Beruflich Dokumente
Kultur Dokumente
Contents
Inventory Control
(For People Who Really Have to Do It)
Volume II in the Useful Management Series
By Robert E. D. Woolsey, Ph.D., F.I.D.S.
And
Ruth Maurer, Ph.D.
Preface: Questions and Answers About This Book
Chapter 1. A Requiem for the EOQ
The Economic Order Quantity Model
Learning the Model and the First Experience of Reality
Killing the EOQ Vampire, the Silver Stake Method
The Cost of the Item
The Holding Rate
The Setup or Ordering Cost
The Annual Demand or Annual Requirement
References
Chapter 2. Lot Sizing Methods of Inventory Control
What This Chapter Is For
The Economic Order Quantity Method
Periodic Order Quantity
Part-Period Balancing
Dynamic Programming
The Method of Silver and Meal
A Better Silver and Meal Method
Silver and Meal Quick and Dirty Fill-In-The Blank Inventory Form
First Silver and Meal Nomogram Example
Second Silver and Meal Nomogram Example
Silver and Meal Nomograph Example
Greenings Nomograph for Forecasted Demands
When Do I Use Which Method?
References
Chapter 3. The Woolsey Never-Fail Spare Parts Reduction Method
If Youre Using the EOQ, At Least Do it Right!
Setting the Scene
With the Computer Jocks and What Happened There
The Long Awaited Recalculation and What Happened Then
Icing on the Cake Recalculating the Order Point
Final Warnings and Suggestions
Flowchart of the Woolsey Never-Fail Spare Parts Reduction Method
Inventory Example Problems
The Learning Organization
Chapter 4. El Pistolero del Inventorio
La Problema y Zopilote
Venustusiano Oso
Una Pregunta por Las Gerentes
El Neuvo Metodo
El Rey
Mas Preguntas de Importancia
Reference
1 Inventory Control (For People Who Really Have to Do It) eBooks Series
2
4
14
37
45
Preface
Questions and Answers About this Book
What is this book for?
This is a cookbook of quick and dirty methods for solving problems in realworld, no B.S. inventory control.
Who is this book for?
This book is for bottom- to middle-level managers, third-world countries, and
small business where computers are either too expensive or labor is too cheap or too
uneducated to justify anything but the use of common sense.
It is also aimed at unpretentious community colleges and business schools that
want to teach people something they can use!
How smart does the reader have to be?
Smart enough to know that there are no pure technical problems in this world.
There are just technical problems imbedded in political problems. This book is dedicated to the idea that if you dont deal with the political problems as well as you
deal with the technical problems, the political problems will deal with YOU!
Does this imply that politics will be covered too?
Big time! The first author starts the book by taking on the most used (wrongly)
method in the inventory world called the economic order quantity. He points out
that this method (like any other) depends on data you can trust, such as delivery
dates and the real need dates seldom given you by the sales types. He also tries to
warn the unwary reader that every method carries a bag of assumptions that need
to be carefully addressed before embarking on any method. A number of chapters
on political aspects of inventory control based on the experiences of the authors are
gleefully included. These chapters are more important than the others.
Do I need calculus, algebra, or statistics for this book?
No, common sense only is required.
Does this book come with a computer program or a diskette?
NO. This book is almost totally made up of certifiably obsolete inventory methods that no up-to-date large corporation or agency would even consider using.
They certainly can be programmed but the approach of the book is for the by-hand
user to have something they can use and understand.
Im an assistant professor at a business school, can I use this book?
YES, if you ever met a payroll in the real world.
Otherwise, NO!
2 Inventory Control (For People Who Really Have to Do It) eBooks Series
Preface
3 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1(Political)
4 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
The next day I showed up and with pencil and paper drew the usual graph of the
Order Quantity, Q, versus time, giving the well-known sawtooth wave shown in
Figure 1.
Q
Q/2
Figure 1
I explained that if we started with Q items in inventory and used them up at some
(known and deterministic) rate, we would run out at time T. However, at the lead
time, we would launch another order that, surprise, would arrive exactly at time T,
and so on and on. Clearly, if we start with Q items, go down to zero, restart with Q
items and go down to zero, and etc., all we have to do is draw a line at Q/2 through
the diagram, as shown above, to get the average inventory. The next step was to get
him to agree that if we had a cost C of the part and a holding rate of I, then CIQ/2
was the average annual holding cost of the item. Clearly, as the number of items in
the order quantity Q increased, the cost increased. I then walked him through the
setup or ordering cost part. If it cost the Air Force S bucks to launch an order, and
the order size was the requirement R, divided by the order quantity Q, then clearly
S(R/Q) would get smaller as Q increased. I then wrote out, just as I had been taught,
the classic formula below.
SR
TEC = CIQ
2 +Q
He had no trouble with the idea that the minimum cost had to occur where the
rising holding cost line crossed the falling ordering cost line, and that it would be at
the lowest point of the total expected cost line made up of the sum of annual holding and ordering costs.
I was delighted; victory was in my grasp. Then things turned to #$%@. He asked
me how I got the good old EOQ from the formula. I told him that all we had to do
was to take a derivative of the total expected cost formula, which I did on the spot.
He looked a bit uncertain, and politely asked (after all I was a second lieutenant)
how I knew that formula was right. I cheerfully answered that all we had to do was
to take a second derivative. He then, still politely (I was still a second lieutenant)
informed me that he didnt know what in Hades a first derivative was, much less a
5 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
second one. At this point I did what I should have done in the first place. I shut up,
thanked him for his time, and learned to do it his way, before making any more hotdog suggestions for improvement.
Killing the EOQ Vampire, the Silver Stake Method
Many years have passed, but the experience is still fresh in my mind. Come with
me now as I take this model apart like a clock, and perhaps convince you never to
use it again. First, let us write the usual formula for the EOQ as follows, using an
appropriate reference that doesnt believe it either: Woolsey [2]. We define as above
the TEC, total expected cost, as the sum of the annual holding cost plus the annual
ordering cost, or:
SR
TEC = CIQ
2 +Q
Where, as usual:
C is the cost of the item in $/unit,
I is the holding rate in % of price/item/unit time,
S is the set-up or ordering cost in $/order,
R is the annual requirements in units, and
Q is the order quantity in # items/order.
