1/2

And t* = [Q* / R] = [2 x C2 / {R x C1 x (1-R /k)}]1/2

Q* = [{2 x 18 x 10,000} / {0.000055 x (1 –10 / 25)}] ½ = 105,000 pieces

And t* = [Q* / R] = [105,000 / 10,000] = 10.5 days.

Remark: Formula for normal EOQ for instantaneous delivery, QI* = [2 x C2 x R / C1]1/2
Formula for EOQ for gradual delivery, QG* = [2 x C2 x R / C1]1/2 x [ k / (k – R)]1/2

That means the normal formula has been multiplied by a factor of [k / (k – R)]1/2

EOQ with Quantity Discounts

Suppliers may offer their goods at lower unit prices if larger quantities are ordered. This practice is
referred to as quantity discounting and occurs because larger order quantities may be less expensive
to produce and ship.

The quantity purchased does not necessarily have to be the usual EOQ amount as formulated from
Case1; rather, it is the quantity that minimizes the sum of annual carrying, ordering and acquisition
costs.

11
 Case 4. Fixed Order Quantity Model with Quantity Discount (assumed Deterministic Demand
and constant Lead Time):

Total Cost
TCA
per Unit
Time, TC

TCB

99,100/-
TCC

89,200/-

87,859/-

0 10,000 25,298 80,000 Batch Size, Q

Demand Rate = D = 8,000 unit per month; Set Up or Ordering cost = S = Rs. 12,000/- per Batch;
Holding Cost = H = Rs. 0.30/- per unit per month.

The discount set is stated as:

“For less than 10,000 units, price is Rs. 11/- per unit; for quantifies from 10,000 to 80,000
units, price is Rs. 10/- per unit; for above 80,000 units, price is Rs. 9.50/-per unit.”

EOQ = Q* = [2 x 8,000 x 12,000 / 0.30] 1/2 = 25,298 units.

TCA = 8,000 x 12,000 / 10,000 + 8,000 x 11.00 + 0.30 x 10,000 / 2 = Rs. 99, 100/-
TCB = 8,000 x 12,000 / 25,298 + 8,000 x 10.00 + 0.30 x 5,298 / 2 = Rs. 87,589/- (minimum)
TCC = 8,000 x 12,000 / 80,000 + 8,000 x 9.50 + 0.30 x 80,000 / 2 = Rs. 89,200/-

Considering the minimum cost, it is best to order a quantity of 25,298 units for this set of quantity
discounts.

But if the discount set is revised as: “for above 80,000 units, price is Rs. 9.00/- per unit”, then TCC
will change to 80,000 x 12,000 + 8,000 x 9.00 + 0.30 x 80,000 / 2 = Rs. 85,200/-.

In this case, the optimum quantity would be 80,000 units (instead of 25,298 units in the initial set of
discount), as now the TCC will be the minimum.

12
Example: A supplier of SS valve has offered Mr. Sen quantity discounts if he will purchase more
than his present order quantities. The new volumes and prices are: -

Range of Order Quantities Acquisition Cost (ac) in Rs. per Valve

1 - 399 2.20
400 - 699 2.00
700 and above 1.80

Mr. Sen has asked you to investigate the new prices under two sets of assumptions: -

A. Orders are received all at once i.e. instantaneous delivery under the estimates: -

Annual Demand Rate = 10,000 valves per year; Ordering Cost = C2 = Rs. 5.50/- per order;
Cost of carrying one unit of inventory per year = C1 = 20% of acquisition cost in rupees per valve per
year = 0.2 x Rs. 2.20/- = Rs. 0.44/- per valve per year.

B. Deliveries are gradual under the estimates; -

Annual Demand rate = D = 10,000 valves per year; Ordering Cost = C2 = Rs. 5.50/- per order;
Cost of carrying one unit of inventory = C1 = 20 % of acquisition cost per valve per year
(Clearly this varies according to the range of order quantities);
Replenishment Rate = Delivery Rate = k = 120 valves per day;
Consumption rate = R = [10,000 valves per year / 250 working days] = 40 valves per day.

