1

SERIES 03

TITLE I. FORECAST MANAGEMENT 1

I. THE BUDGETARY PROCEDURE

A. HISTORICAL ORIGINS

The term "budget" comes from an old French word, "bougette" or "little bag". The bag (from Latin
"bulga" designated in the Middle Ages the chest or bag in which the traveler carried his belongings.
staff (today we use the word "luggage"...), or also the bag in which it would fit.
his coins.

A long journey is not improvised; it requires planning, scheduling the stops at the inn, and having the means.
Necessary sum. The budget is there for that. What better symbol to find for management?
forecast?

The term was then adopted by the English at the end of the Hundred Years' War, at a time when
parliamentary institutions were being born. The "budget" becomes a notion of public law, for
to designate the sum of money allocated by a vote of Parliament to an administrative entity for its
functioning: each minister thus has a small abstract booth (an 'envelope'
budgetary") from which it can draw (the term "ministerial portfolio" is also used, which refers to
to the same image).

Note that this budget allocation mechanism allows the state to solve a delicate problem of
Management: how to decentralize while maintaining control. The budget vote allows for resolution.
"agency relationship" between the State and the various expenditure controllers, to the extent that a
Expenditure can only be incurred if it has been planned within the framework of a budget line.
e
century, we are witnessing the development of large companies that, due to their size,
At the beginning of the 20th
face identical problems of decentralization, coordination, and control. These major
companies will naturally transpose the budgeting procedure for their internal needs of
Management. The budget becomes an essential instrument of emerging management control.

Indeed, how can a large company maintain control over its operations and
development of a large number of establishments, services, subsidiaries? Simply by
centralizing strategic thinking, decentralizing operational management and controlling the
system through a budget allocation procedure for resources, particularly financial.

It can therefore be said that companies have "copied" the administrations and the functioning of the
public accounting, which deserves to be emphasized and contemplated, at a time when the dominant discourse
consists of criticizing "bureaucracy" and demonstrating the efficiency of private business management in
example for administrations!
 2

B. OBJECTIVES OF BUDGET MANAGEMENT

The budget: a management tool

The budget appears as one of the essential parts of the 'cybernetic' regulation mechanism.
mentioned in series 01.

Managers are always in a situation of imperfect information, making it impossible to make

constantly the optimal decisions and one should expect that the actual results will diverge from
objectives. Therefore, it is necessary to establish a feedback control procedure for the discrepancies between
forecasts and achievements, in order to avoid 'slippages'.

In this context, the budget plays a central role as it clarifies all forecasts and
objectives serving as a 'compass' for daily action. The budget control mechanism (see
series 07) obviously assumes the establishment of budget forecasts.

2. The budget: a simulation tool

Very often, before finalizing the budget, we "test" several hypotheses, by measuring
the impact that such or such a decision would have on the forecast result. The budget then plays the role of a
decision-making aid tool. This simulation function explains the great usefulness of the tools
Spreadsheet-type software: you just need to change a data point (an activity level or a price by
for immediately obtaining a new "version" of the budget forecasts
corresponding.

Let's note, for these projection or estimation exercises, the importance of analytical accounting in
partial costs of the 'direct costing' type: it is the knowledge of the structure of charges, fixed and
variables, which allows for predictive calculations.

The budget: a tool for motivation and conflict prevention

The company often constitutes a conflictual environment in which oppositions manifest.
interests among the different stakeholders: between management and unions, between different departments,
etc.

In this perspective, the budgeting process often allows for arriving through negotiation at a
compromise presenting itself as a sort of contract between the various stakeholders. We touch
here to management methods and human resources management. More the procedures
budgets are decentralized and participatory, the more the budget appears as a means of clarifying
this compromise (the establishment of the budget, for example, requires making an assumption about
the increase in the hourly wage). The budget sets the objectives.

C. THE PLAN - PROGRAMS - BUDGETS ARTICULATION

Budget management illustrates the design of management control considered as a
means of declining the strategy.
 3

The plans
The major strategic orientations are explained in the form of multi-year plans (over 3, 4, or 5 years).
setting projections of turnover and results and planning the main operations
of investment and financing.

These plans provide the overall framework for development and are operationalized annually under
the form of programs and budgets.

2. The programs
A program is a short-term activity forecast specifying in detail the levels.
of activity, of production, the volumes of factors used (hours of labor, tons of materials
first, etc.). For example, for a workshop, we establish a monthly production schedule, which
allows for planning the workload of the different machines.

A program is therefore generally expressed in physical units, that is to say in number of

products, in tons, in hours, etc.

3. The budgets
Budgets represent the 'estimation' in monetary units of the implementation of a program.
To move from the program to the budget, one must make economic assumptions about the evolution.
prices and costs (commodity prices, unit variable costs and fixed costs provided by
analytical accounting, etc.).

It is usually distinguished:

a. The "functional" or "operational" budgets, which reflect the programs related to the
different services or departments:
sales budget
production budget
– purchasing budget,
budget for administrative services,

financial or summary budgets that summarize the impact of the implementation of

functional budgets: the cash budget essentially, and, by analogy, the documents
summary forecasts: projected income statement and projected balance sheet.

It should be noted that a classic exercise in budget management consists of preparing these various documents.
forecasts (programs, budgets, projected accounts), usually starting with the
sales forecasts (see for example the corrected case DISTROC further on).
 4

D. THE BUDGETARY CYCLE

Specifically, the budgetary procedure is a cyclical activity that annually regulates the life of
the company.

The classic scenario for a company aligning its accounting and budgeting exercise with the year
civil and practicing a reasonable decentralization of responsibilities, is as follows:

In September, the general management publishes a general orientation note (often also called)
"framing letter", or "pre-budget"), addressed to the main responsible parties, setting the objectives, the
major directives for the coming year, in line with the content of the strategic plan, but taking into account
account for the necessary adjustments due to the evolution of the environment (a segment
investment can be deferred, for example).

In October, each 'component' (each department, or each major directorate, or, in another
level within a group, each subsidiary establishes its own budget in a decentralized manner. Indeed, this
It is the operators who are best positioned to assess the constraints.
This must be put into perspective. For example, the more vertically integrated the activity, the more the units located
Upstream needs the activity forecasts of the units located downstream to establish their own.
forecasts. In this case, we may have an interest in centralizing the procedure, for example by using some
matrix planning methods (see series 05).
These different "budget sketches" are gathered by the management controller to be
harmonized.

In November, this harmonization is gradually realized through an iterative process of

"budget" shuttle. Budget projects indeed shuttle between decentralized units and
the coordination cell led by the financial management or the management control (the expression
"shuttle" is used by analogy with the shuttle of parliamentary bills between the House.
of deputies and the Senate).
This procedure helps to reduce inconsistencies, find compromises, and achieve
arbitrations, particularly taking into account financial constraints.
It is important to understand that the budget not only outlines forecasts but also objectives.
negotiated.
It should be noted that at this level, it is the centrally established cash budget that allows for...
synthesis and to see if "it works or if it fails."

In December, the final budget is decided, with the last adjustments being the responsibility of management.
general. In large companies, this final budget can be a substantial document of
several hundred pages, but whose summary is provided by the projected income statement.
Note that the different budgets, which then form the budget bundle (by analogy with the bundle
fiscal), are obviously divided into monthly payments, to allow for exercising later, during the fiscal year, the
budget control.

During the exercise, the budget forecasts will be compared monthly with the actual results.
effectives, provided by the accounting, in budget control statements allowing for calculation
discrepancies involving responsibilities in case of slippage: this is the 'budget monitoring'.
 5

E. THE CONCEPT OF 'FLEXIBLE' BUDGET

A budget, as a projected estimate, is necessarily dependent on certain variables, and in

specific to the level of activity assumed as a hypothesis.

For example, the operating budget B of a workshop for which the unit of work is the hour -
machine can be put in the form:
B = vN + F

Expression in which:
N = number of hours
v = variable cost of the unit of work
F = fixed costs
The function B = f(N) can be represented graphically as follows:

v v corresponds to the slope

from the budget line

One can make an infinite number of hypotheses about the projected value of N and thus establish an infinity of
budgets.

Generally, in practice, we reason within a "budget range" by presenting the

calculations for at least 3 hypotheses:
most probable average hypothesis (N)m
– 'low' or 'pessimistic' hypothesis (N) b
high or optimistic hypothesis (N) h
 6

Hence the following diagram:

Bh
fork Bm
BUDGETARY Bb

Nb NmNh

Budget calculations can be presented in a table in the form of a 'flexible' budget.

that is to say established within the framework of the various hypotheses considered.

Numerical example: with F = 10,000 and v = 5, and a level of activity of 6,000 hours but being able to
situated in a range of 5,000 - 7,000 hours:

Projected activity level (in hours) 5,000 6,000 7,000

Projected fixed costs 10,000 10,000 10,000
Forecasted variable costs 25,000 30,000 35,000
Total budget 35,000 40,000 45,000

Let us note that the function B = f(N) can obviously be more complicated than a simple linear function.
take here as an example. There can notably be threshold effects (overtime at-
beyond a certain volume of activity for example.

F. THE BUDGETS OF 'GENERAL' SERVICES

Many services within the company, which we refer to here as general services, pose some
budgeting problems very different from what happens for the 'operational' services of which
the activity is directly related to the purchase-production-sale cycle.

In a production workshop, for example, there is a causal relationship between the volume of products.
manufactured and the quantities of raw materials or the number of hours of work required for the
production. It is then possible to 'model' the budget based on variables such as the level
Activity: we have a rational basis to 'quantify' the estimated costs.

On the other hand, in an administrative service, it is much more difficult to measure production, and
to appreciate the amount of expenses necessary for proper functioning.
 7

The budgets for general services are likely to be set arbitrarily (we talk about costs
"discretionary", as they are at the discretion of the managers), and simply renewed
from one year to another.

Attempts at rationalization have nonetheless been introduced in order to optimize spending.

based on the use of techniques such as value analysis or the BBZ approach (Budget
zero base).

1. The services concerned

The services concerned are classified as 'administrative' and 'functional' services (as opposed to
operational), or even 'staff'. Some economists even talk about services
"unproductive"...

These are services that only indirectly contribute to industrial or commercial activity.
making this activity possible. From an accounting perspective, the corresponding expenses are the most
often grouped in the 'auxiliary' analytical centers or sections studied in series 01. We
also often talks about 'support' activities, in the perspective of PORTER and the approach
ABC.

These are essentially services such as:

the general management
the general secretariat,
the postal service,
the telephone switchboard
the accounting services,
financial services,
the IT service,
the documentation,
the legal and litigation service,
the personnel management,
the methods, planning, and scheduling services
the design offices,
– research and development (R & D) services,
the patent management service,
quality control
the marketing research service,
the communication and public relations services,
– and, of course, the management control service!

This list is certainly not exhaustive, and it is quite possible that we have forgotten some.
also debatable (R&D, for example...).

2. The management problems posed by discretionary costs

The expenses incurred in these services are generally classified as 'fixed indirect costs' in the sense of the
cost accounting.

These "overheads" are difficult to control, as it is hard to assess the activity and the
production of the concerned services. The tasks to be accomplished are diverse, heterogeneous, and not repetitive:
it is therefore impossible to specialize or 'Taylorize' work, to define standards or
standards like for operational tasks.
 8

Discretionary charges are largely made up of personnel remuneration.

often highly qualified (managers, IT professionals, specialists...) or by the cost of employment of
sophisticated equipment (computers, laboratory equipment...). The issues of prestige, power,
search for social status generally leads to an inflationary drift of budgets
concerned, a phenomenon particularly aggravated in large companies by trends towards the
bureaucratisation related to size. These budgets are also conducive to the development of a
"slack", from a precautionary mattress.

Conversely, during times of financial difficulty, it is tempting for a general management to

"cut" in these discretionary expenses, easily deemed unnecessary, in order to restore the
profitability, for example by 'downsizing' the staff of head offices or certain departments considered
so plentiful.

These are valuable skills that can be sacrificed in a strategically way.

suicidal in favor of short-term interests. Often, the goal is to 'reduce the workforce by 10%'
it makes one think of the bleeding practiced by Molière's doctors, with the effectiveness that we know, the patient
finding themselves even weaker!

3. The traditional budgeting technique

Traditional procedures for budgeting discretionary expenses do not provide a solution.
truly satisfying to the problems mentioned above.

Most often, the general management calculates a global envelope not to be exceeded for the whole.
from these overhead costs, allowing, after deducting this envelope from the projected margin,
to achieve an acceptable result.

In periods of expansion and good financial health, one simply continues the spending.
of the previous year, possibly increased by an arbitrary percentage, taking into account
for example, taking into account inflation. On the other hand, in times of difficulties, we implement a 'cut'.
budgetary.

Each department then presents its wishes, along with the trends towards the derails mentioned above. There is
there are always good reasons to justify hiring new employees, buying a
more powerful computer, etc. A meeting must then be called to carry out the allocation of the budget.
global. In general, the total demand exceeds the budget, and a bargaining process
is established. It is in this context that the budgetary 'back and forth' procedure studied in series 04.
allows to resolve conflicts and reach a consensus on the 'sharing of the cake'.

It is understood that such a procedure does not guarantee optimality, that the actual needs are not
necessarily satisfied, that the distribution of the budget mainly reflects the power relations between the
different leaders and leads to what is felt as injustices, that the result can
harbor grievances and create conflicts between individuals.

4. The BBZ procedure

The procedure of 'Zero Base Budgeting' gained significant popularity in various circles.
management, especially in large American companies (Texas Instruments, Xerox,
Westinghouse...) after a sensational article published in 1970 in the Harvard Business Review by
Peter PYHRR.
 9

In the BBZ process, we do not settle for incremental improvements; we adopt an attitude.
much more radical: nothing is ever guaranteed, the simple and straightforward renewal of a budget of a
year to year is excluded by principle.

Every budgetary expense must be justified, regardless of past practices, which do not
could serve as a priori justification. In short, we 'start over from scratch' each year, hence the
name of the method.

This first requires breaking down the company's activities into modules, which are
decision-making centers characterized by the existence of a clearly identified responsible person (principle
of uniqueness), and by a certain homogeneity of tasks (certain aspects of the method can be found here
ABC studied in series 02).

At the level of each module, the manager must establish budget proposals, in connection with
with its activity projects. It must clarify its purposes, its objectives, the necessary means for
to reach and establish a 'budget set'.

This budget package must contain:

on one hand, alternative projects (or 'mutually exclusive'), that is to say several solutions
possible budgeted alternatives to fulfill its mission (for example, for a service
computer science required to implement a new program, a project assuming a
internal development involving the hiring of a team of programmers, and a project
assuming the external acquisition of a software package);
On the other hand, complementary projects, that is to say projects that are not strictly necessary.
indispensable (the company can continue to operate without), but which may present a
interest (with the aim of improving quality or productivity, for example). The interest in this
The distinction is particularly to clearly indicate the projects that can be deferred in time, if
The financial means for next year are insufficient.

The different projects are then evaluated and prioritized by a committee.

This phase of prioritization raises the issue of the criteria used to classify, accept, or reject.
the various projects.

One can use classic financial criteria, such as profitability (as for the
investment choices, a question studied in the following part of this series), or methods of
Additive scores. Thus, Peter PYHRR advocates the '5P' system ('Peter Pyhrr's Pragmatic
Pointscore Procedure: each committee member 'votes' by scoring the projects from 1 to 6. We proceed with the
total points obtained (like in figure skating competitions...) and the decision depends
the total score obtained.

