Why Study Inventory Management

Published on June 2016 | Categories: Documents | Downloads: 32 | Comments: 0 | Views: 258
of 11
Download PDF   Embed   Report

Comments

Content


Why Study Inventory Management?
Professors David Pyke and M. Eric Johnson
*
Executive Summary
Studying inventory management can yield significant improvements in both
inventory cost and customer service levels. In the context of a real case, this article
illustrates the process of assessing the opportunities and the benefits that can accrue.
We refer to, but do not present, advanced, yet relatively simple, formulas that can be
implemented on a spreadsheet. Applying these formulas can yield near-term
improvements in both inventory investment and service level through implementing
the right inventory levels, without requiring significant changes to the supply chain.
his process, which we call inventory rationali!ation, is illustrated by moving from
the "current position# point on the graph below to the "efficient frontier# curve.
$eyond inventory rationali!ation, supply chain integration can help reduce forecast
error and therefore can actually shift the entire efficient frontier to a whole new level.
We describe several supply chain initiatives and illustrate their effects on inventory
investment, as can be seen by the reduced-forecast-error curves on the graph. %inally,
we illustrate the use of inventory tools as an input to strategic decisions such as
outsourcing to &hina.
'his note was written by (rofessors )avid (y*e and +. ,ric -ohnson on -uly ./, 0112. 3eproduction without written permission is
prohibited. (ermission to copy may be requested by contacting Ann $unnell at ann.bunnell4dartmouth.edu. he authors may be
contacted by emailing david.py*e4dartmouth.edu and m.eric.5ohnson4dartmouth.edu.
Introduction
+i*e ,dwards, special assistant to the president and owner of &orrugated (roducts
6&(7
.
, loo*ed over the shop floor and wondered if all those piles of inventory were
necessary. &ustomer service was critical to &(, and +i*e had learned enough about
inventory to *now that a significant decrease could be detrimental to &(8s reputation.
9owever, was it possible that those leaning towers of corrugated cardboard contained
more than enough inventory: Was there money to be saved:
&orrugated (roducts was a ;<1 million =orth &arolina company that bought >x/ foot
sheets of corrugated cardboard from large suppliers, and then cut and scored them
according to customer-specific designs. he result was a flat piece of cardboard that
&(8s customers could easily bend into a box. &( made about ?21 different products
or stoc* *eeping units 6S@As7 for delivery to customers who were all within about a
day8s drive from &(8s plant.
+i*e was hired to lead operational change, and ultimately, if things wor*ed out with
the current president and owner, to buy the company. 9is first idea was to implement
some ideas he had learned about inventory management, but he wanted to be fairly
certain that improvement was possible before spending political capital on
widespread change. After a B hour phone call to one of his business school
professors, +i*e had a plan that should help him understand if a reduction in their ;2
million of inventory was possible without damaging customer service. %urthermore,
the president had as*ed him about two possible ma5or initiatives that had been all
over the local press recently C outsourcing to &hina and supply chain management.
9e wondered how his inventory management pro5ect would inform these discussions.
ssessing the !""ortunity
+i*e8s professor suggested that the first step would be to collect some basic data on a
representative sample of items.
0
After some thought, +i*e decided that nine items
would be sufficient. Dead times to replenish finished goods inventory of these items
were one wee*, and &( used a service level of E/F for all items. +i*e8s data are
presented in able .. Annual demand for these items was over 0.> million units, and
the item values varied widely. &(8s current reorder points and order quantities are
presented in the table.
G
=ow, the dollar value of the inventory is defined as the
product of the average inventory and the item8s unit value. =otice that the dollar
value of the total inventory of these items was ;G2,//E. hese data were readily
available from sales records, so +i*e did not have to spend excessive amounts of
time collecting it. 6he standard deviation seen in the table is a measure of forecast
accuracy, or equivalently, variability or dispersion of demand.
>
7
.
he data for the items examined in this case are directly from the company, but other particulars and
names have been disguised to preserve confidentiality.
0
%or a more complete treatment of assessing the opportunity, see 6(y*e, 011>7.
G
he reorder point is the inventory level at which an order is triggered, and the order quantity is the amount
of that order.
>
See 6-ohnson, 011G7.
0
#a$%e &' (orrugated Products Data and (urrent Inventory Po%icies
After collecting these data, +i*e had to study and apply advanced inventory formulas to
create recommended inventory policies 6that is, order quantities and reorder points7.
%ortunately, the formulas were not too complicated, and +i*e8s professor had them built
into a fairly simple spreadsheet.
2
he question at this stage was whether &(8s current
inventory policies were correct, outdated or perhaps wrong.
he results were startling. 9e discovered that some items had way too much inventory
6Item 0.22, for instance7, while others had too little 6Item E?</7. Asing the spreadsheet,
+i*e then calculated the current service levels of the nine items, assuming that the