Ans.:

A. Orders Received all at once i.e. Instantaneous Delivery: -

The EOQs are computed for each of the acquisition costs: -

EOQ2.20 = [(2 x Annual Demand Rate x C2) / C1]1/2

= [{2 x 10,000 x 5.50} / { 0.2. x 2.20}]1/2 = 500 Nos.

EOQ2.00 = [{2 x 10,000 x 5.50} / {0.2 x 2.00}]1/2 = 524.4 Nos.

EOQ1.80 = [{2 x 10,000 x 5.50} / {0.2 x 1.80}]1/2 = 552.8 Nos.

We graph the TMC (Total Material Cost per year) for each acquisition cost. For instance, TMC2.20
can be graphed by substituting several values for Q in this TMC formula:

TMC = (Q /2) x C1 + (D / Q) x C2 + D x Acquisition Cost per unit

TMC2.20 = (Q / 2) x (2.20 x 0.20) + (10,000 / Q) x 5.50 + 10,000 x 2.20.

We note that only EOQ2.00 is feasible because 524.4 valves per order can be purchased at Rs. 2.00/-
per valve. The TMC at two quantities is therefore investigated: 524.4 units per order each at Rs.
2.00/- and 700 units per order each at Rs. 1.80/-: -

13
When Q = 524.4, TMC = (Q/2) x C1 + (D /Q) x C2 + D x ac
= [524.4 / 2] x (0.2 x 2.00) + (10,000 / 524.4) x 5.50 + 10,000 x 2.00
= 104.88 + 104.88 + 20,000
= Rs. 20,209.76/- per year.

When Q = 700, TMC = [700 /2] x (0.2 x 1.80) + (10,000 / 700) x 5.50 + 10,000 x 1.80
= 126.00 + 78.57 + 18,000
= Rs. 18,204.57/- per year.

We conclude that if orders are delivered all at once, 700 valves should be ordered at each inventory
replenishment.

B. Gradual Delivery:-

The EOQs are computed as: -

EOQ2.20 = [2 x D x C2 / C1] 1/2 x [k / (k- R)]1/2

= [2 x 10,000 x 5.50 / {0.2 x 2.20}] ½ x [120 / ( 120 – 40) ] ½ = 612.4 Nos.

EOQ2.00 = [2 x 10,000 x 5.50 / (0.2 x 2.00) ]1/2 x [120 / (120 – 40)]1/2 = 642.3 Nos.

EOQ1.80 = [2 x 10,000 x 5.50 / (0.2 x 1.80)]1/2 x [ 120 / (120 – 40)]1/2 = 677.0 Nos.

We note that only EOQ2.00 is feasible because 642.3 valves per order can be purchased at Rs. 2.00/-
per valve. Two quantities are investigated, Q = 642.3 and Q = 700 units per order: -

When Q = 642.3,
TMC = (Q/2) x [{k – R} / k] x C1 + (D / Q) x C2 + D x ac
= (642.3 / 2) x [{120 - 40} / 120] x (0.20 x 2.00) + (10,000 / 642.3) x 5.50 + 10,000 x 2.00
= 85.63 + 85.63 + 20,000 = Rs. 20,171.26 /- per year.

When Q = 700,
TMC = (700 / 2) x [{120 - 40} / 120] x (0.20 x 1.80) + (10,000 / 700) x 5.50 + 10,000 x 1.80
= 84.00 + 78.57 + 18,000 = Rs. 18,162.57 /- per year.

We conclude that if gradual deliveries are used, 700 units per order should be purchased.

Conclusion: Given a choice, Mr. Sen would prefer to have gradual deliveries of the valves in
quantities of 700 units per order because the TMC of gradual deliveries is slightly less than that for
orders delivered all at once.