It is seen that the BBZ procedure, considered a cure-all by some, and a gimmick by others
others, may be long and costly.
 10

A streamlined version has been proposed, based on a "rotating examination" of the various services. The
Americans speak of the 'sunset procedure,' referring to the image of the
Earth's rotation: when the sun sets in one place, it is still shining elsewhere. Every year, we
"audited" through the BBZune procedure only part of the services. The managers of
these services must then comply with the critical review procedure of their expenditures, and justify the
level of their budget. But once this budget is accepted, these officials are 'at ease'
for example, 4 or 5 years: the order of magnitude of their budget will not be called into question during this time.
time lapses, except for extraordinary events. The following year, it is other colleagues who
"go through the grinder"...

This procedure is obviously less burdensome, less expensive, less 'stressful' for the various
frameworks.

It should be noted that a similar procedure exists in French universities, with their different programs.
are authorized and funded by the State for renewable periods of 4 years: the procedure can
so it can also be used in public management.

G. CONCLUSION

It is necessary to relativize the interest and usefulness of budget management.

First of all, it should be considered that a budget will only be a good management and motivation tool if the
the level of objectives to be achieved is set intelligently: it is necessary to "set the bar" high enough, but
not too much. Hofstede's famous work has shown that budgets that are not ambitious enough lead to
the leaders have given in to ease: the organization is not under pressure. Teams need to
had a "challenge" to meet. But beyond a certain threshold, one can on the contrary demobilize the
operational, who feel that too much is being asked of them. Since anyway we won't be able to
reaching the goal is not worth trying!

Things can be even more serious, and authors like ARGYRIS have highlighted the
devastating psychological effects of excessive budget 'pressure' on managers. The
concerned individuals may experience feelings of frustration, develop pathologies
psychosomatic (eczema...), show aggression, or fall into depression, or even
suicidal. Budget pressure, pushed to the extreme, and conceptualized by some leaders or oneself-
saying specialists in human resource management in the form of 'stress management'
can rightly be considered a form of psychological harassment, or as the version
modern exploitation (some radical analysts go as far as to speak of slavery…).

In the face of this possible 'violence' of the budget, individuals do not fail to resist by developing
defense strategies suited. The most common of these practices is to create 'slack
budgetary", a term used in reference to the theory of organizational slack developed by
CYERT and MARCH (in their famous work: A Behavioral Theory of the Firm). Leslack(en
the slack of a rope or cable; 'to slack off' means to ease one's effort
"mattress", the "room for maneuver", that one tries to negotiate at the time of the construction of
budgets. For example, we know that we can operate with 100, but we ask for 120, and we get
110 after arbitration. It is then very easy to keep this budget, and even being lax, and spending
109, we still have a favorable variance in the end! We understand that budget management can be
completely distorted, emptied of its substance, and ending up with the opposite result of what was intended
searched: instead of reducing costs, we are increasing them!

One can also see in the allocated budget a 'right to spend', which must be used, otherwise we risk
to have a smaller budget the following year. Everyone has heard jokes about the practices
 11

supposed soldiers who would drive the trucks in the barracks yard to use the
fuel budget...

The budget is often criticized (just like any procedure related to planning in
In general, the budget is merely a means to implement a plan) to induce behaviors
bureaucratic and conservative, inhibiting initiatives and possibilities for adaptation to
opportunities. It's planned in the plan and in the budget, so we execute, even if the action is no longer
relevant given the evolution of circumstances. However, in today's world the environment is
increasingly 'turbulent' and unpredictable.

One should not idealize the budget and attribute all virtues to it. Some go even further, by
purely and simply advocating for abandoning budget management, especially to replace it
with a more reactive management. Perhaps it is still going a bit too fast and throwing the baby out with the bathwater.
with the bath water...

H. SELF-CORRECTING EXERCISE: DISTROC CASE

Statement
Elements of response

2. Self-correcting exercise

The company DISTROC markets 3 products, P1, P2, and P3, in 3 geographic areas A, B, and C.
Each zone manager establishes their sales forecasts, specifying the bases considered for the
calculations. These forecasts are then centralized by the sales director who presents a budget
quarterly consolidated to the General Management.

The company is currently preparing its budget forecasts for 1erquarter N+1.

1requestion
For zone C, the sales of 1erThe following was the N quarter:
300,000 products p1,
- 270,000 P2 products,
– 420,000 P3 products.
Establish the quarterly sales program for area C, in number of products (you will use
the same type of table as for the other areas).

2andquestion
Based on the information provided by the area managers, establish the sales budget.
monthly for the 1sterquarter N+1 of the company.

3equestion
Is this budget necessarily the one that will ultimately be retained to establish the budget document?
Why?
 12

4equestion
For the continuation of the case, and regardless of the numerical results of the previous questions, we
suppose that the sales program presented to the general management retains the following assumptions
as being the most likely, for 1erquarter N+1 (in product units):
month
January February Mars
products
P1 480,000 420,000 520,000
P2 440,000 390,000 480,000
P3 690,000 630,000 750,000

Furthermore, for the establishment of all budget forecasts, we reason in a

"range" of plus or minus 10%:
- low or "pessimistic" hypothesis: -10%;
- high or 'optimistic' hypothesis: +10%.
Under these conditions, make a 'flexible' assessment of the projected interim expenses.
quarterly.

5equestion
Establish the flexible quarterly budget for distribution costs (monthly breakdown is not)
requested.

6equestion
DISTROC operates in markets subject to strong cyclical variations, both in terms of
volumes that on prices.
A statistical analysis of the sales history shows that the quarterly revenue follows a
normal law having:
for expected value: 12 million euros;
– for standard deviation: 2 million euros.
[Link] the probability of reaching the high estimate of the budget range.
b. Calculate the probability of being within the budget range (we will use the table provided in
the series 02 of the CPC.

[Link].

ANNEX 1

Sales program transmitted by the director of zone A (in units of products sold) for the
1erquarter N+1 :
month
January February Mars
products
P1 157,000 141,000 170,000
P2 140,000 128,000 156,000
P3 225,000 205,000 249,000
 13

drain the zone B

month
January February Mars
products
P1 212,000 180,000 230,000
P2 201,000 171,000 207,000
P3 302,000 270,000 332,000

ANNEX 2

Assumptions made to establish the forecasts for 1erquarter N+1.

a. The activity is seasonal and the following seasonal coefficients are retained:
January: 1
February: 0.9
- March: 1.1

b. In relation to current prices (end of December N) which are respectively 2, 4, and 3 euros per unit for
P1, P2, P3, and considering the competition, the following decisions have been made at the level of the
pricing policy :
– product P1: rise of 3% at the beginning of February
- product P2: decrease of 10% from the beginning of January
- product P3: increase of 5% in early February.

Sales are expected to grow by 5% in volume for P1 and P2 and by 8% for P3 compared to the
corresponding period in N.

ANNEX 3

Distribution charge structure

The variable costs, which are essentially logistical in nature, are estimated at 0.1 euro per product.
sold.

b. Moreover, the representatives receive a commission of 5% on the revenue.

c. The monthly fixed charges amount to 210,000 euros, as long as the monthly activity volume does not
does not exceed 1,600,000 products (all products combined).

Beyond the threshold of 1,600,000 products, it is necessary to resort to temporary staff to cope.
to handling problems during the shipping of products. Fixed costs then increase
in increments of 15,000 euros for every "batch" of 100,000 products. The temporary workers have
minimum one-month contracts, renewable.
 14

Elements of response

1reProgram of zone C

Monthly average January February Mars

Product Total trimester
N (1) (0,9) (1,1)

300,000
P1 (1.05) = 100,000 105,000 (a) 94,500 115 500 315,000
3

270,000
P2 (1.05) = 90,000 94,500 85,050 103,950 283,500
3

420,000
P3 (1.08) = 140,000 151,200 136080 166 320 453 600
3
(a) For each box: monthly average× volume index× seasonal coefficient.
300,000
For example, for P1 sales in January: 105,000 = × 1,05× 1
3

2eConsolidated sales budget

Month
January February Mars Total
Products
P1 948,000 855 930 (1) 1,061,930 2,865,860
P2 1,567,800 1,382,580 1,681,020 4,631,400
P3 2,034,600 1,924,909 2,354,058 6,313,560
Total 4,550,400 4,163,412 5,097,008 13 810 820
(1) For each box:
(Σ forecasts by zone× price× price variation index. For example, for P2 in February:
415,500
415 500× 2× 1.03 = 855,930

ATTENTION
For P1 and P3, the price increase only occurs whenin February.

3eFinal budget?
No. The budget for question 2 represents the consolidation, the mechanical aggregation of forecasts.
established in a decentralized manner. Now, generally, a process ofshuttleis established between the
management and peripheral units, insofar as the budget does not solely represent a
forecast determined by a number of factors, but also aobjectivenegotiated.

4eFlexible forecast of interim spending

[Link] number of monthly products for the average hypothesis:
January: 480,000 + 440,000 + 690,000 = 1,610,000
February: 1,440,000
– mars : 1 750 000
 15

[Link] of excess "slices"

Month
January February Mars Quarterly total
Hypothesis
1,610,000× 0.9 = 1,440,000× 0.9 = 1,750,000× 0.9 =
Low 1,449,000 < 1,600,000 1,296,000 1,575,000 0
→0 →0 →0
1,610,000 is greater than 1,600,000 1,440,000 < 1,600,000
Average 1,750,000→ 2 3
→1 →0
1,610,000× 1,1 = 1,440,000× 1,1 = 1,750,000× 1,1
high 1,771,000 is greater than 1,600,000 1,584,000 6
→2 →0 = 1,925,000→ 4

c. Projected interim expenses

Hypothesis Number of slices Forecasted cost

Bass 0 0
Average 3 3× 15,000 = 45,000
High 6 6× 15,000 = 90,000

5e Flexible budget for distribution costs

[Link] of commissions for the average hypothesis:

P1: 3% early February:
[480,000× 2) + [(420,000 + 520,000)× 2× 1.03× 0.05 =144 820
P2: minus 10%from January
(440,000 + 390,000 + 480,000)× 4× 0.9× 0.05 =235,800
P3: 5% beginning of February. Same calculation as for P1, that is320 850
Let it be in total:
144 820 + 235 800 + 320 850 = 701 470

[Link] quarterly budget

Element Low hypothesis (–10 %) Average hypothesis High hypothesis (+10 %)

Logistics 4,800,000× 0.9× 0,1 4,800,000× 0.1 4,800,000× 1,1× 0,1
432,000 = 480,000 = 528,000
Commissions 701 470× 0.9
(5b) 701 470 701 470× 1,1 = 771 617
= 631 323
Fixed charges 210,000× 3 = 630,000 630,000 630,000
Interim (4c) 0 45,000 90,000
Total 1,693,323 1,856,470 2,019,617

Global volume: 1,610,000 + 1,440,000 + 1,750,000 = 4,800,000

 16

6eAssessment of probabilities
[Link] of reaching the top of the range:
Regarding the numerical results of question 5, the upper end of the range corresponds to
701470
× 1.1 = 15,432,340 euros
0.05
We have {
Probv≥ 15 432 340 }
⎧ 15,432,340− 12,000,000 ⎫
= Prob t≥⎨ ⎬
⎩ 2,000,000 ⎭
Prob≥ 1,72)= 1− prob(t< 1,72)= 1− 0.957= 0.043

There is only a 4 in 100 chance of reaching the top of the range, which is very low: this hypothesis
is unrealistic.

[Link] of being within the range:

15 432 340
The bottom of the fork corresponds to a revenue of × 0.9= 12,626,460
1,1
The sought probability is equal to:
{ } { } {
Prob 12 626 460≤ v≤ 15 432 340= Probv≤ 15 432 340− Pr ob v≤ 12,626,460 }
Prob(t≤1.72) – Prob(t≤0.31) = 0.957 – 0.621 = 0.336
There is only a 33 in 100 chance of being within the range, which is, again, very low.

[Link]
The previous calculations show that the budget forecasts are far too optimistic: we
high risk of not even reaching the bottom of the range!

II. LINK WITH THE CONCEPT OF CENTER OF RESPONSIBILITY

The budgetary procedure allows general management to decentralize management, to delegate
responsibilities to middle managers who therefore enjoy a certain autonomy, but who in
counterparts must report to their hierarchy on their actions (this is what is referred to as a
In general, 'reporting', from the English 'to report', which means 'to bring back', 'to make a report',
"reporting"). The company is therefore divided into "centers of responsibility," led by
people who, above all, are responsible for the budget allocated to them. The study of management
Budgeting is therefore very often practically inseparable from the study of these centers of responsibility.

A. THE DIFFERENT TYPES OF CENTERS

The division of the company into autonomous entities known as responsibility centers depends
much of the size and type of organization.

In many SMEs with a 'classic' Fayolian hierarchical and functional organization,

the breakdown will be done by function (technical, commercial, administrative). In a larger company.
Important, we will often have a divisional organization by type of products. In a large
In the group, a more complex matrix structure will be observed with operational units that can
 17

depend for example on a geographical direction and a technological direction (the subsidiary
selling televisions in France depending on both European Marketing and Management
Consumer Electronics.

One can define several types of centers of responsibility, depending on their degree of autonomy and
of the scope of the assigned missions.

1. Cost centers
The manager of a cost center is assessed on the quality of services, adherence to deadlines and
especially regarding the level of its costs: we expect a factory manager or a workshop supervisor to limit
the waste, that it respects (and possibly improves) the standards, that it improves the
productivity. Its budget therefore focuses mainly on expenses, not on revenue. By
Simplification, we can say that its mission is to minimize costs. It should be said that it must rather do so.
optimize, taking into account the quality objectives for example.

In an operational cost center (a factory), budgeted costs generally depend on the level.
activity forecast (application of the concept of flexible budgeting), whereas in the functions
support, the costs are much more "discretionary" (they are at the discretion of management), everything
at least in the short term. It is easier to eliminate a position in an administrative department or to
reduce the budget for reception expenses rather than cut back on variable direct costs. In case of
financial difficulties, when it is necessary to save money, it is these discretionary budgets that are
touched first.

2. Product or revenue centers

In contrast, certain areas of responsibility are judged primarily on revenue.
achieved. For example, in a commercial network, the head of a regional agency commits to
a sales budget. In a bank or an insurance company, the account managers have
budgetary objectives regarding the placement of financial products (life insurance contracts,
SICAV, etc.)

Note that having a target expressed in terms of revenue can have counterproductive effects: for
To make revenue, the salesperson can lower prices by granting excessively high discounts, or
taking a financial risk for one's company (risk of payment default) by neglecting to
check the client's creditworthiness.

3. The profit centers

Profit centers help to limit the adverse effects mentioned above: the manager is
judged both on expenses and revenues, on the level of results or margin it generates. Our
A commercial agency, for example, can be judged based on the margin achieved (net sales decreased)
of commercial costs). The manager of a profit center must therefore not be a technician
only a good engineer mastering his process, or only a good salesperson mastering the
client relations), but a more generalist 'manager' capable of acting as an entrepreneur and
to integrate all aspects of management, with a view to maximizing results.