In the rest of this chapter, I am going to replace any symbol about which we have
some doubts with a LARGE question mark. Note that Q is a variable depending on
all the others (we will only concern ourselves with the known constants). We will
now take each of the known constants, in turn, and treat them with the amusement
they deserve.
The Cost of the Item, C
Let us begin with C, the cost of the item, as our first doubtful symbol. For openers, I always get amused when the costs are given in most textbooks without any discussion of first-in-first-out, last-in-first-out, lower-of-cost-or-market, or how-wedo-it-here costing. It is always assumed that the cost accounting system is irrelevant
for lot-sizing purposes. Secondly, I am a firm believer that the only real measure is
dollars after tax at present value, adjusted for inflation. For those of you with short
memories, let me recall for you the time in the good old United States of America
when a peanut farmer was president and:
The inflation rate was 20+ percent.
The prime rate was 20+ percent.
The catalogs no longer came with price lists.
When you called for a price and they gave you one,
it was good only for a short time.
When you ordered and they billed you, it was more.
6 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
Now this little inflation factor becomes even more important among our readers
in countries such as Bolivia, Israel, and Argentina, where people have been paid
twice a day so they could spend it before the exchange rate went down some more.
Trying to use an EOQ in such countries, even as close as Mexico, can be a terrifying
experience; I dont recommend it. On the basis of this argument we award our first
question mark to C and our formula above becomes:
SR
TEC = ?IQ
2 +Q
The Holding Rate, I
We now come to the well-beloved holding rate, usually given as a percentage of
the price per item per unit time. I have been greatly amused for over two decades
now when the typical textbook graph of the total expected cost of the EOQ (plotted against the order quantity) looks like Figure 2 below.
TEC
TEC
Error in TEC*
TEC*
CIQ
SR/Q
Q
Q1
Q*
Q2
Error in Q*
Figure 2
7 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
It is immediately seen that the TEC curve is very flat in the area of the optimal
order quantity, Q*, because of the small angle of the holding cost line usually shown
in most textbooks, as in the above picture. In this case, we can assume that we could
be in error around the optimum Q* between the points marked Q1 and Q2. Note
the wonder of the result; a large error in the constants that make up the optimal
order quantity Q*, when reflected on the TEC line, is seen to be of minimal effect
on optimum total cost TEC*.
In short, a big error in Q seems to result in a small error in TEC. This seems to
imply that if you feed this formula some garbage values of C, I, S, or R, it really wont
hurt you very much in terms of total expected cost. There is only one other thing in
the world that you can feed such garbage to and get something good out of; its
called a pig. If you feed a pig garbage, eventually you get bacon. However, it is
important to note that the creation of bacon requires total commitment from the
pig. Total commitment to the EOQ may be equally fatal to your profits. We will show
this next.
I argue that the variance of the holding rate is more than most people imagine.
Most people, and businesses, tend to assume that the holding rate is the cost of
money, i.e., prime rate plus points. I believe this to be a bloody dangerous assumption, and I will demonstrate it. I believe that there are really only four kinds of products that most businesses really care about. These products are measured only by
market demand and profit margin or markup as shown below in Table 1.
Table 1: The Four Types of Products
Type
1
2
3
4
Mkt. Demand
High
High
Low
Low
Markup
High
Low
High
Low
Translation
Customers beating on doors, profit obscene
Customers beating on doors, profit zip
Nobody wants it, profit obscene
Nobody wants it, profit zip
One doesnt have to be too smart to agree that, if you could, you would only stock
type 1, high market demand and high markup. Let us now realize what else this little classification tells us. It really says that, given a choice, we would like to stock the
parts with the biggest markup in an expanding market.
The next step, however, once you buy this argument, is a killer. It really implies
that the minimum holding rate is the rate of return you are getting on your highest
markup item in an expanding market, because, if you could, you would put everything there (but you cant). This means that far from being the prime rate, the holding rate is lower bounded by your highest markup on your best selling product. In
short, the line you see for holding cost in most textbooks is usually drawn like
Figure 2, reflecting cost of money, or prime rate. Clearly, you should be using a
8 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
holding rate greater than your cost of money or you have no business being in business.
Say that we discover that our best mover in our product line, which also makes a
bundle after tax at present value, has a markup of 60 percent. Lets now redraw our
graph above with the new holding rate but using the same ordering cost line as
before. This is shown in Figure 3.
TEC
CIQ/2
Error
in
O"
TEC*
SR/Q
Q1
Q*
Q
Q2
Error in O"
Figure 3
From the above graph, we note that a) our minimum Total Expected Cost is much
higher, and b) with the same amount of error as in the previous graph, we can be in
deep trouble with its effect on total cost. I hope that I have now convinced you that we
9 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
now have so much variance in our holding rate, I, that we should replace it with the
appropriate question mark in our formula above, resulting in:
SR
TEC = ??Q
2 +Q
The Set-Up or Ordering Cost, S
With this one, I usually get strong arguments like: Why, we have that data to four
decimal places; it comes from job standards taken by our industrial engineers. Now
let us say that we have a happy inventory clerk working efficiently in his tool crib.
He is approached by an industrial engineer who says, somewhat speedily:
Hitherehappyinventoryperson,Imyourlocalindustrialengineerheretodoatimeand
motionstudytoincreaseyoursafetyandproductivity,just ignore me.
While the clerk is trying to figure out what he said, the IE produces his clipboard
and punches his stopwatch to time the next order that has just come in. Now I am
here to tell you as a senior member of the Institute of Industrial Engineers that I
know what will happen next. That clerk will (particularly if unionized), before your
very eyes, turn into a good imitation of Mikhail Baryshnikov doing Swan Lake, and
make every movement at this speed or slower. This, ladies and gentlemen, is what you have to four decimal places as job standards. What you really have
to four decimal places is nonsense. The only way to set real job standards is to do it
yourself, or take your data from behind a distant mountain with binoculars. On this
basis, I award our third question mark to our ordering cost, S, resulting in:
?R
TEC = ??Q
2 +Q
The Annual Demand or Annual Requirement, R
Showing that the annual requirement is deserving of a question mark is really
shooting fish in a barrel. Lets face it, we know where the requirement comes from.