14
 Case 5A. Fixed Order Quantity Model with Safety Stock and Probabilistic Demand (normally
distributed) and Constant Lead Time:

5% of area Q Model
Inventory
Level
Z=1.64
Q*
Order Placement
Q*

Reorder Level, R

Range of
Safety Stock, B
Q* Demand

Time
L
Occurrence of
Shortage

The key difference between Constant Demand Model and the Probabilistic Demand
Model is in the computing of the Reorder Level R. The Economic Order Quantity i.e. Q* is the
same in both cases, i.e. EOQ = [ 2 x 60 x { 365 x 10 } / 0.50]1/2 = 936 units.

Say, average daily demand = d- = 60 unit per day; H = Rs.0.50/- per unit per year.; L = Lead Time in
days = 6 days; Ordering Cost = S = Rs. 10/- per Order or Set up.; Specified Service Probability =
Probability of non stock out = 95 % i.e. corresponding Z = 1.64.; Standard Deviation [S.D.] of Daily
Demand = 7 units per day.

Standard Deviation of Demand during the Lead Time = [ L1/2 x S.D. of Daily Demand]
= 6 ½ x 7 = 17.15 units.

Safety Stock Level = Z x S.D. during Lead Time = 1.64 x 17.15 = 28 units.
Reorder Level = Average daily demand x Lead Time + Safety Stock Level
= 60 x 6 + 28 = 388 units.

So, an order of 936 units is to be placed whenever the number of units remaining drops to 388 units.

15
Case 5B. Fixed Order Quantity Model with Safety Stock and Probabilistic Demand (normally
distributed) and Probabilistic Lead Time (normally distributed):

A mathematical method would be to find the probability distribution for the consumption rate as well
as the lead time from the past data. Then a joint probability distribution of the usage during “a lead
time” may be found. The distribution can then be converted into a cumulative frequency curve. Then
corresponding to a risk level, the maximum demand rate can be read off from the cumulative
frequency curve. The buffer stock is given by the difference between the maximum demand and the
average demand. [For an example see P-21.23 of the book “Production and Operations Management”
by Mr. S.N. Chary – Calculation of Joint Probabilities Method]

The above calculation (i.e. Calculation of Joint Probabilities Method) is somewhat cumbersome. A
formula for use in practice can be as follows: -

The standard deviation of total demand during the lead time σ D is given by,
2
σ D = t- σ 2x + x- 2 σ 2t

where, x- = the average demand per month/week

t- = the average lead time in months/week
σ x = the standard deviation of demand in month/week
σ t = the standard deviation of lead time in months/week

Ex: The average demand rate for a particular raw material for a company is estimated to be 1,000
units per month. The distribution of demand rates is found to approximate a normal distribution with
standard deviation of 200 units per month. The average lead time for the procurement of the raw
material is observed in the past to be approximately 3 months and the normally approximated lead
times have standard deviation of 1 month. If the risk level tolerated by the management is only 5 %,
what is the buffer stock required for this raw material?

Ans: x- = 1,000 units/month.

t- = 3 months.
σ x = 200 units/month.
σ t = 1 month.

Applying the formula, the standard deviation of total demand during a lead time,
σ D = [3 x (200)2 + 1,000)2 x (1)2] = 1,060 units.

Therefore, corresponding to 5 % risk level, we have under normal distribution, 1.645 times standard
deviation from the mean.

Hence the buffer stock = B.S. = Z x σ D = 1.645 x 1,060 = 1,744 units.

16
 Case 6. Fixed Time Period Model with Probabilistic Demand (demand normally distributed)
and constant Lead time:

Place order after review

Inventory P Model
Level
Q2

Q1
Q3
Safety Stock

Time, t
Stock Out
L

T T
L

In P Model, inventory is counted (i.e. reviewed) only at particular times e.g. after every T time period
and then the orders are placed accordingly.

Obviously, order quantities vary from period to period, depending on the usage rate and inventory on
hand. As inventories are counted only after time period T, it is possible that some large demand will
draw the entire stock down to zero immediately after an order is placed. This condition may go
unnoticed until the next review period; then the new order, when placed, still takes time to arrive.