It should be noted that in the case of vertically integrated activities, profit center accounting poses the
delicate issue of internal transfer prices: the result of a center obviously depends on
upstream and downstream billing.
 18

4. The investment centers

These centers imply a delegation of authority that goes beyond simple operations (expenses and revenues),
and expanded to the issue of investment in fixed assets. In general, this formula is not found.
that in large companies and large groups, at the level of divisions and subsidiaries.
The objective here is to maximize the return on invested capital. The manager will therefore have to defend
to its hierarchy, during the budget preparation, the profitability of its projects
investment. It is within this framework that the selection criteria and techniques will be used.
investments studied in financial management. For example, the manager of a subsidiary, to justify
an investment project will make revenue and cash flow projections in order to
calculate an acceptable NPV or IRR according to the group's profitability standards. It is recalled that
The Net Present Value is the sum of the discounted forecasted cash flows; for a project
to be profitable, the NPV must be positive. The Internal Rate of Return is the value of the rate
of updating that cancels the NPV; for a project to be profitable, its IRR must be higher
at the discount rate characterizing the economic environment.

5. The 'nested' arrangement of responsibility centers and associated budgets

Depending on the importance of the company, there can be several levels of delegation, and thus
several levels to define the centers.

For example, there may be several divisions, several factories per division, and several workshops per
factory. Each workshop can serve as a center of responsibility, with its own manager and budget;
the budgets of the workshops can then be consolidated at the plant level, which also constitutes
a center of responsibility, etc.

We observe a sort of nesting, like Russian dolls, of centers and budgets.

At each level, the manager is both the controller and the controlled: the factory director, for example.
he controls the workshops, but must report to his superiors.

B. THE CONTROL OF RESPONSIBILITY CENTERS

Each center has objectives to achieve, which can be negotiated or assigned depending on whether the management is
more or less participative. The performances must be measurable (mainly thanks to)
in the accounting system, but not only: we can also use physical indicators).
The benefit of setting objectives in budgetary terms is that one can think globally, thanks to
to the possibility of aggregating the values of heterogeneous elements (one cannot aggregate quantities
heterogeneous physiques, 'adding cabbages and carrots', but we can make a sales budget
global by adding their values in euros). The head of a center has a commitment to results,
and not just means.

For such a results-based management system to be relevant, two must be implemented.

principes : le principe de responsabilité et le principe de contrôlabilité.

The principle of responsibility means that a leader must be accountable for all consequences.
decisions falling within his delegation of power. He must not be able to impose.
to other centers the costs induced by the mediocrity of its management. For example, if the production is
poor quality, the additional costs that arise at the After Sales Service must be charged
to the production manager, not to the distribution manager.
 19

Symmetrically, the principle of controllability means that a leader should only be responsible for
elements that he controls, on which he has decision-making power. It is not "fair" that the
the results of a profit center, for example, may be diminished by the allocation of certain charges
imposed arbitrarily by Management (for example, in the case of the distribution of headquarters costs). This
leads one to reason about partial costs rather than complete costs (the managers are judged
on a 'contribution' for example), or to neutralize certain uncontrollable elements, by
example using standard costs.

Periodically, the information measuring the actual performances is reported back to inform the
Hierarchy, according to the reporting principle. Achievements can be compared to forecasts.
budgetary, sound the alarm in case of drift, and take corrective measures. The more we are
at the bottom of the hierarchy, the reporting states are more based on physical data or on
the analysis of basic costs: to control a workshop, we will focus on the yields, on the
hourly usage costs of machines, etc. The higher you go up in the hierarchy, the more you use...
aggregated accounting and financial data: a subsidiary will report aggregates to its parent company
very global financiers, without going into the details of how the different services operate: figure
business, payroll, investments, results.

III. THE CASH BUDGET AND SUMMARY DOCUMENTS

PROVISIONAL
The establishment of the cash budget, which synthesizes the effects of the implementation of all the others
budgets, is the opportunity to test the realism of the various activity programs.

A. THE CASH RECEIPTS BUDGET

It generally presents the different months in columns and the different categories of income in rows.
An additional column is also planned to handle short-term receivables related at the end
of the time lag between certain operations (sales) and their cashing, due to the
payment deadlines, which will appear in the projected balance sheet.

For example, suppose that the company grants its clients a month of credit: there will be a delay.
of a month between the sale and the payment.

In January N+1, the company will collect the customer receivables listed on the balance sheet at the end of N. In February,
she will collect the sales of January, in March the sales of February, etc.

The sales of December N+1 will therefore be shifted as an element of the projected balance at the end of N+1, and
collected in January N+2.

Receipts directly related to operations (sales) are generally distinguished, which are subject to
a forecast based on sales budgets, and other receipts related to
specific financial operations (capital increases, borrowings, asset disposals...).
 20

Example of cash receipts budget for N+1:

BALANCE
January February December
Elements ... FORECAST
N+1 N+1 N+1 by the end of N+1

Cash inflows related to operations:

Cash sales VJ(1) VF(2) VD(3)
Credit sales CCDN(4) VACJ5 VACN(6) VACD(7)
– ... Mail
Clients
Other collections:
Asset disposals
– Loans
– ...
Total receipts:
with :
VJ: Cash sales of January N+1.
VF: Cash sales of February N+1.
Cash sales for December N+1.
CCDN: Customer receivables corresponding to credit sales of December N.
Credit sales of January N+1.
VACN: Credit sales of November N+1.
VACD: December N+1 credit sales recorded in the projected balance sheet.

NOTE
All these sales are obviously inclusive of all taxes (including VAT).

B. THE VAT BUDGET TO BE DISBURSED

Given the rules for VAT deductibility, it is better, in practice, to quantify it in a budget.
attach the VAT to be paid monthly, this item being then included in the budget of
disbursements.

Let us remember that the company is responsible for collecting VAT on behalf of the public treasury. It invoices to
its clients the VAT on the amount of its revenue. The client settles the amount of their invoice
The TTC and the company record the 'VAT collected on sales' as a credit to account 4457. However, on the other hand,
the company also pays its own suppliers including tax, and therefore settles, in addition to its purchases excluding tax.
tax, of VAT. This VAT is deductible and represents a claim on the State charged to the debit of
account 4456.

Periodically (in principle every month in the general scheme), the company fills out its declaration.
of revenue and calculates the amount of VAT that it must pay to the State by deducting the
Deductible VAT on purchases from VAT collected on sales. This "VAT to be paid" (account
4455) is paid to the tax authorities in the following month, hence a delay, the VAT for the month m
being in fact withdrawn in m+1.

The VAT to be paid for the month of December N+1 will therefore appear in the provisional balance sheet at the end of N+1.
and will be settled in January N+2.
 21

Annex budget structure of VAT:

January February December BALANCE
Elements ...
N+1 N+1 N+1 FORECAST
VAT collected on sales of
month :
+ +
Deductible VAT on purchases
billed in the month:
– –
Deductible VAT on services
rules within the month: – –
Deductible VAT on real estate
skills acquired in the month:
– –
VAT to be paid: = =

In January, we will release funds VAT on behalf of

the VAT shown in the balance sheet December N+1 who
at the end N (account 4455) will be disbursed in
January N+2

One can alsoadd a line 'VAT credit carryover', in the case where the deductible VAT
is higher than the VAT collected for certain periods.

C. THE DISBURSEMENT BUDGET

Its structure is identical to that of previous budgets, with the months in columns and the different
types of online disbursement: disbursements directly related to operations (purchases, expenses)
personnel, etc.), and other disbursements (dividend payments, loan repayments,
acquisitions of fixed assets, financial investments, etc.

It is also necessary to include an additional column to identify short-term debt items.

forecast balance generated by the time gaps between commitments and payments (by
For example, credit purchases generate an item 'suppliers'.

REMARKS
– allocations to depreciation or provision accountsdo not lead todisbursement
(very classic error source!)
Do not forget the VAT to be paid, nor the payment of taxes (for example, for corporate tax, contributions)
provisional and settlement of the balance;
The disbursements are understood to be all taxes included.
 22

D. THE GENERAL OR RECAPITULATIVE BUDGET

The general cash budget allows for the consolidation of all receipts and all disbursements.
to see how the cash balance will evolve at the end of the month, which allows for a
simulation.

We are generally led to establishseveralsuccessive cash budgets, in order to aim for

iterative trial and error towards a 'zero cash' goal.

Indeed, it is necessary to avoid significant overdrafts, which are sources of bank fees in case of overdraft, or even
even failure, but also the surpluses, which represent a lost profit, a cost
of opportunity.

Starting from a first draft, we will therefore study the possibilities of "erasing" the surpluses or
the projected deficits through appropriate measures such as:
- place liquidity;
to resort to the discounting of commercial papers;
defer certain discretionary expenses;
– etc.

In practice, this summary budget can be presented in 2 different ways, highlighting or

not the monthly cash flow variation:

First method(we assume an initial cash of 100)

January February December
Elements ...
N+1 N+1 N+1
Collections: 1,000 1,200
(–) Disbursements 800 900
Variation in cash flow 200 300
Final treasury 300 600 (X)
(1) (2)
(1) 300 = 200 + 100
(2) 600 = 300 + 300
in the provisional balance

The final cash position is obtained by an algebraic sum of the initial cash position and the variations.
monthly cash flow.

Second method
January February December
Elements ...
N+1 N+1 N+1
Initial treasury 100 300
(+) Receipts 1,000 1 200
Availability 1 100 1 500
Disbursements 800 900
Final cash 300 600

The final cash flow of month m is carried over as the initial cash flow of month m+1.
 23

E. THE FORECAST ACCOUNTS

They allow to "close" the budgetary procedure.

1. The projected income statement

Each item in the projected income statement assumes the centralization and summation of
corresponding elements scattered across the various budgets established at the level of the different services
or of the different functions.

For example, in operational budgets (production, marketing...) as in

budgets for general administrative services include projected personnel costs
that must be summed to assess account 64 of the projected income statement. The same applies to the
depreciation allowances, external services, etc.

The balance of the projected income statement allows us to calculate the projected result.
balance the forecast budget.

2. The forecast balance sheet

Its establishment is facilitated by the identification of a number of elements upstream, at the moment of
the establishment of other budgets (the 'clients' item during the preparation of the budget for
receipts for example).

The 'equity' accounts primarily reflect financial decisions regarding

of investment and financing.

On the other hand, 'bottom of the balance sheet' items, as elements of working capital needs,
result from forecasts regarding the level of activity and the characteristics of the cycle
technical-commercial (delays over time due to payment terms).

F. SELF-CORRECTING EXERCISE: SADA CASE

Statement
CAS SADA

The Aquitaine Food Distribution Company (SADA) operates in a seaside town in the southwest.
a supermarket whose activity is strongly seasonal. Indeed, the revenue excluding taxes
the average, which is 400,000 euros during the winter months, rises to 700,000 euros during the months
in summer. You are asked to participate in the preparation of the budget for the first half of the year
N+1.
 24

For this first semester, the sales forecasts (excluding tax) are as follows:

Month Forecasts (in euros)

January 400,000
February 400,000
Mars 450,000
April 550,000
I 700,000
June 700,000

December N sales were 400,000 and July N+1 sales are expected to be 700,000.
The balance of the SADA after distribution at the end of December N is as follows (in euros):
ACTIVE
Gross immobilizations ............................... 1,200,000
(–) Depreciation 240,000
Net fixed assets ............................... 960,000
Stock of goods ............................... 200,000
Clients 48,000
Treasury 12,000
TOTAL ASSETS:1,220,000
PASSIVE
Capital 600,000
Reserves 220,000
Corporate tax to be paid
(balance N after liquidation) ......................... 5,000
VAT to be paid ... 5,000
Social charges to be paid ............................ 6,000
Suppliers 384,000
TOTAL LIABILITIES:1,220,000

You also have the following information:

[Link] corporate tax balance will be paid in April N+1. Two corporate tax installments of 14,000 will be paid in March and June.
N+1. The corporate tax rate is 50%.
[Link] average storage time for goods is 15 days. It is therefore considered that it is necessary to
beginning of the month a sufficient stock to ensure sales for the first half, supposed to be equal to
half of the monthly sales.
[Link] gross margin rate (the difference between the selling price and the purchase price of goods) is
20% of the sale price. Suppliers are paid within 30 days (purchases made during the month are therefore settled on the
next month).
[Link] variable charges, the amount excluding tax of which represents 10% of the sales excluding tax, are settled each...
months, without deviation from their commitment.
[Link] monthly amortizations amount to 10,000 euros.
[Link] fixed cash outflows amount to 33,000 euros per month. This sum includes the
salaries and social charges, as well as VAT for the taxable part. Social charges are
released the following month.
 25

g. Sales, purchases, variable costs, and 50% of payable fixed costs are subjected to a
20% VAT on the amount excluding tax.
The VAT to be paid for a month is disbursed the following month. Any potential VAT credits are not
not reimbursed by the tax administration, but deferred as being deductible the following month.
[Link] including tax are collected in cash at 90%. The remaining 10% are credit sales.
received the following month.
[Link] is planned to make an investment of 60,000 euros in June N+1, subject to amortization.
linearly over 5 years, aimed at streamlining internal handling and storage operations.
This investment will only be put into service and operated at the beginning of July of year N+1. However, it will be
completely self-funded and settled in June.

Work to do

Whenever feasible, calculations will preferably be presented in tables with the

months in columns.

1. Establish the semi-annual budget for pre-tax purchases.

[Link] the amount excluding tax of the monthly fixed costs (including salaries and social charges).
[Link] the projected monthly results.
[Link] the current monthly break-even point (in turnover in euros excluding tax).
[Link] the monthly budget for the VAT to be paid.
[Link] the monthly budget for expenditures.
[Link] the monthly budget for receipts.
[Link] the general cash budget, showing the monthly variation in cash.
end of the month.

[Link] the forecast balance sheet as of the end of June N+1.

Elements of response
NOTE
The amounts are in thousands of euros, unless otherwise stated.

Question 1: Purchasing budget

Total
Month January February Mars April May June
semesterly
Sales excluding tax 400 400 450 550 700 700 3,200
Stock debut (1) 200 200 225 275 350 350
Stock fin (2) 200 225 275 350 350 350
Stock variation 0 +25 +50 +75 0 0 +150
Consumed purchases (3) 320 320 360 440 560 560
Purchases (4) 320 345 410 515 560 560 2,710
Stock debut: 50% of the sales for the month.
Stock end: 50% of next month's sales.
(3) Consumed purchases: 80% of sales, since the margin is 20%.
(4) Purchases: consumed purchases + stock variation.
 26

Question 2: Amount of fixed costs

Let M be this amount. Since VAT only applies to half, we have:

M M
+ (1,2)=33,000
2 2
M+ 21M
,
= 33 000
2
2.2 M = 66,000
66,000
M= = 30,000 euros
2.2

Question 3: Income statements

Total
Month January February Mars April May June
semesterly
Sales (excl. tax) 400 400 450 550 700 700 3,200
Purchases (excl. VAT) 320 345 410 515 560 560 2,710
Variation of stocks 0 -25 -50 -75 0 0 -150
Other variable charges 40 40 45 55 70 70 320
Margin on variable costs 40 40 45 55 70 70 320
Fixed charges eligible for disbursement 30 30 30 30 30 30 180
Depreciation 10 10 10 10 10 10 60
Results 0 0 +5 +15 +30 +30 +80

Monthly break-even point

The variable cost margin rate is:

20 - 10 = 10% of sales.