current policies were followed. See able 0 for these results. hese numbers enabled
him to chec* with his inventory managers and procurement personnel to see if his results
passed the "sniff test.#
2
%or more detail on finding these advanced formulas, see 6-ohnson, 01127, or 6Silver et al., .EE/7, &hapter
<.
G
#a$%e )' (urrent Service *eve%s
+i*e8s recommended inventory policies included both revised order quantities, which
capture the tradeoff between inventory holding costs and fixed ordering costs, and revised
reorder points, which capture the tradeoff between holding costs and service levels.
hese policies, along with the resulting dollar value of the inventory, are presented in
able G. $ased on the inventory formulas, the service level for all these items would be
E/F.
#a$%e +' ,ecommended Inventory Po%icies and the ,esu%ting Inventory -a%ue
Hne can see that the potential savings were huge. Doo* at Item 0.22 for example.
Inventory managers at &( had set the order quantity to G1,111 units and reorder point to
>2,111 units, leading to an average inventory value of ;..,111 for this item 6able .7.
he advanced formulas, however, recommended an order quantity of G.,?<1 and a
reorder point of 0/,0<0. =ow the two order quantities were fairly similar, but the
recommended reorder point was significantly lower than the current oneI and the
resulting average inventory value was ;?,?2G, a reduction of >1F. he total value of the
recommended inventory was ;0>,<2E, or a savings of G.F from the current value of
;G2,//E. In addition, &( would be providing a consistent E/F service level for each
>
item. +i*e could ad5ust the order quantities and reorder points to more round numbers,
but that would not significantly change the results.
If these nine items were representative of &(8s entire inventory of ?21 S@As, they could
save over ;..2 million in inventory. =ote that this is a one-time savings. =evertheless,
the ongoing savings at a carrying charge of 01F is over ;G11,111 per year, every yearJ
All because of applying a few advanced, yet simple, inventory formulas on a spreadsheet.
+any firms have far more than ;2 million of inventory, so the potential is impressive
indeed.
=ow it is important to point out that the savings will not appear instantaneously. Items
that have been operating at a low service level will li*ely require immediate purchase,
whereas it will ta*e some time to sell off items that have too much inventory. If a firm8s
inventory turnover averages two to four turns per year, it will often ta*e one to two
quarters to see the results of an inventory improvement li*e that at &orrugated. $ut if the
opportunity assessment yields results of this magnitude, it will be well worth the wait.
In our experience with do!ens of firms, both from our own consulting as well as from
student pro5ects, we have seen 5ust two cases where the potential savings was minimal.
In every other case, the savings ranged from about .1F to upwards of G2F. And in
almost every case, the firm could both save inventory investment and improve service at
the same time. Hne *ey insight from our experience, and from the formulas, is that if the
standard deviation of forecast error is less than about 01F of the mean lead time demand,
the opportunity for savings is often on the lower end of the spectrum.
Sensitivity na%ysis
+i*e was quite excited about these results, and he immediately called his professor.
After another B hour conversation, +i*e decided to loo* at the effects of operational
improvements on &(8s inventory. %or instance, what if the president and owner decided
that he was spending too much money on inventory: Alternatively, what if E/F service
was too low to be competitive in this mar*et:
Asing the same spreadsheet and formulas, +i*e performed a simple sensitivity analysis
on the service level for the nine sample items. he results are in %igure .. 3ecall that if
+i*e used the recommended formulas, his inventory investment for these nine items
would be ;0>,<2E. =ow, if he wanted to improve the service level to EE.EF, his
inventory investment would increase by >.F to ;G>,/?1. Alternatively, if he were
willing to provide E1F service, the investment would be ;.?, >21, or a savings of G>F.
his analysis, of course, must involve sales and mar*eting managers, but +i*e now had a
spreadsheet and graphical analysis that could provide valuable input to the discussion.
2
.igure &' Sensitivity na%ysis / Inventory versus Service *eve%
Strategic initiatives' !utsourcing to (hina
+i*e *new that there could be many potential benefits of outsourcing to &hina, including
dramatically lower unit cost, gaining a foothold in Asia, globali!ing the business, and
learning how to conduct business in another country. Hn the other hand, &( might run
several ris*s, including lower quality, poor delivery performance, difficulty handling
supplier relationships, and issues with the domestic wor*force. &orrugated (roduct8s
president reali!ed that one input to the &hina decision should be potential inventory
effects. And +i*e reali!ed that an analysis similar to the serviceKinventory tradeoff he
had prepared could be used.
3ecall that the current replenishment lead times were one wee* for all of &(8s products.
+i*e8s preliminary investigations suggested that lead times from &hina would be roughly
eight wee*s. 9e applied the spreadsheet with the advanced formulas to create the
tradeoff graph in %igure 0, and came to the startling conclusion that inventory investment
for the nine sample items would increase from ;0>,<2E to over ;2.,111, or more than a
.11F increase. If these items are representative of all ?21 S@As, and if +i*e were to