Thus it is possible to be out of stock throughout the entire review period T and the order lead time L.
Safety Stock, therefore, must protect against stock outs during the review period itself as well as
during the lead time i.e. for a total time period of (T + L).

Say, Average Daily Demand = 10 units per day; Review time Period = T = 30 days; Lead Time = L
= 14 days.; Inventory at the beginning of the review period = I == 150 units; Standard Deviation of
daily Demand = 3 units per day.; Required Probability of non stock out = 98 % i.e. corresponding Z
= 2.05.

Standard Deviation during the vulnerable period of (T + L) = (T+L) ½ x S.D. of Daily demand = (30 +
14) ½ x 3 = 19.90 units.

Quantity to Order = d- ( T + L) + Z x S.D. during (T+L) – I

= 10 x (30+14) + 2.05 x 19.90 – 150 = 331 units.

So, we should place an order for 331 units after this review period.
17
Case 7. Single Period Inventory Model

Some single period problems involve determining an order quantity for an item to cover the demand
for a single period only. This type of problem is common for short–lived materials such as fashion
goods, perishable goods, and published materials such as magazines and newspapers. The structure of
theses problems (called mews-boy problems) is well suited for the use of payoff tables.

Ex. Fashion Retailers Inc. is trying to decide how many Silk Scarves to stock for sale for next winter
season. The sales history of this item is as follows: -

Silk Scarves Demanded (SN) in Nos. Probability of Scarves Demanded, p(SN)

100 0.10
200 0.10
300 0.40
400 0.30
450 0.10
Total = 1.00

The silk scarf sells for Rs. 15/- per unit and has a cost of goods sold of Rs. 10/- per unit. If one of
these scarves is stocked for sale but is not sold during the season, it costs Rs. 2/- to discount it next
season.

a. Use payoff tables to minimize the total expected costs. What is the expected value of perfect
information (EVPI)?
b. Use payoff tables to maximize the total expected profits. Compute the EVPI.
c. What stocking strategy is best for the scarf?

Ans.

a. Minimize Total expected Costs

Payoff Cost Matrix

States of Nature Total Expected Costs
EC = Sum of p(SN) x C ij
100 200 300 400 450
1000.0 1500.0 1750.0
100 0.00 500.00 1075.00
0 0 0
1000.0 1250.0
Strategies 200 200 0.00 500.00 645.00
0 0
300 400 200.00 0.00 500.00 750.00 285.00
400 600 400.00 200.00 0.00 250.00 205.00
450 700 500.00 300.00 100.00 0.00 270.00
Probabilities, p(SN) 0.10 0.10 0.40 0.30 0.10

For instance, the EC for the demand of 400 units is,

18
 EC = 0.10 x (600.00) + 0.10 x 400.00 + 0.40 x 200.00+ 0.30 x 0.00 + 0.10 x 250.00 = 205.00

The EVPI is Rs. 205/-, the value of the minimum total expected costs derived from the payoff table
above. In other words, as much as Rs. 205/- per season could be spent for perfect market information
to remove the uncertainly.

b. Maximize total expected profits

Payoff Profit Matrix

States of Nature Total Expected Profits
EP = Sum of p(SN) x P ij
100 200 300 400 450
10
500.00 500.00 500.00 500.00 500.00 500.00
0
20 1000.0 1000.0 1000.0 1000.0
300.00 930.00
0 0 0 0 0
Strategies 30 1500.0 1500.0 1500.0
100.00 800.00 1290.00
0 0 0 0
40 1300.0 2000.0 2000.0
-100.00 600.00 1370.00
0 0 0 0
45 1200.0 1900.0 2250.0
-200.00 500.00 1305.00
0 0 0 0
Probabilities, p(SN) 0.10 0.10 0.40 0.30 0.10

For instance, the EP for the demand of 400 units is,

EP = 0.10 x (-100.00) + 0.10 x 600.00 + 0.40 x 1300.00 + 0.30 x 2000.00 + 0.10 x 2000 .00
= 1370.00

In this case,

EVPI
= [0.10 x 500.00 + 0.10 x 1000.00 + 0.40 x 1500.00 + 0.30 x 2000.00 + 0.10 x 2250.00] – 1370.00

= Rs. 205.00

c. The best stocking strategy is 400 units of scarves. This alternative is preferred regardless of
whether the total expected cost or the total expected profit criterion is used.