Let SR be the threshold sought:

0.1 SR – (10 + 30) = 0
40
SR = = 400soit 400,000 euros
0.1
The breakeven point of 400,000 euros corresponds well to the results for January and February of the
previous question.
 27

Question 5: VAT Budget

Month January February Mars April I June

VAT collected 80 80 90 110 140 140
Deductible VAT:
on purchases 64 69 82 103 112 112
– variable surcharges 8 8 9 11 14 14
fixed surcharges 3 3 3 3 3 3
VAT to be paid
for the month 5 0 _ – 0 11
VAT credit – – 4 11 – –
VAT to be paid: 5 5 0 0 0 0

BALANCE BALANCE
INITIAL FORECAST

Question 6: Disbursement budget

BALANCE
Month January February Mars April May June
PROVISIONAL
Suppliers
(Initial assessment) 384 – – – – –
Purchases (including tax) – 384 414 492 618 672 672
(SUPPLIERS)

Charges variables 48 48 54 66 84 84
Social charges
(Initial assessment) 6
Fixed charges 27 33(1) 33 33 33 33 6
(SOCIAL CHARGES)
VAT to be paid
(question 5) 5 5 – – – – 11
(TVA)

IS :
Balance N – – – 5 – –
Deposits – – 14 – – 14 28
(STATE,IS)

∆
Investment – – – – – 60 ASSETS
TOTALS 470 470 515 596 735 863
(1) Including social charges. Since these charges are fixed, we may not distinguish them.
 28

Question 7 : Cash Inflow Budget

BALANCE
Month January February Mars April May June
PROVISIONAL
Sales including tax 480 480 540 660 840 840
Clients (Initial Assessment) 48

Cash sales: 432 432 486 594 756 756

(90 %)
Credit sales: 48 48 54 66 84 84
(10 %) (CLIENTS)

TOTAL OF
480 480 534 648 822 840
RECEIPTS

Question 8: General cash budget

Month January February Mars April May June

Collections 480 480 534 648 822 840
Disbursements 470 470 515 596 735 863
Monthly variation of
10 10 19 52 87 minus twenty-three
treasury
Cash at the end of the month 10 + 12 = 22 22 + 10 = 32 32 + 19 = 51 103 190 167

Question 9: Forecast balance sheet

ACTIVE
Gross fixed assets: 1,200 + 60 = .................... 1,260
(–) Depreciations: 240 + (6× 10) = .................... 300
Net fixed assets 960
Stocks (question 1) ... 350
Clients (question 7) 84
State, advances paid on corporate tax ................................. 28
Treasury (question 8) ............................................ 167
Total assets:............ 1,589

PASSIVE
Capital 600
Reserves 220
Half-yearly result (question 2) .............................. 80
Suppliers (question 6) ........................................ 672
State, VAT to be paid (question 5) ............................... 11
Social charges to pay ......................................... 6
Total liabilities:............ 1,589
 29

IV. SALES BUDGET MANAGEMENT

Sales forecasting is often the starting point of the budgeting process: budgets
depend on the projected level of activity.

A. SALES FORECAST

1. Statistical methods
The simplest method is to start with an analysis of past sales and extrapolate it.
trend. Therefore, one must have a sales history.

If the evolution of sales is regular (which can be verified by observing a type of progression
Arithmetic), one can perform a linear adjustment (also called a regression) of sales V
with respect to time t, modeling sales in the form of a function of the type V = a t + b, the
The problem is to determine the coefficients a and b using the 'least squares' method.

The coefficient a is given by the formula:

(Sum of tV - n M(V) M(t)) / (Sum of t)2- n M(t)2)
with n = number of periods; M(V) = average of sales, and M(t) average of the period indices.

The regression line passing through the midpoint is:

b = M(V) - a M(t)

For example, let's assume that the trade statistics of the last 5 years regarding sales
in tonnes of product P are as follows:

Years (t) Quantities sold (V)

N–4 40
N–3 42
N–2 45
N–1 43
N 48

A time scale must be defined, for example taking t = 0 for N–4, t = 1 for N–3, etc.
In order to calculate the coefficient a, it is necessary to present the following table:

Years t V t2 TV
N–4 0 40 0 0
N–3 1 42 1 42
N–2 2 45 4 90
N–1 3 43 9 129
N 4 48 16 192
Totals 10 218 30 453

The average values of t and V are:

M(t) = 10 / 5 = 2
M(V) = 218 / 5 = 43.6
 30

We have:
a = ( 453 – ( 5× 2× 43.6) / (30 - (5× 2× 2)) = 17 / 10 = 1.7 and
b = 43.6 - (1.7× 2 ) = 40.2

We have therefore:
V = 1.7 t + 40.2

To obtain the forecasted sales in N+1, simply set t = 5:

Projected sales in year N+1 = (1.7× 5) + 40.2

= 48.7 tons

In some cases, when the sales of a product depend on another economic variable X, we
can make the adjustment not based on time, but based on this other variable, if one
has a forecast. For example, electricity production depends very broadly on the
economic growth, therefore the evolution of GDP. Tire manufacturers, for the market
from the 'first fit', establish their own sales forecasts based on the sales forecasts of
new cars made by car manufacturers, etc. It will always be a matter of modeling the
phenomenon in the form of a function V = f(X). In more complex cases, one may have a
function dependent on multiple variables: it is necessary to use econometric methods.

The activity may not evolve regularly: it may, for example, accelerate during periods of
launch, starting activities. For example, when launching a new product, in case of
Success growth can be exponential for certain periods, which can be diagnosed.
by observing increases that are not arithmetic but geometric. It is necessary then
a regression that is not linear, but logarithmic. Thanks to a change of variable
logarithmic, we reduce it to a linear regression.

Sales are modeled by an exponential function of the type V = b at

We pose:
Y = log V,
À = log a,
And B = log b.

It is known that the logarithm of a product is the sum of the logarithms:

log V = log b + log at
log V = log b + t log a

let it be:
Y=B+tA

B and A are calculated by linear regression, then we return to a and b using the power, inverse function.
of the exponential.

We will refer to the SET case presented later for an example of calculation.

One may also need an adjustment by a power function of the type:

V = b Xa.

In this case, we reduce it to a linear regression by doing: log V = a log X + log b

 31

If the evolution of sales is more erratic, adjustment methods lose their meaning, and one cannot
what to do with very short-term forecasts. It's the same problem as for prediction of
the evolution of a stock's share price. We can then use "exponential smoothing", which
it consists of making a forecast for N+1, noted PN+1, using the forecast that was made for N,
rated PN, and the actual achievement for the same period, rated RN. The model 'corrects' in
some sort of error in smoothing the numerical series.

We calculate PN+1 = c RN + (1-c)PN

The coefficient c is between 0 and 1, it needs to be 'adjusted' through learning based on

the amplitude of erratic variations to achieve the best smoothing.

One can refine by inventing models taking into account several previous periods, N–1.
and N–2 for example.

One can also bring out the trend by calculating 'moving averages', that is to say by
slidingly replacing each value of the numerical series with the average of 2, 3, or n values
from the series. For example, with order 3 averages, we replace the observation related to the
period N by the average of the observations from periods N-1, N, and N+1 (one can also take N-2,
N–1 and N). Here again, this helps to smooth the phenomenon.

Note that research is currently oriented towards the use of 'neural networks', which are
simulation software capable of learning by capitalizing on information over time
period, with interesting predictive power.

2. Marketing approaches
Statistical methods assume that one can rely on history and are based on
the hypothesis that the general trend will continue. They are therefore ineffective in the case of
new products, or to predict changes in trends. They are therefore not sufficient,
and should be supplemented by more qualitative marketing analyses.

Regarding trend changes, we can refer to the life cycle theory of

products. A product generally goes through 4 phases: launch, development, maturity, decline.

In the launch phase, we jump into the unknown. The crucial point is the decision of
launch. This is where market studies are crucial. The assessment of demand
forecast is obviously full of uncertainties. A technique often used is to make a
test launch, in a single region for example, to see customer reactions. If the launch
is a failure, at least at the start, the new product generates losses and poses a dilemma: should one-
to persist, possibly increase advertising efforts, or on the contrary stop the expenses? One can see
that the management controller is quite powerless to give advice: it's a
responsibility of sales representatives and the general management to make the decision. But the system
The accountant must allow for quantifying the progression of expenses in order to clarify the choice: one cannot
do not accumulate losses beyond reason. If the launch is a success, that is where
Generally, there will be an exponential increase in sales.

Then, once the launch is successful, the sales growth rate will stabilize, and we will enter into
a phase of regular evolution. We know that sales will increase by x% every year. The product
will conquer its potential market, which depends on the global market, but also on the market share of
the company that is the result of its notoriety. Joining a panel for example allows the company to
to position itself in relation to its competitors and help anticipate market evolution.
 32

The product then reaches the maturity phase. This is the case for products like refrigerators.
or televisions: when nearly 100% of households are equipped, there is no longer a so-called market of
replacement (we buy a device when the old one breaks down, the cost of repair being
hardly lower than the price of a new one), and the growth rates become null: we sell every year the
same quantity. Note that all the art of marketing is precisely, in connection with research-
development, to escape this stagnation through an innovation policy: refrigerators
classics make up a replacement market, but ice vending machines are
them, in the development phase! The same goes for televisions, whose sales are boosted
with flat screens or 'home theater' devices

Finally, the product is experiencing a decline phase, linked to obsolescence, due to the fact that the product is outdated.
make technological or sociological developments. Sales begin to decline, as well as the
results, and we must consider removing the product from the catalog before incurring losses. The product can
totally disappear from the market (it's only found in flea markets...), or take refuge in
some very particular "niches" (automobiles have made draft horses disappear, and therefore the
the blacksmith profession and the production of horseshoes, except for the needs of horse racing by
example).

The following diagram illustrates this evolution:

Diagram: Product life cycle

Sales DEVELOPMENT
MATURITY

LAUNCH
DECLINE

Stop of the
production

Weather

In every sector of activity, professionals generally have a good understanding of the cycle of
life (average duration of a model in the automotive industry, for example). Furthermore, at the level of
the design of products, very often the lifespan is "programmed" (case of household appliances
for example). We can therefore manage the renewal of the catalog. Even for a product in
one can be wrong about the exact moment of the trend change, concerning the whole of a
product range the global forecasts can be satisfactory. To do this, the practice that
consists of periodically asking field sales representatives, who are in daily contact with
clients, how they 'feel' the evolution of the market (which allows for example to calculate the
percentage of "optimistic" salespeople), provides valuable insights. Still
should we take into account the current behavior (so-called "post-modern") of consumers, who have
more and more tendency to 'zap' between brands...
 33

3. The voluntarist aspects: the business goals

On the other hand, it should be considered that the determination of projected sales is not solely
forecasting matter. We are not in a fully deterministic universe (as in the case of
weather forecasts for example) in which the objective conditions of the environment
Market data would allow for calculating the business volume. A budget is also an act.
volunteering, a motivational tool that allows displaying more or less ambitious goals.
In a way, "when we want, we can."

To motivate salespeople, there is a whole series of techniques: financial incentives (bonuses,

advantageous commissions), gifts and various rewarding prizes (to the best seller of
the year, which wins a trip to the Caribbean and its photo in the internal newspaper for example).

Furthermore, we can also motivate the client: through advertising, through promotional actions.
through coupons, discounts, and an attractive pricing policy.

4. The effects of elasticity and trade policy

In fact, there is a whole set of "levers" that a sales department can play on.

The price aspect is often decisive, especially when one is in a situation of 'competition.
monopolistic
the brand or the quality of after-sales service for example.

The company must therefore set its price and faces a problem of elasticity (review the
Optimal pricing policy paragraph of series 02.

It should be noted that this 'elasticity effect' of demand relative to price can also play through
in relation to the price of another product, particularly in cases of substitutability. This is referred to as elasticity.
"crossed." For example, if the price of oil and thus domestic heating fuel increases, the
consumers will be encouraged to switch to other sources, gas or solar energy for example.
The variation in the price of oil leads to a change in the demand for gas or heating devices.
solar: we will have an elasticity, this time generally positive.

More generally, trade policy is not limited to the price factor alone. Other elements,
on which we can play, are equally important: the product characteristics (more or less
good quality for example), the mode of distribution (hypermarket or specialized stores), and the mode
advertising communication (ads in mainstream newspapers or in magazines of
(luxury). Marketing specialists explain that what matters is that there is consistency between these
different elements of the 'marketing plan' (the 'marketing mix'): for example for products
cosmetics, several policies are possible: low-end products in discount stores,
luxury products in perfumeries. This consistency issue can represent a formidable
dilemma for producers. For example, can we sell a brand champagne in a
supermarket, just a few meters from the cans of green peas or diapers? Note that this
often pushes producers to diversify into 'sub-brands': the same product for
the essential is declined under different brands and packaging to reach different segments of
customers through different distribution channels.

All this shows that ultimately sales depend a lot on policy decisions.
commercial, and therefore the intelligence and creativity of the teams. Even in sectors
historically in decline, there are still companies that "manage to stand out" thanks to
a relevant policy
 34

5. Taking seasonal variations into account

Sales generally follow a phenomenon of seasonal variations. It is therefore necessary
to break down the overall budget, generally annual, into budgets established for shorter periods.
A classic practice is to establish monthly budgets, which will facilitate the establishment of others.
downstream forecasts (cash flow for example). But we can also establish according to another
periodicity.

Let's take the example of an annual budget broken down into monthly budgets.

For example, starting from the time series made up of the sales history over several
over the years, we determine the general trend, the 'trend', for example by linear adjustment. Then,
for each available monthly data, corresponding to a period i, we calculate a coefficient
seasonal by relating the observed value to the calculated value on the trend line. We
can then calculate average monthly coefficients, characteristic of seasonality. In high
In the high season, the coefficients are above 1, and in the low season, they are below 1.

To establish the monthly forecasts, for each month we calculate the forecasted value on the
trend line that is multiplied by the corresponding seasonal coefficient.

A numerical example is provided by the first questions of the JARDIPRO case presented later.

It should be noted that strictly speaking, in the case of monthly budgets, the sum S of the coefficients
seasonal workers should be equal to 12, these coefficients oscillating on average around 1. If this is not the
case, and if the difference is significant, the coefficients must be adjusted by multiplying them by 12 / S.
But in most cases, this correction can be overlooked.

B. THE SALES BUDGET

Multidimensional aspects of the problem

To establish the company's sales budget, several dimensions must be taken into account. We
we have already identified two of the dimensions to consider: the products (there is no reason why
the sales of different products evolve in a homothetic way) and the periods (one cannot
content of an overall annual forecast: due to seasonality, it must be broken down by month or
in quarters). But there are others.

We may have to take into account different markets, defined either geographically (sales
domestic and export sales for example), or by types of clientele (civil markets,
military markets.

Generally, one must also take into account the structure of the commercial network, if one wants to then
monitor performance. For example, if we have a network of regional agencies, each agency
has its budget.

Other criteria may apply in certain sectors, forcing consolidations.

still different. For example, in the telecommunications or energy sectors, the recent
regulatory evolution, which imposes on historical operators (FRANCE TELECOM,
EDF and GDF) the end of the monopoly on part of their activities requires a distinction between sector
regulated and deregulated sector, open to competition (one can expect a decrease in activity in the
deregulated sector, new entrants naturally taking market share. But let us note that
 35

the overall effects are ambivalent: for example, GDF loses part of its monopoly on gas,
but can now sell electricity...).