It is extracted by the right- or left-hand rule from some appropriate orifice by the
marketing department. Everybody knows that marketing routinely inflates their
demand forecasts. Let us, for the inexperienced, discuss why this happens. It should
be clear to anyone with production experience that where you stand depends on
where you sit. A sales manager, when asked by his CEO about his projection for the
next quarters sales, must say something like this (or lose his job):
Sales, sir, will be UP, and I mean UP, were going to blow the competition
AWAY!
Now lets imagine for a moment, that the sales manager has an honesty attack and
tells the ultimate biggy something like:
Well, boss, we are going to be hanging on by our fingernails this quarter, our
competition is going to eat us alive.
10 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
11 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 1
is unchanging through the ages, yea verily, like unto the Rock of Gibraltar, it is (are
you ready for this), the:
2
But wait, that 2 is there because we have made the heroic assumption that we have
constant demand, which means that the demand pattern looks like the picture we
started this whole thing with (see Figure 4 below).
Q
Q/2
Figure 4
Has anyone who is reading this in their life ever seen a demand pattern that looks
like that? I have now worked and taught on five continents, and I have yet to see an
example of constant demand. We know perfectly well that demand patterns looks
more like a graph of earthquakes measured on the Richter scale than the graph
above. After all, how often have you seen your marketing types penalized for giving
you nonsense production forecasts? We experienced types also know that the zero
on the graph above is really somewhere below the bottom of this page, the difference being known as, dare I say it, SAFETY STOCK. Under these circumstances, if
you want to get Q/2, all you have to do is close your eyes and draw a line anywhere
on the graph. It really wont make any difference, trust me.
So it turns out we cant even trust the 2. At this point we award another question
mark to the 2, and a final one to the variable Q, giving the desired result:
?? = ?
TEC = ???
? + ?
Recommendation
If you, or your firm are using the EOQ model, does it occur to you that an
accounting audit of costs might be in order? If you continue to love and use the EOQ
without knowing what it is costing you, I can only suggest that you deserve each other.
12 Inventory Control (For People Who Really Have to Do It) eBooks Series
References
1. Osteryoung, J. S., Nosari, Eldon, and McCarty, Daniel E., Use of the EOQ
Model for Inventory Analysis, Production and Inventory Management, Vol. 27,
No. 3, (1986), pp. 39-46.
2. Woolsey, R.E.D., and Swanson, H.S., Operations Research For Immediate
Application, A Quick & Dirty Manual, New York, Harper & Row Publishers,
(1975), pp. 39-41.
3. Woolsey, R.E.D., and Lienert, C. E., Ordering Inventory When The Forecast Is
Ridiculous, Production and Inventory Management, Vol. 27, No. 1, (1986), pp.
144.
The above paper appeared originally as:
4. Woolsey, R.E.D., A Requiem For The EOQ: An Editorial, Production and
Inventory Management,Vol. 29, No. 3, Third Qtr,. 1988, pp. 68-72.
13 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
14 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Finally, if you can afford it, and you can control your suppliers and your sales forecast sufficiently, you shouldnt touch any of the above methods with a barge pole
but should use a lot size of 1.
The Economic Order Quantity Method
This method is rightly considered the fundamental model in inventory control
and can, correctly used, still generate more consistent results than rule of thumb
methods. The usual reference is to Harris [1]. This method makes the assumption
that we start off with Q items in inventory, over a given time, we will use these items
at a known and constant rate until we run out. Sometime before running out, we
note the bloodshot eyeball of the inventory control clerk looking ahead and realize
that we are going to run out! At some time before running out, the clerk launches
another order. This trigger point is a set level of inventory that will just last until we
run out, using the present demand rate. In short, the Trigger Point may be defined
as (Annual Demand)*(Lead time in days)/(365 days). In theory, if the clerk launches the order on or before the lead time for it to arrive, this persons backside is sufficiently covered. It is, however, good to remember that the lead time for the order
to arrive is supplied by the vendor, who has every reason to lie to you. If he or she is
sufficiently truthful about the extended time it will take to deliver your order, you
just might seek out another supplier. If the order arrives exactly at the time you run
out, you have the beloved sawtooth shaped curve seen below.
Q
Q/2
Figure 5
All that is required in the above situation is to note that if we start with Q in
inventory and go down to zero repeatedly, we can draw the average inventory line
seen above labeled Q/2.
We now define that the cost of the item is C dollars per unit. We further define
that the holding cost per item per day (expressed as a percentage of the cost, C) is I.
For example if we have an item that costs $25 dollars with an annual holding rate
of 12 percent, and our order quantity is 40 items per order, the average holding cost
per year is:
Holding Cost = C * I * (Q/2) = $25 * .12 * (40/2) = $60
15 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
In short, we can express the annual holding cost as AHC = CIQ/2. It is instructive
to note the dimensions of the above holding cost explained below.
Holding Cost ($/year) = (Price ($/unit))*(Holding Rate ($/$ . year))*
(Q units/order)/(2 (1/order)).
As it is assumed that C and I are known, we just have the relationship Cost =
Constant * Q. This may be easily graphed with Q versus cost as shown in Figure 6.
Cost
O
Figure 6
What the above graph tells us is: The more we hold in inventory, the more its
going to cost us in holding cost. Now lets derive the second part of the Economic
Order Quantity Model, the annual ordering cost.
Let us define that the cost to launch an order, get it in, and shelve it is S dollars per
order. Let us also define the forecasted demand for the year as D items per year. Now
if we order Q items in an order, it should be apparent that the number of orders per
year is:
# orders per year = (D items per year)/(Q items in an order).
And it follows, using the above example, that if we have an annual demand of 120
items and an ordering cost of $20 per order, the cost of such annual ordering should
be:
Annual Ordering Cost = S * D/Q = $20 * 120/40 = $60.