19
 Inventory Control Systems

All inventory systems are plagued by two major problems: -

1. Maintaining adequate control over each inventory items, and

2. Ensuring that accurate records of stock on hand are kept.

We present below some inventory systems used in practice.

a. Optional Replacement System: An optional replenishment system forces reviewing the

inventory level at a fixed frequency (such as weekly) and ordering replenishment supply if the
level has dropped below some amount. This is a P – model. The maximum inventory level (called
M) can be computed based on demand, ordering cost, and shortage cost. Because it takes time and
costs money to place an order, a minimum order of size Q can be established. Then, whenever
this item is reviewed, the inventory position (called I) is subtracted from the replenishment level
M. If the number (call it q) is equal to greater than Q, order q. Otherwise, forget until the next
review period.

Stated formally, q=M-I

If, q≥ Q, order q. Otherwise, do not order any.

b. Two – Bin System: In two-bin system, items are used from one bin, and the second bin provides
an amount large enough to ensure that the stock can be replenished. This is a Q - model. Ideally
the second bin would contain an amount equal to the reorder point (R) calculated earlier. As
soon as the second bin supply is brought to the first bin, an order is placed to replenish the
second bin.

Actually, these bins can be located together. In fact, there could be just one bin with a divider
between them. The key to a two-bin operation is to separate the inventory so that part of it is held
in reserve until the rest is used first.

c. ABC Inventory Planning: If the annual usage of items in inventory is listed according to rupee
volume, generally, the list shows that a small number of items account for a large rupee volume
and that a large number of items account for a small rupee volume. The ABC approach divides
this list into three groupings by value: A items constitute roughly the top 15 percent of the items,
B items roughly the next 35 percent, and C items the last 50 percent or so. Of course, in practice,
segmentation may not occur so neatly. The objective, though, is to try to separate the important
from the unimportant ones.

The purpose of classifying items into groups is to establish the appropriate degree of control over
each item. On a periodic basis, for example, class A items may be more clearly controlled with
weekly ordering, B items may be ordered biweekly and C items may be ordered monthly or

20
 bimonthly. Note that the unit cost of items is not related to their classification. An A item may
have a high dollar volume through a combination of either low cost and high usage or high cost
and low usage. Similarly, C items may have a low rupee volume because of either low demand or
low cost.
Sometimes an item may be critical to a system if its absence creates a sizable loss. In this case,
regardless of the items’ classification, sufficiently large stocks should be kept on hand to prevent
runout. One way to ensure closer control is to designate this item an A or a B, forcing it into a
category even if its rupee volume does not warrant such inclusion.

d. VED (Vital, Essential, Desirable) Analysis: In addition to rupee value of materials, there is
sometimes a “nuisance’ value to the materials. What we discussed in ABC analysis was the rupee
value due to their presence. Whereas the “nuisance value” is the cost associated with materials
due to their absence.

Certain materials are important by their absence and not necessarily by their presence. If they are
not available, they hold up production and, therefore, there are high costs of shutdown or slow-
down of production. By themselves these materials may not be priced high in the market. The
investment in these materials may be small but lack of any of them, the production process may
come to a grinding halt.