We are therefore in a 'multidimensional' logic. It is necessary for the needs of management.

to be able to present budget forecasts according to different 'axes' (this is actually the same
problem in management accounting.

2. A classic example: the product-market-time triptych

The problem is that when presenting a budget, we generally only have 2 tables available.
dimensions, with the information arranged in rows and columns.

For example, with an organization by geographic markets, each regional agency can establish
his budget with products in rows and months in columns:

Agency budget 1
Month J F M A M J J A S O N D Totals
product 1
oduit 2
……
I am doing
…….
rate

The establishment of this budget is based on a program, in the same format, established in quantities.
The programs and budgets of the various agencies are centralized and consolidated to have the
global budget of the company.

Schema B: Consolidation of agency budgets

….

Agency budget 2 Global budget

Agency budget 1

Downstream, the total quantities by product can be used to establish the program of
production, and the totals in euros per month can be used to establish the budget of
cash collections.
 36

3. The information system to be implemented

If we do the calculations "manually", the multidimensional problem is very difficult to manage.
The solution involves the use of database management systems (such as ORACLE)
example). This tends to generalize with the use of Integrated Management Software (ERP,
in English ERP). The basic forecasts are entered in a multidimensional manner (by
example such quantity of such product in such agency) to feed the database from
which allows to extract the desired information on demand (in particular, downstream, to present
the states of budget control

C. THE COMMERCIAL EXPENSE BUDGET

1. The deterministic aspects: the costs induced by the volume of activity

Some fixed costs reflect at a given moment the structure, organization, and importance of
commercial services (for example, the depreciation of premises, the salaries of staff who
administers sales). Some variable costs of a 'logistical' nature are largely proportional.
to the sales (for example, shipping costs by mail for a sales company
correspondence). The same goes for the variable part of the salespeople's remuneration (commissions).
For these elements, we are in the logic of the "flexible budget" discussed above. The forecasts
Budgets are dependent on projected activity, and can therefore be quantified linearly.
(Budget = Fixed Costs + Variable Costs based on Sales)

2. The dynamic aspects: the public promotional levers

On the other hand, for important positions such as the advertising budget, the logic is reversed: it is not
the advertising budget which depends on the forecasted sales, but the forecasted sales which
dependent on advertising effort (or else, it is to doubt the relevance of all the teachings.
of marketing!). Therefore, you need to start by deciding on the advertising budget, having an idea of its
commercial effectiveness, in order to then assess the projected sales (for a given price level).
Advertising is included in the budget costs, but it turns out that it is more of a
investment, which is often observed in the form of a 'lingering effect,' as in
Physics: if for example we have regularly done a lot of advertising every year, which has reinforced
the product's notoriety, and if suddenly we stop the advertising, the notoriety will continue for
for a certain period of time and in the short term, sales will decrease very little, as customers are accustomed
to buy the product. Obviously, if we continue to not advertise, sales will decline.
inexorably. But the effect is delayed.

We are therefore actually faced with a system of complex interactions, for which predictability
depends a lot on volumes and experience. For example, a SME that decides for the first
Doing a television advertising campaign to break out of anonymity has no
a way to seriously predict the results and the growth generated: it is the lottery. However,
for a multinational that operates a portfolio of well-known brands and allocates annually a
advertising budget allocated to support its market share, the problem is routine and the spending
Advertising can be considered as a percentage of turnover.
 37

D. SELF-CORRECTING EXERCISES

CAS JARDIPRO

Statement
The company JARDIPRO primarily manufactures and sells through large retail chains.
specialized in small hand tools used by individuals for gardening tasks. The
sales, very seasonal, are concentrated in the spring months. Trade statistics
monthly provides over the last two years (N–1 and N) the following amounts of sales excluding
taxes_in_thousands_of_euros

MONTH N–1 N
January 20 23
February 21 24
Mars 33 35
April 49 53
May 80 87
June 103 112
July 68 73
August 62 66
September 74 80
October 59 64
November 42 45
December 22 25
Total 633 687

First question
Graphically represent the evolution of sales.

Second question
The evolution of sales shows a slight growth that has been confirmed for several years.
In order to quantify this trend more formally, establish the equation of the line of
"trend", in the form of an R function of the type y* = ax + b, with:
y* = monthly sales in thousands of euros
x = variable time (we will take x = 1 for January N–1)
a and b = constants determined by "fitting" (or "linear regression").
(It is recalled that the slope coefficient of the regression line is given by:
∑ xy− nx y
a= , with n = number of observations, and the regression line passes through the point of
∑ x 2− nx2
coordinatesx , y).
The coefficients a and b will be calculated with a precision of at least 4 decimal places. Represent.
graphically the regression line.
 38

Third question
For each monthxithe observed amount of sales,yican be related to the value of the function of
yI
regressionyi* to calculate a seasonal coefficient
yi*

Calculate these coefficients for the 24 months of the history, then by averaging them arithmetically.
results concerning N–1 and N, the 12 coefficients characterizing monthly seasonality (for example,
the coefficient characterizing the month of January will be obtained by averaging the results obtained
for January N–1 and January N.
These coefficients will be calculated with 4 decimal places.

Fourth question
By combining the trend effect and monthly seasonality, establish the sales forecasts excluding taxes.
annualized for N+1, rounding to the nearest thousand euros. Calculate the revenue.
annual budget excluding taxes.

Fifth question
In all that follows, regardless of the results of the previous question, it is assumed that the figure
The projected annual turnover retained for year N+1 is 800 thousand euros, the most likely assumption.
The sale of products gives JARDIPRO a margin on variable costs of 40% relative to the revenue.
business, and the annual fixed costs amount to 300 thousand euros.
The company's managers reason within the following budget "range": the
Projected revenue can be between a 'low' or 'pessimistic' hypothesis at most.
5%, and a 'high' or 'optimistic' hypothesis of more than 5%, depending on the economic situation.
In this context, establish the sales budget and the annual projected income statement.
"flexible" of the company.

Sixth question
20% of JARDIPRO's sales are billed in dollars outside the euro zone. All costs are
in euros, except 10% of variable costs and 25% of fixed costs, which are billed in dollars.
Currently, one dollar is worth one euro. In year N+1, the dollar can gain or lose d% of its value by
in relation to the euro. For example, if the dollar gains 10% of its value, one dollar will exchange for
1.10 euros.
Calculate the annual projected result value of JARDIPRO, in the most central hypothesis.
likely from the budget range, depending on d. Should the company fear a rise in
dollar?

Seventh question
From what percentage p of sales invoiced in dollars is the company insensitive to
fluctuations in the value of the euro against the dollar (with a total turnover)
forecast of 800 thousand euros?
 39

Response elements of the JARDIPRO case

First question: Graphical representation

(See next page)

Second question: Trend line

The calculator must allow obtaining:

R = 0.66261x +46,7174

Graphically
y = sales in k€ 112
103

80
74

(R)
y = 55
46.7174

25
20 22

x = month
x = 12.5

Third question: Seasonal coefficients

It is necessary to arrange the calculations keeping in mind that we will have to calculate an arithmetic average of the values.
obtained for N–1 and N, for each month. It is therefore better to take the months in the following order:
1 (January N–1), 13 (January N), 2 (February N–1), etc.
 40

Calculation table:
yi
Periods xi yI y*i Average
yi*

January
N–1
N
1
13
20
23
47,3800(1)
55,3313
0.4221
0.4157
}0, 4189(2)
February
N–1
N
2
14
21
24
48,0426
55,9939
0.4371
0.4286
}0, 4329
Mars
N–1
N
3
15
33
35
48.7052
56.6566
0.6775
0.6178
}0.6477
April
N–1
N
4
16
49
53
49.3678
57.3192
0.9925
0.9246
}0.9586
I
N–1
N
5
17
80
87
50,0305
57.9818
1.5990
1,5005
}1.5498
June
N–1
N
6
18
103
112
50.6931
58,6444
2.0318
1.9098
}1.9708
July
N–1
N
7
19
68
73
51,3557
59,3070
1,3241
1.2309
}1.2775
August
N–1
N
8
20
62
66
52,0183
59,9696
1,1919
1,1006
}1,1463
September
N–1
N
9
21
74
80
52.6809
60.6322
1,4047
1.3194
}1.3621
October
N–1
N
10
22
59
64
53,3435
61,2948
1,060
1.0441
}1.0751
November
N–1
N
11
23
42
45
54.0061
61,9574
0.7777
0.7263
}0.7520
December
N–1
N
12
24
22
25
54.6687
62,6200
0.4024
0.3992
}0, 4008
calcul desy* on the
i right of the trend:
• for x = 1, we have* (0.66261× + 46.7174 = 47.3800
then, for x = 13: y* (0.66261× 13) + 46.7174 = 55.3313 and so on.
0.4221+ 0.4157
0.4189=
2
0.4371+ 0.4286
0, 4329= and so on.
2

Fourth question: sales forecast

For example, for January N+1, which is the 25thandmonth on the timescale:
y*25(0.66261× 25) + 46.7174 = 63.2826
y*25= 63.2826× 0.4189≈ 27thousands of euros
 41

Calculation table:

Forecast
Month x y* Coefficient
(in kt)
January N+1 25 63,2826 0.4189 27
February 26 63,9453 0.4329 28
Mars 27 64,6079 0.6477 42
April 28 65,2705 0.9586 63
May 29 65.9331 1.5498 102
June 30 66,5957 1,9708 131
July 31 67.2583 1.2775 86
August 32 67.9209 1,1463 78
September 33 68,5835 1.3621 93
October 34 69,2461 1.0751 74
November 35 69.9088 0.7520 53
December 36 70.5714 0.4008 28
Total 805

Fifth question: Sales budget and 'flexible' income statement

The "flexibility" here means that calculations are made for different levels of activity.

Calculations in thousands of euros:

Low hypothesis Hypothesis High hypothesis

Elements
(× 0.95 central (× 1.05
Turnover 760 800 840
Charges variables (CA× 0.6 456 480 504
Margin on variable costs
(CA× 0.4 304 320 336
Fixed charges 300 300 300
Projected result +4 + 20 + 36
 42

Sixth question: Projected result based on the percentage appreciation of the dollar
We have:
Forecasted result = Sales – (Variable costs – Fixed costs)
800 – (480 + 300

80% 20% 90% 10% 75% 25%

in euros in dollars in euros in dollars in euros in dollars
Let it be:

800 – (480 + 300)

640 160 432 48 225 75

in euros in dollars in euros in dollars in euros in dollars

Projected result:
d d d
R P= 640+ 160(1+ ) − 432− 48(1+ ) − 225− 75(1+ )
100 100 100
d
R P= 640+ 160− (432+ 48+ 225+ 75)+ (160− 48− 75)
100
Rp= 20 + 0.37d

This result is logical: for d = 0 (stability of the dollar), we find, Rp= 20, as in the question
previous.

If there is a decrease in the dollar (if the dollar loses d% of its value compared to the euro) then R pbecomes
Rp= 20 - 0.37d.

The company has no need to fear the rise of the dollar; on the contrary: its cost increases would be more
which offset by the increase in its billings.

Seventh question: Percentage of sales invoiced in dollars

Let p be this percentage; for example, if p = 20, 20% of sales are invoiced in dollars.

JARDIPRO is insensitive to the dollar effect if the increase in costs is compensated by the increase in sales.
so if
P d d d
800− × (1+ ) = 48(1+ ) + 75(1+ )
100 100 100 100

One can 'simplify' by

d
1+ let it be
100
8 p= 48+ 75
48+ 75 123
p= = = 15,375
8 8
The company must therefore bill a little more than 15% of its sales in dollars.
 43

2. CAS SET

Statement
The cooperative "Saveurs et traditions", based near Périgueux, has been bringing together for 4 years
farmers from the Southwest eager to add value to their products, and distributed by
correspondence of regional specialties (preserves, foie gras, rillettes, etc.).

Initially, sales were driven by targeted mailing operations in certain categories.

socio-professional (especially liberal professions) and on periodic follow-ups using the
client file. Noticing a slowdown in these sales, Management decided last year
to experiment with the possibilities of e-commerce via the Internet. The success has been striking: in
A few months ago, online sales became nearly as important as sales.
traditional, to the point that the SET decided to focus all its commercial efforts on this channel.

The statistics for the last 6 months are as follows (in number of orders received and
(processed):

Month Orders by mail Total

postal
1 850 210 1,060
2 831 261 1,092
3 808 321 1,129
4 790 402 1 192
5 771 496 1,267
6 748 616 1,364

The average order amount is 40 euros (excluding taxes), and the margin rate on costs
The variable cost in relation to sales (excluding taxes) is 30%.

At its inception, the company's production capacity had been sized to meet the
processing up to 1,000 orders per month. At this level of activity, the overhead costs
monthly amounts were 10,000 euros.

An increase in this level of activity has subsequently led to and will lead to an increase "by
levels of these 'fixed' costs.

Indeed, it is necessary to hire additional staff, who are paid monthly, for the
tasks for preparing shipping packages and handling. It is estimated that the monthly fixed costs
increase by 4,000 euros for every 400 additional orders.

First question
Calculate the monthly variations, absolute and relative, of orders by traditional mail.
What phenomenon is highlighted? (it will be notedxthe variable time, andythe variable requested by
traditional mail).

Second question
Model the demand for traditional mail over time, and make a forecast for the
Next 3 months (months '7', '8' and '9').
 44

Third question
Calculate the monthly variations, absolute and relative, of email orders.
What phenomenon is highlighted? (it will be notedzthe variable "demand by Internet".)

Fourth question
Model the demand by email over time, and make a forecast for the
next 3 months.

Fifth question
For the establishment of the monthly total sales budget for the next quarter, we reason
within a "range" of plus or minus 10% compared to the theoretical forecasts
provided by the previous questions. Present the corresponding 'flexible sales budget.'

Sixth question
What urgent decisions need to be made now to implement this budget?
What general lesson can we draw from this case about the usefulness of forecasting management?

Seventh question
Represent on a graph, based on the monthly activity level expressed in number of
orders, the margin on variable costs and the amount of fixed charges, and comment on the issues
posed by the search for the break-even point in this scenario.

Eighth question
Based on the sales forecasts, calculate the projected results for the next
quarter, based on two hypotheses:
a. We configure the structure on the high value of the range;
[Link] configure the structure based on the theoretical forecast. In this case, in the event of realization of
the most optimistic hypothesis (high value of the range), the company misses the sales
corresponding to the inevitable clipping of demand.
Comment on the results obtained.

Elements of response to the Set case

1requestion

My month Orders by Absolute variation Relative Variation

mail Δy=ymHimŠ1 Δy
ym ymŠ1

1 850 -
2 831 - 19 - 0.022
3 808 - 23 -0.028
4 790 -18 - 0.022
5 771 - 19 -0.024
6 748 - 23 -0.029

There is a regular decrease in demand for traditional mail, with a variation

absolute which is on average about -20. We can therefore make the following assumption: demand for
Traditional mail is a decreasing linear function of time.
 45

2equestion
It is necessary to perform a linear adjustment and model the demand using a function of the type:
y = ax + b
The desired function is obtained with a calculator (linear regression):

y870.4668 - 20.2286x

Hence the following forecasts for the next quarter:

y(7) = 870.4668 - (20.2286× 7) = 728.8666
≈ 729
y(8) = 870.4668 - (20.2286× 8) = 708,638
≈ 709
870.4668 - (20.2286× 9) = 688,4094
≈ 688

3andquestion

Month Commands Variation Variation

m by Internet absolute relative
zm Δz=z−mz mŠ1 Δz
zm-1
1 210 -
2 261 + 51 0.243
3 321 + 60 0.229
4 402 + 81 0.252
5 496 + 94 0.234
6 616 + 120 0.242

This time we notice that the variable 'demand via the Internet' seems to be increasing according to a
geometric progression. We can therefore assume that the statistical adjustment with respect to
time can be represented using an exponential function.