outsource all of &(8s products to &hina, the inventory investment would grow to more
than ;.1 million. Hf course, the unit cost savings might overwhelm the increased
inventory and transportation costs, but clearly &( should include the inventory effects in
this important decision. =ote that this same graph captures the effect of an operational
improvement. If, by sourcing from a local supplier or other operational improvements,
+i*e could reduce the replenishment lead time from one wee* to . day, his inventory
investment would decrease by about 0/F, with no decrease in service.
=ow these numbers and percent changes are specific to this case. =evertheless, the
general shape of %igures . and 0 are consistent across many inventory systems that we
have seen.
?
.igure )' !utsourcing to (hina' *ead time versus Inventory
What had +i*e learned thus far: 9e *new that &(8s current position was not optimal,
assuming that the nine items were representative. 9e also *new that if &( made some
strategic choices about customer service, he could provide valuable input to the
discussion using the inventoryKservice tradeoff curve. %inally, he had discovered that the
same tools could provide input to long term strategic decisions about outsourcing and to
initiatives for operational improvement. %igure G highlights &(8s current position relative
to an "efficient frontier# on the inventoryKlead time tradeoff curve. +i*e could easily
move &( to the efficient frontier if lead times remained at one wee*, simply by applying
the recommended inventory policies. Hr, he could shift &( along the efficient frontier,
assuming he could influence lead times accordingly. A similar picture applies to the
inventoryKservice level tradeoff.
.igure +' (orrugated Products0 (urrent Position ,e%ative to an Efficient .rontier
<
Strategic Initiatives' Su""%y (hain Im"rovements
he other initiative raised by &(8s president was supply chain management. +i*e had
studied supply chain management in school, so he *new that the potential benefits were
huge. Supply chain initiatives such as &ollaborative (lanning, %orecasting and
3eplenishment 6&(%37 can provide a supplier with visibility to their customers8
production and inventory plans. herefore, instead of reacting to customer orders, the
supplier can plan ahead based on actual customer demand, thereby reducing forecast
error. A further development, called Lendor +anaged Inventory 6L+I7, assigns the
responsibility for managing the customer8s inventory to the supplier. L+I, li*e &(%3,
gives better visibility to customer demand well in advance of when the customer would
typically place an order. %orecast error again can decrease substantially.
+i*e wondered if &( could save inventory investment if he pursued &(%3 or L+I with
their largest customers. 9e *new that &( had the capability to wor* with customers in
this way, and he was excited to get some support for the discussions. o quantify the
effect, he decided to see what would happen if, as a result of the collaboration, the
standard deviation of forecast error for each product decreased. he same spreadsheet
and inventory formulas yielded the graph in %igure >. %or example, the graph shows that
if &( could reduce the standard deviation of forecast error to <1F of its current value,
they would cut inventory investment for these nine items to about ;0.,111, or a savings
of about .>F. If they could reduce standard deviation by half, the savings would be
00F.
.igure 1' Su""%y (hain Im"rovements' ,eduction in .orecast Error
/
E
Summary
+i*e8s experience at &orrugated (roducts is similar to many other examples of dramatic
inventory improvement C lower costs and better service C that we have observed. Why
are such opportunities so prevalent: Sometimes, managers are simply too busy to review
their inventory policies. )emand patterns change, new products are introduced, supply
chains become more integrated, suppliers move offshore, but inventory policies are not
updated. In these cases, it is fairly simple to recompute the policies and ma*e the
appropriate ad5ustments. Hther times, as in &(8s case, managers do not have access to
advanced formulas and insights. he materials referenced in this note can be very helpful
in this regard, as can a number of other inventory teaching materials. hese cases are
represented by the "current position# point on %igure 2. In spite of implementation
challenges, it can be relatively straightforward to move from the current position to the
efficient frontier 6the highest curve on %igure 27, by rationali!ing inventory via advanced
inventory formulas.
Hften, however, the most impressive improvements come from shifting the efficient
frontier to a whole new level. hese shifts require fundamental changes in supplier and
customer relationships and forecast accuracy. Supply chain management innovations can
help companies achieve these results, and the inventory management tools discussed in
this note can be used to analy!e the associated inventory benefits.
.igure 2' Su""%y (hain Im"rovements' Shifting the Efficient .rontier
.1
,eferences
-ohnson, +. ,. 6011G7. Understanding Supply Chain Variability and Measuring orecast
!erformance. uc* School of $usiness =ote, )artmouth &ollege, 9anover, =9.
-ohnson, +. ,. 601127. Calculating Safety Stoc". uc* School of $usiness =ote,
)artmouth &ollege, 9anover, =9.
(y*e, ). %. 6011>7. #pportunity Assessment. uc* School of $usiness =ote, )artmouth
&ollege, 9anover, =9.
Silver, ,. A., (y*e, ). %., M (eterson, 3. 6.EE/7. $nventory Management and !roduction
!lanning and Scheduling 6G ed.7. =ew Nor*O -ohn Wiley M Sons.
..

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close