This gives us a clue to another kind of classification of materials which has to deal with the
critical nature of the items i.e. whether they are “vital” to the production process, or ‘essential” or
just”desirable”. “Vital’ (V) items have extreme criticality, desirable (D) items are not critical and
essential (E) items are somewhere in-between. Such VED ranking can be done on the basis of the
shortage costs of materials, which can be either quantified or qualitatively expressed.

e. Combination of ABC and VED Analysis: Since the conventional ABC analysis and VED
analysis are both important, it is possible to combine both these classification and arrive at a joint
classification of materials. Such a classification is shown below: -

V E D
A AV AE AD

B BV BE BD

C CV CE CD

As shown above we may classify materials into nine categories for management control purposes.
The classification is for giving guidelines or helping management to have better control over the
materials.

21
f. FSN (Fast, Slow, Non-moving Items) Analysis: It has been found in some organization that they
have large inventories of materials sitting idle in the stores for years together. Sometimes, nobody
is aware of their existence. There may be a variety of reasons for this, chiefly the obsolescence of
the materials. The materials might have been bought in large quantities a few years ago and
would have become obsolete over time and therefore forgotten. These materials, however, have
some salvage value or at least a scrap value.

It has been found advantageous in such situations, to carry out an analysis on the basis of the rate
of movement of the materials in the stores like fast-moving, slow-moving and non-moving items.
The non-moving items can be, in such cases, listed and the list can be sent to the different
departments or organizations that may be interested in these materials. The items which have
become obsolete can be sold for whatever salvages value they can obtain.

ABC Analysis of Selective Inventory Control

Ex. In the past Subimal Works has used a P-Model inventory system that involved taking a
complete inventory count of all items each month. However, increasing labour costs are forcing
Subimal works to examine alternate ways to reduce the amount of labor in inventory stockroom, yet
without increasing other costs, such as shortage costs. Here is random sample of 20 of their items.

Item Annual Item Annual Item Annual Item Annual Item Annual
Number Usage Number Usage number Usage Number Usage Number Usage
1 1500 5 9600 9 800 13 42000 17 4000
2 12000 6 750 10 15000 14 9900 18 61000
3 2200 7 2000 11 13000 15 1200 19 3500
4 50000 8 11000 12 600 16 10200 20 2900

Annual usage figures are in Rupees.

a What you recommend them to do to cut back their labour cost?
b Item 15 is critical to continued operation. How would you recommend it to be classified?

Ans.: We shall suggest them to implement ABC Plan as follows: -

Percentage of Total Usage Percentage of Total Usage

Annual Annual
Item No. rounded to two places of Item no. rounded to two places of
Usage usage
decimal decimal
18 61000 24 17 4000 4
4 50000 20 19 3500 1
13 42000 17 20 2900 1
10 15000 6 3 2200 1
11 13000 5 7 2000 1
2 12000 5 1 1500 1
8 11000 5 15 1200 0
16 10200 4 9 800 0
14 9900 4 6 750 0
5 9600 4 12 600 0

22
The purpose of classifying items into group is to establish the appropriate degree of control over
them. Segmentations, in practice, do not occur very neatly.

a We classify A items as Item No. 4, 13 and 18; B items as 2, 5, 8, 10, 11, 14 and 16; C
items as remainders.
b We shall classify the Item no. 15 as A item.

23
 STORE ISSUE PRICING
Ex. The Store transactions during a month are as follows: -
Opening Balance: Nil;On 1st, receipt of 400 units @ Rs. 2.50/-; On 6th, issue of 200 units; On
th th th
8 , receipt of 500 units @ 2.80/-; On 12 , issue of 300 units; On 15 , receipt of 300 units @
2.40/- units; On 28th, issue of 500 units.
Price the issue on (I) FIFO; (ii) LIFO; (iii) Simple Average and (iv) Weighted Average
Method.
Ans.:

FIFO LIFO Simple Average Price Weighted Average Price

Issue Issue Issue Remarks Issue

Date Date Date Date Remarks
Qty. Rate Qty. Rate Qty. Rate Qty. Rate

6th 200 2.50 6th 200 2.50 6th 200 2.50 6th 200 2.50
[200 x 2.50 +
200 2.50 [2.50+2.80]
12th 12th 300 2.80 12th 300 2.65 12th 300 2.71 500 x 2.80 ] /
100 2.80 /2
(200 + 500)