4equestion
To model the demand through the Internet, we are looking for a function of the type:
z= b× ax
It is possible to obtain the exponential adjustment directly with the exponential regression mode.
from a calculator and entering the pairs:
(xi, zi).
Otherwise, it is possible to reduce to a linear adjustment by using the properties of logarithms:
z= b× ax
lnz = ln(b× ax)
lnz= lnb + xlna
Let Z = lnz; B = lnb; A = lna
Z = B + xA
 46

Entering the pairs (x, lnz) allows obtaining A and B using the linear regression mode.
x z lnz
1 210 5.3471
2 261 5,5645
3 321 5.7714
4 402 5.9965
5 496 6.2066
6 616 6.4232
A = 0.2152 hence a = exp(0.2152) = 1.24
B = 5.1317 therefore b = exp(5.1317) = 169.3

The sought function is:

z= 169.30× (1,24)x

Hence the following forecasts for the next quarter:

z(7) = 169,3× 1.247≈ 763
z(8) = 169,3× 1.248≈ 947
z(9) = 169,3× 1.249≈ 1,174

5equestion
Sales forecasts in quantities:
fork
Month y z Theoretical total -10% + 10 %
7 729 763 1,492 1,343 1,641
8 709 947 1,656 1,490 1,822
9 688 1,174 1,862 1,676 2,048

Flexible sales budget:

Month Low hypothesis Theoretical budget High hypothesis
7 1 343× 40 = 53 720 59 680 65,640
8 1,490× 40 = 59 600 66 240 72,880
9 1 676× 40 = 67 040 74,480 81,920

6equestion
This budget can only be implemented if SET has sufficient processing capacity.
orders. We know that fixed costs increase in stages. Therefore, it is necessary to plan as soon as
now the structural arrangements to cope with the expected increase in activity
(hiring of order preparation staff for example).
Should this program be framed on the high assumption, taking the most optimistic value of the
fork?
The company risks finding itself with excess capacity.
In a more general context, this example illustrates the nature of budget management: it is not about
to make mechanical forecasts, but to have a simulation instrument allowing
to enlighten the decision-making (here, the necessity to evolve the structure to support a
rapid development). But this management tool is not a cure-all, it does not miraculously solve everything
the problems, and in particular it does not make the disappearriskinherent to any business.
 47

7e question
Up to 1,000 orders, the fixed costs are 10,000. Then, they increase in increments of 4,000.
batches of 400 orders.
If C is the number of monthly orders and FF the fixed costs:
C≤1 000→ 10,000
1,000 < C ≤ 1,400→ FF = 14,000
1,400 < C ≤ 1,800→ 18,000
1,800 < C ≤ 2,200→ FF = 22,000
2200 < C ≤ 2600→ FF = 26,000
2600 < C ≤ 3000→ 30,000
The unit margin is 40× 0.3 = 12 per order.

We obtain the following diagram:

Marge s/CV
FF

36,000 FF

30,000

28,000

26,000

24,000
(+)
22,000

18,000 (-)

14,000
(-)
12,000

10,000

(-)

0 C
0 1,000 1,400 2,000 3,000
SR1 SR2 SR3 SR4
834 1 167 1,500 1,834

This graph highlights 4 break-even points:

SR1, such that 12 SR1 = 10,000
10,000
let SR1 = = 834
12
 48

SR2, where 12 SR2 = 14,000, hence SR2 = 1,167

SR3, such that 12 SR3 = 18,000, so SR3 = 1,500
SR4, such that 12 SR4 = 22,000, thus SR4 = 1,834

It is only from SR4 that SET is guaranteed to be profitable regardless of the level of activity.
hatching). Below, we go through profit ranges (shaded areas) that alternate with
loss ranges (–).

8equestion
[Link] we configure the structure on the high assumption, the fixed costs will be:

Month Request Capacity Fixed costs

necessary
7 1 641 ≤1 800 18,000
8 1,822 <=2 200 22,000
9 2048 ≤2 200 22,000

The monthly result is 12 C - FF.

From which the following table:

Month Low hypothesis Average hypothesis High hypothesis

7 (1 343× 12) – 18 000 = (–) 1 884 ( –) 96 + 1 692
8 (1 490× 12) – 22 000 = (–) 4 120 ( –) 2 128 (–) 136
9 (1 676× 12) - 22,000 = (-) 1,888 (+) 344 + 2 576
Total (–) 7 892 ( –) 1 880 + 4 132
quarter

[Link] we configure the structure on the average hypothesis, the fixed fees will be:

Month Required Capacity Request Fixed costs

7 1,492 < 1 800 18,000
8 1 656 < 1 800 18,000
9 1,862 < 2 200 22,000

If the high hypothesis occurs, sales will be limited to 1,800 for month 8 and 2,200 for the
month 9.
Hence the following results table:

Month Low hypothesis Average hypothesis High hypothesis

7 (1 343× 12) - 18,000 = (-)1,884 (1,492× 12) – 18,000 = (–) 96 (1 641× 12) – 18 000 = + 1 692
8 (1 490× 12) – 18 000 = (–)120(1 656× 12) - 18,000 = + 1,872 (1 800× 12) – 18 000 = + 3 600
9 (1 676× 12) - 22,000 = (-)1,888 (1,862× -21,988 +344 (2 200× 12) – 22 000 = +4 400
Total quarter (–) 3,892 + 2 120 + 9 692

We achieve better forecast results. For month 7, the results are the same, and
For month 9, the issue of choice does not arise since we must move to the level of 22,000 in charges.
fixes anyway.
On the other hand, for month 8, we should delay investments, as the risk of overcapacity is
too big.
 49

V. STOCK MANAGEMENT AND PROCUREMENT

Inventory and supply management aims to define the program
procurement that allows managing stocks at the lowest cost taking into account two requirements
contradictory:
Limit the level of stocks.
Avoid stockouts.

Stocks are a necessary 'evil':

They are heavy consumers of financing ('the stock is money that sleeps'), of space,
time...
- However, even if the "zero stock" objective is frequently announced, they ensure a function of
regulation by fluidifying the relationships between successive workstations, between suppliers and
the clients.

Stocks contribute to the economic performance of the company: an excess weighs on costs (therefore
on profitability), a shortcoming exposes to customer dissatisfaction or, internally, to a break in
the activity.

Moreover, obsolescence, unsold goods, and speculation are other factors to consider.

A. THE ADMINISTRATIVE MONITORING OF STOCKS

The growing computerization of inventory management systems (generalization of barcodes and

Optical readers have simplified the administrative management of stocks.

In a company, the diversity of stocks is such that it is often not appropriate to track them all.
Indeed, it can be lengthy and costly to carry out a complete follow-up of all the elements.

The administrative stock tracking methods aim to determine the stocks that will be
the subject of a precise follow-up.

This segmentation is based on the principle of management by exception, according to which a small number
Stored references represent a significant share of consumption, while the rest
represents a weak [Link] importance is generally assessed by the value of
consumptions.

The segmentation carried out is based on an empirical law observed by PARETO (1848-1923) known as the law
from 20-80: about 20% of the stored references represent 80% of the consumption value
total.

Thisapproximate distinctionleads to classify the stored references into two groups: the
references that will be subject to rigorous monitoring and those that will undergo a more
please.

An extension of this method (ABC method, no connection with activity-based accounting)

consists of dividing the stocks more finely into three groups:

Percentage of references Percentage of consumption in value

Group A 10% 65%
Group B 25 % 25%
Group C 65 % 10%
 50

Limits of these methods:

Quantitative approach: some low-cost references can be fundamental.
The restocking policy depends on the quality of the relationship with suppliers, of the
suppliers' ability to react to deliver unplanned orders...

EXAMPLE1

Company A has decided to adopt differentiated management of its stored references based on
20/80 method.

Work to be done:
Establish the concentration curve of the consumption of the references used, and deduce from it,
on one hand, the references to be managed individually and, on the other hand, the references that will be subject to
quarterly orders.

Annex
References Unit purchase cost Quantity consumed
A 57 1,000
B 17 2,300
C 13 1,200
D 72 2,500
E 27 750
F 72 660
G 95 8,000
H 60 530

Elements of response
Ranking of consumptions according to their value:
References Unit purchase cost Quantity consumed Total value
A 57 1,000 57,000 3
B 17 2,300 39 100 5
C 13 1 200 15 600 8
D 72 2,500 180,000 2
E 27 750 20 250 7
F 72 660 47 520 4
G 95 8,000 760,000 1
H 60 530 31,800 6
Total 1,151,270
 51

– Calculation of cumulative increases in percentage of consumptions and references:

Cumulative growth of Cumulative rising of Cumulative total of
References
consumptions € consumptions in % references in % (1)
1G 760,000 66.01% 12.50%
2D 940,000 81.65% 25.00%
3A 997,000 86.60% 37.50%
4F 1,044,520 90.73% 50.00%
5B 1,083,620 94.12% 62.50%
6H 1,115,420 96.89% 75.00%
7E 1,135,670 98.64% 87.50%
8C 1 151 270 100.00% 100.00%
There are 8 references. Each reference therefore represents 1/8 (or 12.5%).

Concentration curve:

Cumulative consumption in %

100.00%

90.00%

80.00%

70.00%

60.00%

50.00 %

40.00%

30.00 %
20.00%

10.00 %

0.00%
0.00% 12.50% 25.00% 37.50% 50.00% 62.50% 75.00% 87.50 % 100.00%
Cumulative references in %

According to the 20-80 rule, references G and D, which represent a quarter of the references and 81.65% of the
The value of consumptions must be subject to close monitoring.

The ABC method would categorize reference G in the first group, followed by references D and A.
second group, and finally the references F, B, H, E, and C for the last group.
 52

B. MAIN STOCK MANAGEMENT MODELS IN CERTAIN FUTURE

1. Presentation of the parameters

Data from a problem

(KNOWN ELEMENTS)
Notation Comments

Demand (or consumption). D Physical units to be satisfied during the period

of management.
It may have an internal origin (necessary consumption for
a production) or external origin (customer requests).
Stock management period. P Duration for which the request is expressed.
Launch cost (or cost of cl Internal costs incurred by the company to make a
placing an order. order.
Storage cost for one unit of cs Cost of ownership of the stock for a unit of product.
product. a unit of time.
It can be expressed directly in euros and/or in the form
of rates.
– Enhance cs in the case of a possession rate:
cscost of the good× rate
– A year can mean:
P=1
P = 12
P = 360
according to the unit of measurement of costs

Cost of shortage for one unit of cp Cost related to stockout for a unit of product
product. for a unit of time.

UNKNOWN
(A DETERMINER)
Notation Comments

The quantity to order each Q Do not confuse with the request that is known.
restocking.
The number of orders to place n
during the management period.
The duration of the approval period T Time interval that separates two supplies
missioning successive.

There is a relationship between the parameters to be determined:

D D P
n= Q= T=
Q n n

For example, if a demand of 10,000 is expected for a year:

If Q = 5,000 then n = 2 and p = 6 months
Q = 2 000 alors n = 5 et p = 2,4 mois

Objective: to minimize the total cost of inventory management for the management period
Total launch cost CL
Total storage cost CS
Total cost of the shortage CP
Total inventory management cost CG
 53

2. The WILSON model

a. Main hypotheses
The Wilson model assumes that:
The demand is regular and certainly known.
Restocking is instantaneous.

As a result:
Stockouts are not taken into account,
The existence of safety stocks is unnecessary.

Graphical representation of stock evolution

Stock

Breakups being impossible, the

necessary stock is equal to Q.

Average stock

Time
T T T T
n supply periods T for the management period

b. The function to study

The objective is to determine the procurement program (Q, n, T) that minimizes CG, the cost.
total inventory managementfor the management period.
Shortages are, by assumption, not possible:
CG = CL + CS

Study for a supply period T

Stock The average stock = (initial stock + final stock) / 2
Q
Q – Cost of storing the average inventory: × cs × T
2
– Cost of placing an order: cl

Weather
T

For the management period P: these costs repeat n times

Q
CG = cl× n + × cs × T× n
2

P
Or T =
n
 54

Q P
CG = cl× n + × cs× ×n
2 n
Q
CG = cl× n + × c× P
2

The CG study can be performed interchangeably based on the variable Q or the variable n.
The search for the value that nullifies the derivative allows determining the optimal values.

EXAMPLE2

Company B plans to sell 7,200 items next year, which it purchases for €150 each.
wants to determine its optimal procurement program.
The cost of placing an order is €180.
The storage manager estimates that the cost of storage consists of the carrying cost of inventory.
which expresses the cost of tied-up money, that is, 0.5% per month, and the physical cost of storage, that is
€0.02 per item per day.
Cost elements Value Unit of time of cs Value of P
Ownership rate 150× 0.5% = 0.75 month P = 12
Cost 0.02 day P = 360

Q Q
CG = 180× n + × 0.75× 12 + × 0.02×360
2 2
CG = 180 n + 8.1 Q

The choice of the variable to retain for expressing the total cost of inventory management is generally dictated.
in the statements.
7,200 7,200
CG(Q) = 180 + 8.1 Q CG (n) = 180 n + 8.1
Q n
1296000 58320
Let CG (Q) = + 8.1 Q Let CG (n) = 180 n +
Q

The search for the value that nullifies the derivative allows us to determine the optimal values.

1296000 58320
Studied function CG (Q) = + 8.1 Q CG (n) = 180 n +
Q n
Study interval ] 0 ; +∞ [ ] 0 ; +∞ [
1296000 58320
Derivative CG’(Q) = – + 8.1 CG’ (n) = 180 –
Q 2
n2
Canceled for(1) Q2= 160,000 n 2= 324
Let Q = 400 Let for n = 18
Where from Q* = 400 articles n* = 18 commands
n* = 7 200 / 400 = 18 orders Q* = 7,200 / 18 = 400 articles
T* = 360 / 18 = 20 days T* = 360 / 18 = 20 days
CG (400) = 3,240 + 3,240 = 6,480 € CG (18) = 3,240 + 3,240 = 6,480 €
 55

Graphical representation

Costs Costs
CG (Q) CG (n)
CS (Q) CL (n)

6,480 6,480

3,240 3,240
CL (Q) CS (n)
Q n
400 18
(1)
The second derivative being positive, the point that nullifies the derivative is a minimum.

The optimal program of company B therefore consists of placing 18 orders of 400 items each,
a order every 20 days, for a total inventory management cost of €6,480.

The graphical study of functions revealsthat at optimum the total launch cost is
equal to the total storage cost, and that the optimum corresponds to a minimum.

c. Limits of the model

The Wilson model assumes that stockouts are not possible. In practice, the
companies can seek to limit this risk by holding safety stock or can, in
on the contrary, integrate this risk to limit their average stock.

The Wilson model does not take into account that the ordered quantities often have a
impact on the purchase price.

These shortcomings have led to adaptations of the Wilson model.

3. Model incorporating a safety stock

To avoid stockouts, companies can hold a safety stock that allows for
to face unexpected events (delivery delays, surge in demand, ...).