16 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Figure 7
What the graph in Figure 7 tells us is: The more often we order, the more its
going to cost us in ordering cost. Combining the two graphs in one where Q is now
plotted against both costs, we create the combined graph in Figure 8.
Cost
Figure 8
17 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Picking any points on the Q line, say Q1, Q2 and Q3, we add up the contribution
of both the holding cost and the ordering cost to generate the graph for the Total
Annual Cost as shown in Figure 9.
Cost
O
Figure 9
The graph in Figure 9 tells us that we need to balance the contributions of holding and ordering costs to minimize our total cost. The minimum will obviously
occur where the increasing holding cost line crosses the decreasing ordering cost
line. Put another way, this is where CIQ/2 = SD/Q.
However, solving the above expression for Q gives us the famous Economic Order
Quantity, Q*, or:
2*S*D
CI
Q*=
Lets now take the following example, and apply the method above:
Month
Demand
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
10
10
11
10
18 Inventory Control (For People Who Really Have to Do It) eBooks Series
12
10
Chapter 2
2*20*120
= 40
25*.12
This tells us that the cost minimizing lot size is 40 items an order. Now lets check
this month by month to confirm that this is the right answer:
Month
Demand
Order
Start
End
Average
1
10
40
40
30
35
2
10
0
30
20
25
3
10
0
20
10
15
4
10
0
10
0
5
5
10
40
40
30
35
6
10
0
30
20
25
7
10
0
20
10
15
8
10
0
10
0
5
9
10
40
40
30
35
10
10
0
30
20
25
11
10
0
20
10
15
12
10
0
10
0
5
We see at once that as we order three times, our annual ordering cost is 3
orders/year * $20/order = $60/year. Our annual average holding cost is
(35+25+15+5+35+25+15+5+35+25+15+5)/12 = 20 items/year * $25/item * .12 =
$60/year.
Political Discussion
The fundamental problem with this method is the assumption that we have constant demand. In other words, we assume the sawtooth illustration we saw before is
correct. The hard facts are, however, that you will be hard pressed to find a forecast
that is not subject to, at least seasonal trends, such as the example below:
Month
Demand
1
10
2
10
3
15
4
20
5
70
6
7
8
9
10
180 250 270 230 40
11
0
12
10
Using a cost of $2 per unit per month, a ordering cost of $300 and an annual total
demand of 1105, the optimum EOQ is found to be Q* = 166. Applying this to the
above problem as before we get:
Month
Demand
Order
Start
End
Average
1
10
166
166
156
161
2
10
3
15
4
20
5
70
6
180
166
156 146 131 111 207
146 131 111 41 27
151 138.5 121 76 117
7
250
223
250
0
125
8
270
270
270
0
135
9
230
230
230
0
115
10
40
166
166
126
146
11
0
12
10
126 126
126 116
126 121
19 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
We note that we order six times at $300/order, so our ordering cost is $1,800. But
our average inventory is now 1,532.5, which gives an annual average holding cost of
$2*1,532.5 = $3,065. We notice at once that our happy assumption that the minimum cost will occur when the increasing holding cost line crosses the decreasing
ordering cost line is violated. This is because we no longer have satisfied the necessary
assumption of constant demand. In short the lumpiness of demand is doing us in.
We need some way to smooth out this lumpiness. Thus, the next method.
Periodic Order Quantity
We can often reduce inventory carrying costs by finding an economic time interval between orders. Divide the EOQ found above which is ______, by the mean
demand rate (total demand divided by the number of periods), which is in this case
1105/12 = _____. Doing this, we get _______ months. We can just round this up to
___ months and apply it to the example below:
Month
Demand
Order
Start
End
Average
1
10
20
20
10
15
2
10
10
0
5
3
15
35
35
20
27.5
4
20
20
0
10
5
70
250
250
180
215
6
7
180 250
520
180 520
0
270
90 395
8
9
270 230
270
270 270
0
40
135 155
10
40
40
0
20
11
0
10
10
10
10
12
10
10
0
5
1
10
2
10
3
15
4
20
5
70
6
7
8
9
10
180 250 270 230 40
11
0
12
10
Lets recall that our ordering cost is $300/order. If we were to order only for the
first month, our holding cost would be $2/item/month *10/2 average items = $10.
As this is nowhere close to $300, we consider ordering for two months. If we ordered
the second months demand of 10 to come in at the first of January, we would hold
20 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
it all of January and half of February, which gives a cost of $2/item/month * 3/2 *
10 = $30. But to get the total cost for ordering for two months, we would have to
add in the $10 we got before for Januarys demand, giving a total of $10 + $30 = $40.
So for up to now we have:
Ordering for one month is $2 * 10 * 1/2 = $10.
Ordering for two months is $10 + $2 * 10 * 3/2 = $40.
Clearly, ordering for three months would give:
$40 + $2 * 15 * 5/2 = $115.
And ordering for four months would give:
$115 + $2 * 20 * 7/2 = $255.
And ordering for five months would give:
$255 + $2 * 70 * 9/2 = $885.
Now as ordering for four months is $255, which is closer to the ordering cost of
$300 than ordering for five months for $885, our first order would be for four
months and 55 items.
Using the space below, calculate the other orders and enter them in the table
below and calculate the expected holding and ordering cost for this method.
Month
Demand
Order
Start
End
Average
1
10
55
55
45
50
2
10
0
45
35
40
3
15
0
35
20
27.5
4
20
0
20
0
10
5
6
7
8
9
10 11 12
70 180 250 270 230 40 0
10
_________________________________
_________________________________
_________________________________
_________________________________
1
10
55
55
45
50
2
10
3
15
4
20
45
35
40
35 20
20 0
27.5 10
5
70
70
70
0
35
6
180
180
180
0
90
7
250
250
250
0
125
8
270
270
270
0
135
9
230
270
270
40
155
10
40
11
0
40
0
20
0
0
0
12
10
10
10
0
5
21 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Dynamic Programming
What this method does is to implicitly examine every possible way to order the
inventory and then choose the best of these. This method requires that you fill in
the table shown below. Note that for a 12-period forecast for one item you might
have, in a worst-case scenario, to fill in 144 entries in the table below. This means
that if you were using this method to lot-size 36,000 items for a year that you might
have to make calculations to fill in 36,000 * 144 or 5,184,000 entries. Also, the really bad news is that you have to do the whole thing over again every time ANY single periods forecast changes. The good news is that the method gives the optimum
solution, the bad news is that the computational burden is ferocious. I will cheerfully admit that there have been many advances in reducing this computational burden. For examples, see the Production & Inventory Management Journal issues for
1990-1993. However, the hard facts, in my opinion, are that this method is outstanding in theory but hard to understand, and virtually impractical in the real
world.