[ 2.80+ [400 x 2.71 +

400 2.80 300 2.40
28th 28th 28th 500 2.60 2.40] 28th 500 2.58 300 x 2.4] /
100 2.40 200 2.80
/2 ( 400 + 200)

Ex. Priced Stores Ledger

The following is an extract of the records of receipt and issue of Sulpher during a month.
Issues are priced on the principles of First In First Out (FIFO) method. The Stores verifier has found
a shortage of 10 MT on 22nd and left a note accordingly.

Priced Stores Ledger

Description : Sulpher Maximum Level: 600 Store Code No.: 02 340 751 7
Normal Lead Time: 15 Minimum Level: 200 Unit of Qty. : MT
days. Reorder Level : 300 Location : Godown No. 3

Vouch Quantity Rate Value

Date Source
er No. Receipt Issue Balance Receipt Issue Balance
1 - Balance - - 500 200.00 - - 1,00,000/-
3 - Issue - 70 430 200.00 - 14,000/- 86,000/-
4 - Issue - 100 330 200.00 - 20,000/- 66,000/-
8 - Issue - 80 250 200.00 - 16,000/- 50,000/-
38,000/
13 V/1 Receipt 200 - 450 190.00 - 88,000/-
-
14 - Return 15 - 465 200.00 3,000/- - 91,000/-
16 - Issue - 10 515 200.00 - 2,000/- 98,600/-
45,600/
20 V/2 Receipt 240 - 525 190.00 - 1,00,600/-
-
24
 Issue
22 - Adjustm - 10 515 200.00 - 2,000/- 98,600/-
ent

75 200.00 15,000/-
24 - Issue - 215 -
225 190.00 42,750/-
40,850/-
60,800/
25 V/3 Receipt 320 - 535 190.00 - 1,01,650/-
-
26 - Issue - 115 420 190.00 - 21,850/- 79,800/-
27 - Return 35 - 455 190.00 6,650/- - 86,450/-
19,000/
28 V/4 Receipt 100 - 555 190.00 - 1,05,450/-
-
N.B.:
a RETURN means return from production shop.
b The deficiency of 10 MT found on “Stock Verification” has been written off as Stores Issue at
the earliest possible date.

25
 SELECTIVE INVENTORY CONTROL TECHNIQUES
SL. NAME BASIS OF CLASSIFICATION REMARKS
NO.
1 ABC ANALYSIS THE INVENTORY ITEMS ARE
A=HIGH
CLASSIFIED ON THE BASIS OF THEIR
CONSUMPTION;
(ALWAYS,BETTER,CO USAGE (OR CONSUMPTION VALUE) IN
B=MODERATE;
NTROL) MONETARY TERMS C=LOW COSUMPTION
VALUE ITEMS
2 VED ANALYSIS BASIS OF CLASSIFICATION IS THE THE VED ANALYSIS
CRITICATILY OF THE ITEMS IS MAINLY FOR
(VITAL ESSENTIAL, SPARE PARTS
DESIRABLE)
3 SDE ANALYSIS USEFUL IN THE
STUDY OF THOSE
(SCARCE, DIFFICULT, ITEMS WHO ARE
EASILY AVAILABLE SCARCE IN
IN THE MARKET) AVAILABILITY
4 HML ANALYSIS SIMILAR TO ABC ANALYSIS, EXCEPT USUALLY HELPS
THAT COST PER ITEM IS TAKEN CONTROL OVER
(HIGHEST COST, CONSUMPTION AT
MEDIUM COST, LOW DEPARTMENTAL
COST) LEVEL
5 FNSD ANALYSIS ITEMS ARE CLASSIFIED IN THE USED TO COMBAT
DESCENDING ORDER OF THEIR USAGE OBSOLESENCE IN
(FAST MOVING, (MOVEMENT ) VALUE ALL TYPES OF
NORMAL MOVING, INVENTORY
SLOW MOVING, DEAD
ITEMS)
6 XYZ ANALYSIS BASED ON CLOSING INVENTORY
VALUES OF DIFFERENT ITEMS IN
(HIGH INVENTORY STORES
VALUE, MODERATE,
LOW)
7 SOS ANALYSIS BASIS OF CLASSIFICATION IS THE PROCUREMENT &
NATURE OF SUPPLIERS HOLDING
(SEASON, OFF STRATEGIES FOR
SEASON) SEASONAL ITEMS
LIKE AGRICULTURAL
PRODUCTS
8 GOLF