The existence of a safety stock increases the average stock:

Stock

(Q+ S)+ S
Q+S Average stock =
Q 2

Q
Average stock = +S
S 2
Safety stock (S)
Weather
 56

EXAMPLE2 (continued)

Company B decides to maintain a permanent stock of 50 items to deal with unforeseen events.
⎛Q ⎞ ⎛Q ⎞
CG = 180 n + ⎜ + 50× ⎟ 9 + ⎜ + 50× ⎟⎠ 7.2
⎝ 2 ⎠ ⎝ 2
CG = 180 n + 8.1 Q + 810

Studied function 1296000 58320

CG (Q) = + 8.1 Q + 810 CG (n) = 180 n + + 810
Q n
Study interval ] 0 ; +∞ [ ] 0 ; +∞ [
Derivative 1296000 58320
CG'(Q) = - + 8.1 CG’ (n) = 180 -
2
Q2
Cancel for(1) Q2= 160,000 n2=324
Let Q = 400 Let n = 18
Where from 400 articles n* = 18 orders
n* = 7 200 / 400 = 18 orders 400 articles
T* = 360 / 18 = 20 days T* = 360 / 18 = 20 days
CG (400) = 6,480 + 810 = 7,290 € CG (18) = 6,480 + 810 = 7,290 €

The optimal program of company B is therefore unchanged. It consists of placing 18 orders of 400.
each article, thus an order every 20 days.
On the other hand, the total cost of managing the stock becomes equal to €7,290.

Conclusion: thesafety stock has no impact on the determination of the program

supply(Q*, n*, T*) because this is a constant whose derivative is zero. In
in return, the overall cost of managing the stock is higher:it is necessary to store and finance the stock of
security.

4. Alert stock (or critical stock)

The alert stock or critical stock is thestock level that triggers an orderIt is about
stock level that allows to meet demand during the delivery time and to maintain the
safety stock if it exists.

Alert stock = consumption during the delivery period + safety stock

It can be determined graphically or algebraically, and two situations need to be distinguished: the case where the
the delivery time is shorter than the supply period T and the case where the delivery time is
superior to the T supply period.

If the delivery time exceeds the length of the supply period, it is appropriate
to remove the current order(s).

EXAMPLE2 (continuation)
Calculate the alert stock of company B if the supplier delivery time is 5 days, without
safety stock and with safety stock
Calculate the alert stock of company B if the supplier delivery time is 25 days, without
safety stock and with safety stock
 57

Elements of response
When the delivery time from suppliers is 5 days:
The stock at the beginning of each supply period is 400 items, and the duration of the
The supply period is 20 days, the daily demand of company B is 20 items.
cycles (20 = 400 / 20; or 7,200 / 360 = 20).
The demand over 5 days is therefore equal to 100 items (100 = 20× 5).

Without safety stock:

The safety stock being zero, the alert stock equals 100 items.
Stock

A delivery L must take place at the

end of the T period. An order
C must be done 5 days in advance.
By "projection", the alert stock
SA
SA is determined.

5 days
C L

With safety stock:

The safety stock being 50, the alert stock is equal to = 100 + 50 = 150 items.

Stock

SA

5 days
C L

When the suppliers' delivery time is 25 days:

The demand over 25 days is therefore equal to 500 items (500 = 20× It will be necessary to remove the
ongoing orders since the delivery time is more than 20 days.
Without safety stock:
The safety stock being zero, the alert stock is equal to = 500 - 400 = 100 items.
 58

Stock
This order will be delivered
during the delivery period
Q

SA

25 days
C L

With safety stock:

The alert stock is equal to = 500 + 50 - 400 = 150 items

Stock
This order will be delivered
during the delivery period
Q

SA

25 days
C L

5. Model with sliding scale rates

Suppliers offer a volume discount based on the quantities ordered, thus providing a reduction in
because one of the hypotheses of Wilson's model.

a. Determination of the total procurement cost

The objective is modified: it is to minimize thetotal procurement cost, I noted THIS, forto hold
account for savings generated by the decreasing purchase price.
CA = Purchase cost + Total inventory management cost

The total procurement cost will bestudied for each price range.
 59

b. Determination of the minimum

Finding the optimal solution requires calculating the value that nullifies the derivative for each
price hypothesis, andto verify the consistencybetween the obtained value and the supply conditions
h agreements proposed by the supplier.

If the value obtained is compatible with the interval, the function is studied for this point; if it is not
the function is studied for the bounds of the interval.

EXAMPLE3

Company C expects its consumption of raw materials to be 900 liters for the year.
to come
The storage cost is estimated at 10% per year and the cost of placing an order is €800.
The price conditions of his usual supplier are as follows:
Ordered quantities Q Price per liter
Q < 100 liters 400 €
100≤ Q < 300 liters 350 €
Q≥ 300 liters 300 €

Determine the optimal quantity to order.

Elements of response
Q
900× p + 800 n + × (p× 10 %× 1
2
900 Q
CA (Q) = 900 × p + 800 + × 0.1 p
Q 2

Q< 100 100≤ Q< 300 Q ≥ 300

p = 400 p = 350 p = 300
CA (Q) 720,000 720,000 720,000
360,000 + + 20 Q 315,000 + + 17.5 Q 270,000 + + 15 Q
Q Q Q
Derivative 720,000 720,000 720,000
− + 20 − + 17.5 − + 15
2 2
Q Q Q2
Is canceled for Q*≈ 189.73 Q*≈ 202.83 Q*≈ 219.09
Note Incompatible with the slice Compatible with the slice Incompatible with the range
Study CA of the boundaries Study CA (Q*) Study the CA of the boundaries

STUDY of CA 1,080,020 € CA (202)≈ 322,099 € Revenue (300) = 276,900 €

CA (99)≈ 369,252 € 284,300 €
Note: for the last interval, the maximum quantity to order is not infinite (∞) but D.

The comparison of the total procurement costs of company C indicates that the total cost
The supply is minimal for an order of 300 liters.

The optimal procurement program therefore consists of placing 3 orders of 300 liters.
each one, that is one order every 4 months, for a total supply cost of €276,900.
 60

6. Model with shortage (or model with disruption)

TheThe stock level is intentionally limited to reduce storage costs..
Stockouts occur when the stock (S) at the beginning of the supply period is
less than the economic quantity needed to meet the demand for the period (Q).

Stockouts translate into lost earnings: the cost of shortage.

This model assumes that demand is 'captive', meaning that unmet demands
are not lost, but simply deferred to the next period.

To solve issues with shortages, two approaches are possible: one approach
"mathematics" and a simpler approach but based on memorizing a link with the
basic Wilson model that does not demonstrate the optimum.

a. Presentation of the model

Study for a supply period T

Limited stock at S, lower than

need Q of the period
S

Ts Tp
Q-S

T = Ts + Tp

Launch cost = cl
S+ 0 S
The average stock held during the storage period Ts is equal to: =
2 2
S
Storage cost = × cs × Ts
2
The stockout starts at the beginning of period Tp and reaches Q - S at the end of the period
0 + Q − S Q− S
The average shortage during the period T is
p equal to: =
2 2
Q− S
Cost of shortage = × cp× Tp
2
 61

According to the properties related to similar triangles (or Thales' theorem), and knowing that
P
T = , it is possible to write:
n
TS T S P S
= Where Ts =T Let it be T= ×
S Q Q n Q

Tp T Q−S P Q− S
and = Where Tp=T Let it be Tp= ×
Q− S Q Q

For the management period P: these costs are repeated n times.

CL = cl × n
S
CS = × cs× Ts× n
2
Q−S
CP = × cp× Tp× n
2

⎡S ⎤ ⎡ Q− S ⎤
CG = [ cl× n ] + ⎢ × cs× Ts× n ⎥ + ⎢ × c×
p
T×pn ⎥
⎣2 ⎦ ⎣ 2 ⎦
⎡S P S ⎤ ⎡ Q− S P Q− S ⎤
Let CG = [ l×cn ]⎢ + [× cs× × × n ⎥ + ⎢ × c p× × × n⎥
⎣2 Q ⎦ ⎣ 2 n Q ⎦
⎡S S ⎤ ⎡ Q− S Q− S ⎤
Where CG = [ l×
c n ] ⎢+ × c× ×P ⎥ + ⎢ 2 × c× × P⎥
⎣2 Q ⎦ ⎣ p Q ⎦
D
Or n = ,
Q
⎡ D⎤ ⎡S S ⎤ ⎡ Q− S Q− S ⎤
So CG (S, Q) = ⎢ c×
1 ⎥ + ⎢ × cs × × P⎥ ⎢ × c p× × P⎥
⎣ Q⎦ ⎣2 Q ⎦⎣ 2 Q ⎦

It is a function of two unknowns S and Q.

The search for values that cancel the first derivatives allows for the determination of values
optimal.

EXAMPLE4

Company D plans to resell 1,500 product Xs for the upcoming year. It wants to determine its
optimal supply program.
The cost of launching an order is €243.
The storage manager estimates that the cost of storing a product X is €0.75 per month, and
that the cost related to the shortage of a product is €0.225 per day.

1,500 S S Q− S Q− S
243 + × 0.75× × 12 + × 0, 225× × 360
Q 2 Q 2 Q

364 500 S2 (Q− S)2

CG (S, Q) = + 4, 5 + 40, 5
Q Q Q
 62

364 500 S2 (Q2− 2 SQ+ S2)

CG (S, Q) = + 4.5 + 40.5
Q Q Q

364,500 S2 S2
CG (S, Q) = + 4.5 + 40.5 Q - 81 S + 40.5
Q Q Q

364,500 S2
CG (S, Q) = + 45 + 40.5 Q - 81 S
Q Q

• It is necessary to start by calculating the first derivative with respect to S (the variable Q is
considered as a constant :

S
CGS (S, Q) = 90 - 81
Q

S 81
The derivative is zero for =
Q 90

S
Let it be for 0.9
Q

• First derivative with respect to Q (the variable S is considered a constant):

364 500 S2
CG’Q(S, Q) = - - 45 + 40.5
Q2 Q2

S
Or, from the derivative with respect to S, we know that = 0.9
Q
364 500
CGQ(S, Q) = - -45 (0.9)2+ 40.5
Q2
364 500
CGQ (S, Q) = – + 4,050
Q2
The derivative is zero for Q2= 90,000
Let Q* = 300 (we will admit that it is a minimum).

We deduce that:
S
= 0.9 hence S* = 0.9× 300 = 270 products
Q
n* = 1,500 / 300 = 5 orders
T* = 360 / 5 = 72 days
S
Ts= T where Ts = 72× 0.9 = 64.8≈ 65 days
Q
Q− S 300− 270
Tp=T where Tp= 72× = 72× 0.1 = 7.2≈ 7 days
Q 300

364,500 2702
CG (270,300) = + 45 + 40.5× 300 - 81× 270 = 2,430 €
300 300
 63

The company must therefore have an initial stock of 270 products and place 5 orders per year of
300 articles, which means an order every 72 days. During each supply period, it
will be able to meet the demand for 65 days (exactly 64.8) and will be out of stock for 7 days
(7.2 exactly). The total stock management cost is €2,430 per year.

b. Link with the Wilson model

The shortage rate, notedαallows for a link between the Wilson model (without shortage) and the
model with shortage and check the value that cancels the first derivative with respect to S:

cp
α=
c p + cs
Express cpand cs in relation to the management period P.
Note:α = S / Q

WILSON model Model with break

Qw 1
Q* = Qw×
α
nw n* = nw× α
Tw 1
T* = Tw×
α

EXEMPLE4 (continued)

Verify the results obtained from those of the Wilson model.

Compare the results and conclude.

Elements of response
If company D does not incorporate the shortage cost (Wilson model), the total inventory management cost
east
Q
CG = 243 n + × 0.75× 12
2
1,500
CG (Q) = 2 43 + 4.5 Q
Q

364 500
+ 4.5 Q
Q

364 500
CG' (Q) = - + 4.5
Q2
The derivative is zero for Q≈ 284,604
Where not≈5.27 and T≈ 68.30
CG (284,604) = 2,561.44

0.225× 360
The company's shortage rate is:α = = 0.9
(0,225× 360)+ 0.75× 12
 64

S
Here we find the value which cancels the first derivative with respect to S of the model with shortage.
Q
Wilson Model Model with shortage
1
Q≈ 284, 604 Q* = 284,604 × 300 products
0.9
n≈ 5.27
T≈ 68.30 n* = 5.27× 0.9 = 5 orders
CG = 2,561.44 €.
1
68.30× 72 days
0.9
CG = 2,430 €.

Conclusion: the model with stock breakageallows to reduce the total cost of inventory management
since the beginning period stock is deliberately smallerthat the quantity
necessary Q.
On the other hand, the company risks not meeting the entire demand for the period.

C. STOCK MANAGEMENT MODELS IN UNCERTAIN FUTURE

Unlike previous models, demand is uncertain and, to protect against fluctuations in the
request, it is appropriate to create a stock.

The objective of models in uncertain futures is todetermine the stock level S* at the beginning of
each supply period Tto minimize management costs.

Models are numerous and sometimes complex...

Only the case of costs not proportional to time will be studied.

1. Demand is a continuous random variable

It will be accepted that the random management cost is minimum for the stock value S such that:
Cp
P(D≤
Cp+ Cs

EXAMPLE5

A butcher shop has found that the daily sales of beef is a random variable that follows
a normal distribution N(500 kg; 75 kg).
The meat is sold for €25 per kilo and the butcher makes a profit of €10 per kilo.
Unsold meat is sold as animal meat at a price of €7 per kilo.
Determine the optimal stock of meat that the butcher shop should display for sale each day.
deduct the safety stock.
Determine the stock level to maintain in order not to exceed a stockout rate of 10%.
Determine the average number of days of out-of-stock for a year (open 315 days) for a stock.
daily of 510.5 kg.

Elements of response

– cp= lost revenue due to demand exceeding stock = 10 €

 65

Purchase cost = 25 - 10 = 15 €
cscost of an unsold item = 15 - 7 = 8 €
The management cost is minimal for:
10
P(D < S) = 0.5555
10+ 8
S − 500 S− 500
P(T < 0.5555 we set t =
75 75
P(T < t) = 0.5555
S − 500
t ≈ 0.14 where 0.14 and S* = 510.50 kg
75
494.5 kg
Safety stock = 510.5 - 500 = 10.5 kg
- P(D > S) = 0.1
⎛ S− 500 ⎞
P(T < ⎜T< = 0.9
⎝ 75 ⎟⎠
S− 500
≈ 1.28
75
596 kg
Safety stock = 596 - 500 = 96 kg
For one day, the probability of being out of stock is 1 - 0.5555 = 0.4445
Daily sales are independent of each other.
The risk of rupture follows a binomial distribution B (315; 0.4445).
This law can be approximated by a normal law (n > 30 and npq > 10)
N (315× 0.4445 ; 315× 0.4445× 0.5555
N (140 ; 13.25)
So about 140 days of stock outage for 315 days of opening.

2. The demand is a discrete random variable

It is appropriate to present a matrix of inventory management costs (rows) based on demand.
(columns) and to retain the decision that minimizes the expected cost.