Lets start filling in the table as follows.
The intersection of column one and row one is the inventory cost to order for one
month in period one. We should recall from part-period balancing that this is $300
to order plus an average holding cost of $2 * 10 * 1/2 for a total of $310.
Month
Demand
1 310
2 340
3 415
4 555
5 1110
6 x
7 x
8 x
9 x
10 x
11 x
12 x
Order
1
10
620
665
765
1255
x
x
x
x
x
x
x
x
55
2
10
655
715
1065
3
4
5
6
7
15 20 70 180 250
735 925 1405 1955 2525
945
70
8
270
3055
3175
3175
3245
9
230
3395
3395
3445
10 11 12
40 0
10
3175 3485
3505
The intersection of column one and row two is the inventory cost to order for two
months in period one. As this is just $310 plus the additional holding cost of $2 *
10 units held for 3/2 of a period we get a total of $340.
The intersection of column one and row three just adds to the above $340, the
additional holding cost of $2 * 15 units held for 5/2 of a period, a total of $340 +
$75 or $415.
22 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
The intersection of column one and row four just adds to the above $415 the
additional holding cost of $2 * 20 units for 7/2 of a period, a total of $415 + $140
or $555.
The intersection of column one and row five adds in the non-trivial additional
cost of holding 70 items for 9/2 periods or $2 * 70 * 9/2, giving $1,110. Notice, this
incremental cost of $640 is greater than the additional ordering cost of $300 for
another order, so we could stop further calculations down this column as the situation will just get worse.
The intersection of column two and row one should contain the cost to order for
one month in period two, assuming that you ordered for one month in period one.
This is simply $300 plus to order, plus average holding cost of $2 * 10 * 1/2 or $310,
plus the cost of ordering 10 units in period one which is also $310 for a total of
$620. We may then proceed as before until we step over our ordering cost of $300
as before. The intersection of column three and row one should be the minimum
cost to order for one month in period 3 given that you made the best decision for
ordering for the first two months. As our choices are to order for month one in
month one and month two in month two at $620 or to order for month one and
two in month one at $340, the choice is obvious. So we add to this $340 the cost of
ordering in month three (which is $300) and the holding cost (which is $2 * 15*1/2
=15) which gives $655. The reader is invited to check the other entries in the table
if he or she is so inclined.
Month
Demand
1 310
2 340
3 415
4 555
5 1110
6 x
7 x
8 x
9 x
10 x
11 x
12 x
Order
1
10
620
665
765
1255
x
x
x
x
x
x
x
x
55
2
10
655
715
1065
3
4
5
6
7
15 20 70 180 250
735 925 1405 1955 2525
945
70
8
270
3055
3175
3175
3245
9
230
3395
3395
3445
10 11 12
40 0
10
3175 3485
3505
After the table is completely filled in, we start at the cost in the upper right hand
corner of the table (the entry in row 1, and column 12). We now look diagonally
down the table moving always down and to the left, looking for the cheapest cost on
23 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
this diagonal. Once we have found it (its $3,245) we enter the total items (= 280)
for this last order into the order row at the bottom. We then look at row one in the
last column before the column that generated the previous order, and do the procedure again, filling in the orders as shown above.
Putting the above results into our usual table gives:
Month
Demand
Order
Start
End
Average
1
10
55
55
45
50
2
10
3
15
4
20
45
35
40
35 20
20 0
27.5 10
5
70
70
70
0
35
6
180
180
180
0
90
7
250
250
250
0
125
8
270
270
270
0
135
9
230
280
280
50
165
10
40
11
0
50
10
30
10
10
10
12
10
10
10
0
5
Political Discussion
We have found the optimal solution, but the computational cost is definitely nontrivial. Further, as we must find out the optimal order now by going back to front,
this means that what we do now is a function of the forecast at the end of the forecast period. It is important to realize that the further out we get in the forecast, the
greater the expected error. This method says go out to where the forecast is least reliable, and figure out what to do now! I must tell you that I have real problems with
this concept. My experience tells me that the forecast is wrong and that it will
change. If you agree with me, this says that the computational requirements to get
the optimum is going to be rarely, if ever, justified by the use of this method. But
there is good news yet, the method of Silver and Meal follows next.
24 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
TEC
Error in TEC"
TEC*
CIQ/2
SR/Q
Q1
Q*
Q
Q2
Error in O"
Figure 10
Now, I do believe that the minimum cost for inventory situations occurs where
the ordering costs and the holding costs balance. I further believe that as we get further and further out in a forecast the expected error of the forecast is roughly equal
to the time period squared multiplied by the size of the forecasted order in that period. In equation form we could write:
Expected Error of the Forecast = T2*D(T)
I believe this because my experience tells me that errors in forecasts go up roughly as the square of the time periods from now. And it is certainly true that the bigger the order in the future, the greater the probability that it will CHANGE!
25 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Lets take a look at the measure of the above equation; assume that the time period is months and the demand is in tons/month. If this is the case the measure is:
Expected Error + Months2*Tons/Month = Months*Tons
It only remains to find some combination of the cost of the item (C), the holding
rate (I), and the ordering cost (S), that has the same measure of months*tons. We
dont have to look far to find:
S/CI = Months-Tons
We therefore conclude that a rough and ready approximation to when we should
order is given by:
S
T2*D(T) > CI
In short, the above rule says:
Launch the order when the time period, squared, times the demand in that
period becomes greater than the ratio of the ordering cost to the holding cost.