(Govern. Controlled,
Ordinarily available,
Locally available,
Foreign items)

26
 MUSIC-3D Analysis or Multi-unit selective inventory control three dimensional approach

The fundamental idea behind selective inventory control techniques is to put efforts where results are
worth. It is well known that even if an organisation uses millions of items, only few items are
important – from the finance view, availability considerations, seasonality, criticality of performance,
etc. the materials are classified according to their importance and increased attention is paid to these
important items.

We have seen that the ABC classification or always better control system is based on the annual
consumption value; SDE classification is based on availability and VED classification is based on
criticality of items. If we adopt three levels for each, such as ABC, SDE and VED, then we land in 3
x 3 x 3 = 27 groups and it becomes a difficult job in practice to follow-up.

Hence we advocate two levels for each of the three dimensions – high consumption value / low
consumption value; long lead-time / short lead-time and critical / non-critical items. This integrated
picture is called MUSIC-3D or multi-unit selective inventory control three dimensional approach.

We have the following table indicating the 2 x 2 x 2 = 8 categories of items; as explained earlier, the
categorization could be arbitrary and started on a thumb rule basis.

High consumption value items Low consumption value items

Long lead time Short lead time Long lead time Short lead time
Critical Critical Critical Critical
Critical LLT 1 SLT 2 LLT 3 SLT 4
HCV HCV LCV LCV
Non-Critical Non-Critical Non-Critical Non-Critical
Non-
LLT 5 SLT 6 LLT 7 SLT 8
critical
HCV HCV LCV LCV

These classifications should be reviewed periodically, at least once a year. It may be ideal to carry out
this analysis separately for each class, like raw materials, imports, spares, etc.

Let us consider, as an example, the item in cell 3. These are critical from operation view; take a long
lead-time for procuring, but with low consumption value. For this category, the purchase quantity
must be very large with annual ordering. The coverage can be even two years, as the working capital
involved is small, provided there is adequate space and the items are not perishable in nature.

Items in cell 6 constitute the opposite items in cell 3. They have short lead-time, high consumption
value and are non-critical. The purchase quantity should be nil, as the item is non-critical and have
high consumption value, or these items should be bought as and when needed, provided the
competent authority permits. In other words, the purchase should be done just at the time of
requirement basis, and delegated to the highest level, with the associated delays in the procurement.
The concept of zero inventory is applicable here, as the value of the item is very high but it is non-
critical in nature.

The problems before the purchase manager are the items falling in cells 1 and 2, which are critical in
nature and have high consumption value. He has to exercise very strict control on these items by very

27
accurate forecasts and adequate consumption norms for these items. The ordering may be on a
staggered basis, with frequent orders or weekly deliveries. The expediting and follow-up has to be
maximum for these items, he has to really chase these items, and he may develop as many sources as
possible for each of these items. The posting for these items should be immediate and up-to-date. The
information on stocks status, consumption, withdrawal, delivery, supply status, transport status etc
must be continuously monitored through a computer on these few items falling under cells 1 and 2.

MUSIC-3D, unlike other approaches, simultaneously considers all the three dimensions, namely
availability, criticality and consumption value and thus is a powerful approach in the direction of cost
reduction and application of scientific management practices in the making of materials as a profit
center.

28