Now your first reaction should be that no way in Hades could lot sizing be that simple and cover your backside. Lets test it on our well-beloved problem and see. Recall
that the first five months of forecasted demand look like:
Month
1
2
3
4
5
Demand
10 10 15 20 70
We first try T = 1, with a holding cost of $2/item/period and an ordering cost of
$300, which gives:
T2*D(T) > S/CI, or,
12*10 > 300/2
Clearly, the answer is no, so lets try T=2, which gives:
22*10 > 300/2
Again the answer is no, so lets try T=3, which gives:
32*15> 300/2
Once more, the answer is no, so try T=4, which gives:
42*10 > 300/2
The answer is yes! This tells us that we should order four periods demand or 55
items to come in at the start of the first period. This is also the first answer found
by dynamic programming above.
The method above is really rough and ready, and should not be used for anything
other than a ball-park estimate of where the lot sizes are. Lets see if we cant define a
more representative model than the (admittedly) quick and dirty one above.
26 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
27 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
j=T
))
The stopping rule is simple. As soon as TEC(T+1) > TEC(T), STOP! However, we
dont want to have to calculate however many steps it may take to find the optimum,
so lets design a nomogram or a fill-in-the-blank form to do it for us. To do this, we
must first modify the above general formula by dividing both sides through by CI;
this gives:
TEC(T) = 1 * S + j=T
TEC(T)
(j - 1)*D(j)
CI
j=1
T CI
( (
))
I hope that the reader will agree with me that dividing both sides of the above
equation by a constant shouldnt change the stopping point. It will, however, change
the values of the total costs at various times by the factor of CI. Lets test the above
problem again to convince ourselves that even though the costs change, we stop at
the same place.
If we order for one month, and we assume that we count our inventory at the end
of the month, clearly our cost per unit time is:
TEC(T=1)= 1 *(S/CI)= 1 *($300/2)=$150/month
T
1
If we order for two months on the same basis, our cost per unit time is:
TEC(T=2)= 1 *(S/CI+(D(T))= 1 *($300/2+10)=$80/month
T
2
If we order for three months, as before, our cost per unit time is:
TEC(T=3)= 1 *(S/CI+(D(T)+2*(D(T)+1))
T
= 1 *($300/2+(10+2*15))=$63.33/month
3
Now ordering for four months, our cost per unit time is:
TEC(T=4)= 1 *(S/CI+(D(T)+2*(D(T)+1)+3*(D(T+2))
T
= 1 *($300/2+(10+2*15+3*20))=$62.50/month
4
1
TEC(T=5)= T *(S/CI+(D(T)+2*(D(T)+1)+3*(D(T+2))+4*D(T+3))
1
= 5 *($300/2+(10+2*15+3*20+4*70))=$62.50/month
We therefore see that our stopping point is the same as before. To minimize cost
per unit time, our first replenishment should be for 55 units, just as before. Lets
now use this modified method to design the promised fill-in-the-blank form.
28 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Reference
Hesse, Rick and Woolsey, R.E.D., Applied Management Science, Science Research
Associates, Inc., Chicago; Palo Alto, Calif., 1980, pp. 63-65.
29 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
0
10
0
150
1
150
1
10
10
160
2
80
2
15
30
190
3
63.33
3
20
60
250
4
62.5
4
70
280
530
5
106
5
180
Its easy to see that the values generated coincide with those from the modified
formula above.
The next replenishment would look like:
A
B
C
D
E
F
0
70
0
150
1
150
1
180
180
330
2
155
2
250
This tells us we should order 70 as the second replenishment, if the forecast doesnt change! Continuing on with the form we would generate the orders as shown
below with associated costs.
Period
Demand
Order
Start
End
Averg.
1
10
55
55
45
50
2
10
3
15
4
20
45
35
40
35 20
20 0
27.5 10
5
70
70
70
0
35
6
180
180
180
0
90
7
250
250
250
0
125
8
270
270
270
0
135
9
230
280
280
50
165
10
40
11
0
12
10
50
10
30
10
10
10
10
0
5
Chapter 2
1
35
2
45
3
67
0
35
0
187
1
187
1
45
45
232
2
116
2
67
134
366
3
122
3
77
4
120
This tells us that our first replenishment is to order 80 units to come in at the start
of the first period. At this point a grizzled old foreman looks over your shoulder and
says, You know if your demand in March was 58, youd get a different answer. Is
he right? Lets use another Q&D and see.
A
B
C
D
E
F
0
35
0
187
1
187
1
45
45
232
2
116
2
58
116
348
3
116
3
77
231
579
4
144.7
4
120
Yes indeed; the foreman saw the trigger point. Now lets discuss how he did that.
Remember if the entry in the F row ever goes up, WE STOP, and order up TO that
point. Lets look at the original problem again:
A
B
C
D
E
F
0
35
0
187
1
187
1
45
45
232
2
116
2
67
134
366
3
122
3
77
4
120
31 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
Look at the second entry in the A row (=1). Now divide that entry into the first
entry in the F row (=187). This gives 187/1=187. This is telling you that you will
not order for the first period unless the demand in the second month is greater than
187. Say the February demand is 188, this gives:
A
B
C
D
E
F
0
35
0
187
1
187
1
188
188
375
2
187.5
2
67
3
77
4
120
As the cost goes up, but not by much, we would order for the first period only.
Following exactly the same logic and looking at the original problem again we
have:
A
B
C
D
E
F
0
35
0
187
1
187
1
45
45
232
2
116
2
67
134
366
3
122
3
77
4
120
We now do what the foreman did, and divide the third entry in the A row (=2),
into the second entry in the F row (=116). This gives 58, so he knows that you
wouldnt order for just two months if the demand in March was 58 or less. The whole
idea of this method is, once you have filled in the forecast, the bumps in demand
should pretty well tell you where the break points are. As there may actually be times
when the big dogs need to know where the ball park is, we now display another
Silver and Meal Form that only requires the use of a straightedge. This is called the
Silver and Meal Nomograph and appears in the following section.
32 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
33 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
800
700
I)
(C
CO
LD
HO
70
600
200
50
.5
500
40
(T
IN
30
AN
ST
EM
400
PERIODS
2
1.
20
ORDERING COSTS
ST
300
CA
300
10
400
RE
500
FO
200
600
10
0
T2 D(T)
800
Figure 11
34 Inventory Control (For People Who Really Have to Do It) eBooks Series
700
20
S /CI
100
2
1
Chapter 2
35 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 2
References
1. Harris, F. W., How Much Stock To Keep On Hand, Factory, The Magazine of Management,
Vol. 10, No. 3, pp. 240-241, 281-284.
2. Kaimann, R.A., EOQ vs. Dynamic Programming Which One To Use For Inventory
Ordering? Production & Inventory Management, Fourth Quarter, 1969, pp. 66-74.
3. Kaimann, R. A., A Comparison of the EOQ and Dynamic Programming Inventory Models
With Safety Stock Considerations, Production & Inventory Management, Third Quarter, 1972,
pp. 72-91.
4. Silver, E.A., and Meal, H.C., A Simple Modification of the EOQ for the Case of A Varying
Demand Rate, Production & Inventory Management, Vol.10, No. 4, Fourth Quarter, 1969, pp.
52-65.
5. Woolsey, R.E.D., Ordering Inventory When the Forecast Is Ridiculous, Production &
Inventory Management, Vol. 27, No. 1, First Quarter, 1986, pp. 144-148.
6. Woolsey, R.E.D., A Nomograph For Ordering Inventory When The Forecast Is Ridiculous,
Production & Inventory Management, Vol. 27, No. 4, Fourth Quarter, 1986, pp. 128-133.
7. Woolsey, R.E.D., A Requiem For The EOQ, Production & Inventory Management, Vol. 28,
No. 3, Third Quarter, 1988, pp. 64-66.
36 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3 (Political)
June 7, 1996 Revision
37 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
PRICE
$4.44
DEMAND
287
Clearly, the item was ordered in lots of 28 whenever the on-hand fell to 14.
The ADP whiz was somewhat unnerved when the prof asked him to find the exact
statement in the program where the EOQ was calculated. A search of the program
(which happened to have been written in FORTRAN) revealed the following statement:
C SET THE ECONOMIC ORDER QUANTITY
102 EOQ= SQRT(2*D*R/C)
Here, (clearly) D is the annual demand in units.
C is the price of unit in dollars.
R is the ratio of ordering cost to holding cost.
As the demand and price are part dependent, we would really like to know what
the devil R is.
A search earlier in the program turned up the statements:
C SET RATIO OF ORDERING COST TO HOLDING COST
18 R = 12
But, if the program thinks that R = 12, Then this implies
R = (Ordering Cost)/(Holding Cost) = 12, which seems to imply that
Ordering Cost = 12 * Holding Cost.
At this point the prof asked what the analyst thought the holding rate really was
and received a most surprising answer. The civil servant said that, as it was government money, it really didnt cost anything to hold something in inventory. The prof
made a note of this and silkily asked him if he might bring over a friend of his from
the local newspaper that was giving them so much trouble (the city editor) so that
the civil servant could repeat that statement for the press. At this point, this persons
boss took this young man outside and screamed at him for a time. One supposes
that the boss explained to this budding civil servant the results of a headline that
said:
LOCAL RTD OFFICIAL SAYS TAXPAYERS MONEY REALLY FREE
This would no doubt be followed by an announcement that this particular civil
38 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
servant had just been sent to Great Sand Dunes National Monument to take inventory of the number of flies found on state issue flypaper, until a more worthy assignment appeared. After this explanation, the offending civil servant, white-faced from
the previous discussion, asked humbly if the prof had any ideas for setting a holding rate.
The professor pointed out that as the RTD issued bonds, one had only to look at
the bond rating and a number could quickly be attached to the cost of money for
the RTD. This was done and found to be 11.4 percent. Plugging this interesting bit
of information into our formula above, we get:
Ordering Cost = 12 * Holding Rate = 12 * 0.114 = $1.37
The prof then asked the assembled multitude if they believed that they could:
Note that they were out of stock.
Do the paperwork for purchasing to launch the order.
Launch the order.
Receive the order.
Restock and update the record.
Pay the vendor.
And do all this for a grand total of $1.37.
At this point, the head warehouseman said that no way could it be done for less
than $20. But, if he thinks that then this implies:
Holding Rate = (Ordering Cost)/12 = 20/12
This implies a holding rate of 166 percent a year. The assembled multitude roundly declared that both the ordering cost of $1.37 and the holding rate of 166 percent
per year were manifestly absurd. We now have a dilemma; if an ordering cost of
$1.37 is absurd or the holding rate of 166 percent a year is absurd, then perhaps we
should have some doubts about R = 12. We have now been told that the ordering cost
must be at least $20, and the holding rate is 11.4 percent. Given these two bits of
information, we are forced to conclude that the FORTRAN statement that reads:
R= 12
Should perhaps really read:
R= 20/(0.114) = 175.43
Could it just be that this change will make a slight difference in our EOQ calculations? Could it also be that a consultant could make a profit here by offering to be
paid a (small) percentage of the savings?
The Long-Awaited Recalculation and What Happened Then
Lets use the above knowledge to save some money. We first find a typical item in
the spare parts list like the one we have seen before, namely:
PART # EOQ/ORDER POINT PRICE DEMAND OVER LAST # YEARS
CT1D4640
28/14
$4.44
287
3
It seems to say that a demand of 287 units was satisfied over the past three years
39 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
by ordering in lots of 28 whenever an order point of 14 was reached. Now find the
average annual cost of inventory for this item, using the present EOQ and the newly
extracted ordering cost and holding rate, using the tried and true formula for total
expected cost:
TEC = CIQ/2 + SD/Q,
We then plug in the price, the new holding rate, the new ordering cost, the old
annual demand, and the old optimal order quantity, which gives:
TEC = (4.44)(.114)(28)/2 + (20)(287/3)/28, or
TEC = Holding Cost(= $7.09) + Ordering Cost(= $68.33) = $75.42/year
Recall one of the many amusing assumptions made by the economic order quantity model. At the optimal order quantity, the annual holding cost should equal the
annual ordering cost. Looking at the difference between $7.09 and $68.33, we immediately realize that something is seriously amiss. The reason why this result is so
skewed is simple; the EOQ and order point that they have been using come from cost
data that probably hasnt been updated since the system was installed. Using the wellbeloved formula for the optimal EOQ,
EOQ =
2*S*D
C*I
2(20)(287/3)
(4.44)(.114) = 87 units
40 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
41 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
get a piece of the action, can at least get a promotion and/or a medal out of it. In
conclusion, for those of you in both the public and private sectors I can only say,
good luck and good hunting.
References
1. Woolsey, R.E.D., and Swanson, Huntington S., Operations Research For Immediate
Application, A Quick & Dirty Manual, Harper & Row, New York, (1975), pp. 39-41.
and Inventory Management Journal, Vol. 29, No. 4, Fourth Qtr. 1988, pp. 64-66.
42 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
Step 2.
Step 3.
Use trapping method to find out holding rate and setup or ordering cost.
Step 4.
Step 5.
Step 6.
Step 7.
Step 8.
Step 9.
Obtain (contractually) fee of 1 percent of above difference before telling them magnitude of saving.
Step 10.
Step 11.
Step 12.
43 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 3
PRICE
$35
DEMAND
1000
OVER # YEARS
1
EOQ/OP
47/15
If the holding rate is 1.5 percent a month and the ordering cost is $25, what is the
present total expected cost?
TEC =
What is the correct EOQ? _________
What is the total expected cost with the correct EOQ? _________
TEC =
HOW MUCH DO WE SAVE? ________
A quick check of the lead time shows that it has changed to 18 days.
The new order point is ____________
PRICE
$35
DEMAND
4000
OVER # YEARS
8
EOQ/OP
32/18
If the holding rate is 1.5 percent a month and the ordering cost is $80, what is the
present total expected cost?
TEC =
What is the correct EOQ? _________
What is the total expected cost with the correct EOQ? _________
TEC =
HOW MUCH DO WE SAVE? ________
A quick check of the lead time shows that it has changed to 10 days.
The new order point is ____________
44 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 4 (POLITICAL)
June 7, 1996 Revision
45 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 4
low and then the fun began. I went over to a nearby bin and picked up a tungsten
carbide drill bit worth about $800. Now it is important to recall that this kind of
inventory is called a consumable! I showed it to him and, facing the wall of Kardex
cards, asked politely: Donde est la Carte? (Where is the card?) It didnt take long
before the plant manger and I realized that he couldnt find it! I put the drill bit back
and chose a Kardex card for a tool at random from the wall of files, showed it to him
and asked: Donde est el implemento? (Where is the tool?) It took an even shorter time to realize that the supervisor hadnt a clue, even though the Kardex card had
an I.D. number on it in red two inches high.
I then noticed that on the wall next to the half-door was a nail; on the nail was a
pad of pull tickets for checking out material from the crib. On top of the pad was a
pencil, on top of the pencil in geologic depth was dust! Any fool can now plainly see
that the supervisor doesnt use the system.
Pointing out the unused pad, I gently inquired of the supervisor as to why he didnt use the system? Sweating visibly, he told me that:
a) He only had this job because he had hurt his back, and
b) was going to retire in six months. He told us the following story.
Venustusiano Oso
On the first morning he had to run the crib by himself, he had been told absolutely that no tools were to be issued without a pull ticket. Unfortunately, the first guy to
show up was Venustusiano Oso (The Bear!). Now this hombre stood some six feet,
six inches in his stocking feet (not that he wore socks) and weighed in the vicinity
of your average sumo wrestler. Venustusiano stood on his side of the half-door
holding a broken drill bit shouting: OTRA! (Another!) The supervisor timidly
pointed to the pad and said softly: Llena el formulario, por favor. (Fill out the
form, please.) The only result was that Venustusiano thrust himself through the
half-door, grabbed the supervisor by the front of his shirt and said (with an increase
in volume): OTRA!!
Needless to say, Venustusiano shortly departed clutching his drill bit, and a badly
shaken supervisor was starting to enter the needed data on a pull ticket.
Unfortunately, another sumo wrestler was standing at the half-door wanting another part. This person made it abundantly clear that if he didnt get the same quick
service (without filling out the $#@*& form) that grave bodily harm would shortly
take place relative to the supervisor.
We can all see what the inevitable result is here. No forms were ever filled out
which meant that the database for the million-dollar Big Blue tool control system
was unmitigated nonsense. The supervisor clearly entered in whatever he either had
time for or could remember.
46 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 4
47 Inventory Control (For People Who Really Have to Do It) eBooks Series
Chapter 4
ly aware of a Colt .44 making an indentation into his overflowing gut. The super
purposefully cocked the single action piece and (politely) repeated: Llena el formulario, por favor. Venustusianos eyes glazed over, and he was heard to say
(respectfully): Porque no? Donde est el?
Needless to say, the form was filled out, Venustusiano got his drill bit, and everybody else filled out their forms. And, indeed the supervisor collected his $18,000
and retired to a better life somewhere. It was rumored that the high point of his
retirement dinner was when he gave his gun to his successor!
Mas Preguntas de Importancia
Now what is the meaning of this morality tale? In Spanish, we could certainly say
that: Por razn si possible, por fuerza si necessario. However let us end by asking
again, two questions.
Pregunta Primero:
What was the relationship of the supervisor to his
company, before I came, in one word starting with
E?
Answer:
He was an employee!
Pregunta Secundo:
What was the relationship of the supervisor to his
company, after I left, in one word starting with O?
Answer:
He was an owner!
Question:
Esta Differencia?
Answer:
S!
Question:
Que es el nombre de la differencia?
Answer:
La nombre es productividad!
Pregunta:
Comprende?
Answer:
S?
Todos los hombres:
Viva Mexico!
El Zopilote Sabio, Roberto Eugenio Donaldo Woolsey Blanco
Reference
Simpson, Lesley Byrd, Many Mexicos, University of California Press, Berkeley & Los Angeles,
1967.
48 Inventory Control (For People Who Really Have to Do It) eBooks Series
Biographies
49 Inventory Control (For People Who Really Have to Do It) eBooks Series