EXAMPLE6

The probability law of the weekly demand for product P is as follows:

D 0 10 20 30
P(D) 0.1 0.4 0.3 0.2

Determine the optimal stock to hold at the beginning of each week, knowing that the shortage cost is
6 € per missing item and the storage cost is 4 € per unsold item.
 66

Elements of response

Request
0 10 20 30 E (Cost)
Stock
0 0 60 120 180 96(3)
10 402 0 60 120 46
20 80 40 0 60 36
30 120 80 40 0 56
(1)
The demand is for 10, but nothing can be sold since the stock is zero. So the shortage cost is for 10 products.
(2)
The demand is 0, but the stock is equal to 10. Therefore, storage cost for 10 unsold products.
(3)
With the statistical function of a calculator, enter for each line the cost and the associated probability and search for the
average. If you can't do it, apply the formulaΣ ci× [Link], 96 = 0× 0.1 + 60× 120.4× 0.3 + 180× 0.2

The optimal solution is a stock of 20 products P at the beginning of each week to minimize
the expected cost of inventory management.

D. BUDGETING OF SUPPLIES

Budgeting reveals the timing of consumption forecasts.

order, delivery, and stock level.

When consumption is regular, there is noywithout any particular difficulty to carry out the
budgeting.

When consumption is irregular, the supply manager has a choice between two
budgeting methods :
– Ordering constant quantities with variable periodicity: this procedure facilitates storage.
– Order variable quantities at constant intervals: this procedure simplifies the work
administrative by introducing regularity in the schedule.

The budgeting of supplies can be achieved through a graphical method and/or by a

"accounting" method.

EXAMPLE7

Company E plans to sell 1,500 goods M for the coming year at €119.6 including tax.
Company E applies a margin rate of 25% on the purchase cost and applies the normal VAT rate.
The cost of placing an order is €400 and the holding cost is estimated at 0.5% per
month.

The projected monthly sales are as follows:

Month J F M A M J J A S O N D
Sales 50 150 100 75 80 80 120 125 90 200 160 270
Cumul 50 200 300 375 455 535 655 780 870 1,070 1 230 1,500

The stock on January first is 300 units of merchandise M.

The lead time for supply is one month and the safety stock is 15 days.
The initial stock must be replenished at the end of the year.
 67

The optimal procurement program is determined from the Wilson model:

The storage cost requires prior calculation of the purchase cost = 100 / 1.25 = 80 €.
1,500 Q
CG ( Q) = 400 × + × (0.5%× 80)× 12
Q 2
600,000
CG' (Q) = - + 2.4
Q2
The resolution gives: Q* = 500 n* = 3 T* = 4 mois

a. Order constant quantities at variable periodicity

The steps of graphical resolution are as follows:
Trace the cumulative consumption
– Postpone the initial stock,
Determine the first break date (when the stock is equal to the cumulative consumption),
Postpone from this break the safety stock then delivery time to get the date of
delivery and the order date,
Postpone the delivery to the delivery date.

2,000 Cumulative consumption

1,900
1 800
1,700
Do not forget to report L3
1,600 Q = 500
1,500 on date L3
1,400
1,300
1,200
Delivery
1 100
Q = 500
1,000 on the date L2
900
800
Delivery
700 Q = 500
600 on date L1
(and not R1)
500
400
300
200
100
0
0 1 2 3 4 5 6 7 8 9 10 11 12
C1 L1 R1 C2 L2 R2 C3 L3 R3

Attention to the reading: R2 at the beginning of September. (and not

15 days before the breakup, there must be August); L2 end of August and not end of July
a delivery ordered 1 month in advance
 68

The accounting method or table method requires presenting a December N–1 line for the
report of the initial stock and a theoretical stock column.

Month Need Stock Date of Delivery Stock Date of Date of Quantity

theoretical rupture rectified order delivery ordered
Dec N–1 300
J 50 250
F 150 100
M 100 0 31/3 500 500 February 15 March 15 500
A 75 425
M 80 345
J 80 265
J 120 145
A 125 20 500 520 July 21 August 21 500
S 90 - 70 6/9 430
O 200 230
N 160 70 500 570 October 23 November 23 500
D 270 - 200 8/12 300

The first break occurs at the end of March. A delivery will be made 15 days prior, on March 15th, and the
The order was placed on February 15. The March stock is then corrected to be increased to 500.

The second break occurs in September. At the end of August, the stock (before delivery) is equal to 20 and
Knowing that the consumption for September is 90 over 30 days, a linear interpolation allows
to specify the date:
⎧ 30→ 90
⎨ J → 20 90 J = 30× 20 J = 6.67 = 6 (as a precaution, round down by default)
⎩

The third break occurs in early December. At the end of November, the stock (before delivery) is equal to 70 and
Knowing that the consumption for December is 270 for 31 days, a linear interpolation allows
to specify the date:
⎧ 31 → 270
⎨ J→ 70 270 J = 31× 70 J = 8.03 = 8 (for safety, round down by default)
⎩

b. Order variable quantities at constant intervals

Once the initial dates for rupture, delivery, and order are determined, it is advisable to...
report 4 months later (value of T), then calculate by difference the amount to be ordered.
 69

2,000 Cumulative consumptions

1,900
1 800
1,700 Do not forget to report L3
1,600 By difference between D and the
previous deliveries
1,500
Q = 1,500 - 355 - 575 = 570
1,400
1,300
1,200 1,230
end
1 100 Delivery for
Nov.
1,000 reaching 1,275, that is
Q = 1 230 – 655
900 Q = 575
800
Delivery for
700 arriving at 655, that is
600 Q = 655 – 300 Cumul 655
500 Q = 355 late July
400
300
200
100
0

0 1 2 3 4 5 6 7 8 9 10 11 12
C1 L1 R1 C2 L2 R2 C3 L3 R3

Reporter 4 months later the dates

Accounting method or table method:

Stock Date of Stock Date of Date of Quantity

Month Need Delivery
theoretical rupture rectified order delivery ordered
Dec N–1 300
J 50 250
F 150 100
M 100 0 March 31 355 February 15 March 15 355
A 75 280
M 80 200
J 80 120
J 120 0 31/7 575 June 15 July 15 575
A 125 450
S 90 360
O 200 160
N 160 0 30/11 570 October 15 November 15 570
D 270 300
 70

The frequency is 4 months. The volume of orders will correspond to the outputs of the period to
to come 355 = 75 + 80 + 80 + 120
575 = 125 + 90 + 200 + 160
The last delivery is obtained by difference: 570 = 1,500 - 355 - 575

E. LE JUST IN TIME

Inventory and supply management has been profoundly transformed under pressure from
new just-in-time (JIT) production organizations.
Classical methods are no longer really suited to current production conditions.
flexible and with reduced stocks.

Downstream production management starts with demand to initiate production and seeks to
satisfy the demand just as it arises, in the right quantity, in the right quality.
This approach, launched by Toyota in the 1970s, aims to reduce costs and timelines through the
just-in-time production, that is to say without waiting or stock.
The JAT was popularized by the slogan of the 5 zeros to achieve (zero paper, zero stock, zero defects, zero
deadline, zero breakdown) which symbolizes total quality.

A crucial issue: the elimination of stocks

For Japanese authors, the JAT is more than just a production management model: it is a
philosophy.
Thus, for the Japanese, stocks are not a regulator to prevent dysfunctions, but
a manifestation of their existence (breakdowns, social conflict, absenteeism, supply disruption
ment...).

Organizational conditions for the success of the JAT:

– Improve the flow of information. Indeed, production is carried out in response to the
request, it is necessary that it arrives without error.
The two main tools of the JAT are:
WInternally, the KANBAN system is an information system based on labels.
to ensure downstream management.
Each stage of production produces only for the ordered quantity when it receives the label.
the downstream stage.
The simplicity of kanban is only apparent. For the system to work, it is necessary that the
the production process has been made reliable (zero breakdowns, flawless coordination between positions,
guaranteed external supplies...)
WExternally, electronic data interchange (EDI)
Rationalize the localization.
For example, the seats of Renault vehicles are manufactured by a subcontractor located a few
kilometers. Every 30 seconds, when a vehicle is launched on the assembly line, a
EDI transmission automatically triggers the order with the subcontractor. A truck delivers.
every 20 minutes, and about 2 hours elapsed between the production order and the
delivery.
Rationalize the workshops:
WReliability of machines (importance of maintenance), versatility of machines
WQuick changeover of tools (SMED :single minute exchange of die)
WMinimize movements, eliminate unnecessary operations...
Development of self-control at each stage (hence motivation, staff training)
Structural easing: reactivity.
 71

Limits of just-in-time production methods:

Inability of the JAT to adapt to significant volume fluctuations. Other means must
to be implemented such as job flexibility or work schedule adjustment.
Difficulty in solely aligning with the demand coming from the market. Even Toyota, initially
you JAT, are required to make forecasts to guide production volumes, even if it means
customize the products at the last moment.

To this end, the OPT (Optimized Production Technology) is a method that is the extension of
MRPManagement of Productive Resources) and of thekanbanand which will be presented in the next
paragraph.

VI. PRODUCTION MANAGEMENT

The role of the production function is to provide the goods and services corresponding to a
demand expressed in the market, through an effective combination of factors of
production to achieve set objectives in terms of volume, quality, deadline, cost.
Production management is a set of activities that allows organizing and
coordinate the physical flows and the information flows necessary for preparation and implementation
in the implementation and control of production processes.
Two main production management systems can be distinguished:
The Taylor-Ford model: upstream management where production flows are driven by
commercial forecasts
The just-in-time model: downstream management where production is driven by demand.

The factors underlying this evolution of production management systems have been presented.
in the chapter 'Full Costs'.

A. THE MANAGEMENT OF THE PRODUCTION SYSTEM

The key problem that production management must solve is that of alignment between the
production and sales.

Upstream piloting
[Link] of the PBC method (Component Requirements Planning)
Based on the forecasts made on final demand, a planning of work and resources
raw materials and in man is established.
It is therefore upstream of production that decisions are made: what to produce? when
produce?

Scheduled request Master Plan Calculation of needs

It is noteworthy that all physical and informational flows are oriented in the same direction: towards
commercial forecasts towards the finished product.
 72

The MRP method (Management Resources Planning), translated by Resource Management of

Production or MRP (Material Requirements Planning) is a management technique for overseeing the
production from the upstream.

The method starts from a finished product and uses the bill of materials, the operating ranges, and the stocks.
current stocks and desired stocks to calculate, through matrix calculations, the dependent needs.
The MRP method is a comprehensive business model that allows, based on sales forecasts,
to plan the entire activity, to coordinate all functions, to reserve capacities,
to manage the stock and ensure its availability at the necessary times.

In use, the MRP method is all the more effective as demand is predictable, that the
products are subject to stabilized nomenclatures, that they are produced in batches and that components
Identical ones are used in the production of several different products.

[Link] elements of matrix calculus

A matrix of order (n× p) is a table of numbers called elements, made up of n rows and p
columns.

• Sum and difference of two matrices

The sum and difference of two matrices are only possible for matrices of the same order.
The sum A + B of two matrices A = (a ) n,
andp B = (b ) isn,defined
p by adding to each
element of A to the corresponding element of B. The difference is calculated according to the same principle.

⎛ 1 3⎞ ⎛ 0 2⎞
Let two matrices A = 0 8⎜ and B⎟ = 1 5 ⎜ ⎟
⎜ 2 5⎟ ⎜ 4 1⎟
⎝ ⎠ ⎝ ⎠

⎛ 1 + 03+ 2 ⎞ ⎛1 5 ⎞ ⎛ 1 − 03− 2 ⎞ ⎛ 1 1 ⎞
A + B = 0⎜+ 18+ 5 = ⎟1 ⎜
13 ⎟ A - B = 0⎜− 18− 5 =−1⎟ 3⎜ ⎟
⎜ 2+ 45+ 1 ⎟ ⎜6 6 ⎟ ⎜ 2− 45− 1 ⎟ ⎜ −2 4⎟
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠

• Product of two matrices

Product M× M1 of 2two matrices is not possiblethat if the number of columns of M is 1equal to the
number of lines of M. 2

The matrixobtained is a matrix whose number of rows is that of M and the 1number of
columns that of Mr.2

⎛ 1 4⎞ ⎛ x⎞
⎛ 0 7 2⎞ ⎜ ⎟
Let three matrices C = 2⎜ 5, D =
⎟ ⎜⎝ 9 8 4⎠ ⎟ andX=⎜ y⎟ .
⎜ 3 6⎟ ⎜⎝ z⎟⎠
⎝ ⎠

• C× X= Impossible (C = 2 columns and X = 3 rows)

• X× C= Impossible (X = 1 column and C = 3 rows)
• X× D = Impossible (X = 1 column and D = 2 rows)
⎛ x⎞
• D× X= ⎛ 0 7 2⎞ ⎜ ⎟ ⎛ 0x+ 7y+ 2z ⎞
⎜⎝ 9 8 4⎠ ⎟ × y⎜ ⎟ = ⎜
⎜⎝ z⎟⎠ ⎝ 9x+ 8y+ 4z ⎠⎟
 73

⎛ 1 4⎞
⎛ 0 7 2⎞ ⎜
• D× C= ⎜⎝ 9 8 4⎠ ⎟ × 2 5 ⎟ ⎛ (0× 1)+ (7× 2)+ (2× 3) (0× 4)+ (7× 5)+ (2× 6) ⎞
⎜ 3 6⎟ = ⎜ ⎟⎠
⎝ ⎠ ⎝ (9× + (8× 2)+ (4× 3) (9× 4)+ (8× 5) + (4× 6)
⎛ 20 47 ⎞
= ⎜
⎝ 37 100⎠⎟
⎛ 1 4⎞ ⎛ (1× 0) + (4× 9) (1× 7)+ (4× 8) (1× 2)+ (4× 4) ⎞
⎛ 0 7 2⎞
• C× D= ⎜ 2 5×⎟ ⎜⎝ 9 8 4⎠ ⎟ =(2⎜ × 0) + (5× 9) (2× 7)+ (5× 8) (2× 2) + (5× 4) ⎟
⎜ 3 6⎟ ⎜ ⎟
⎝ ⎠ ⎜⎝ (3 × 0) + (6× 9) (3× 7)+ (6× 8) (3× 2) + (6× 4) ⎟⎠

⎛ 36 39 18 ⎞
45⎜ 54 24 ⎟
⎜ ⎟
⎜⎝ 54 69 30 ⎟⎠

EXAMPLE1

The forecasted orders for the first quarter of year N for three products are as follows:
J F M
A 1 2 1
B 2 1 1
C 0 1 2

It does notythere are no bottlenecks. The nomenclatures are as follows:

[Link] ensembles by product (matrix M1)
A B C
E1 1 1 2 Assembly time: 3 months.
E2 2 0 1
E3 1 1 2

[Link] subsets by set (matrix M2)

E1 E2 E3
SE1 1 2 1 Machining lead time: 2 months
SE2 1 1 1
SE3 0 1 2

[Link] parts by sub-assembly (matrix M3)

SE1 SE2 SE3
P1 1 1 1 Machining time: 1 month
P2 1 1 2
P3 0 1 1
 74

[Link] materials (weight in kg) needed per piece (matrix M4)

P1 P2 P3
MP1 2 0 2 Delivery time: 1 month
MP2 1 1 1
MP3 1 1 0

Work to be done:plan the component needs.

Elements of response
One must start from the forecasted orders to work back up through successive matrix calculations to the
supplies of raw materials.

[Link] of needs in sets: M1× Forecasting order matrix:

A B C J F M O N D
E1 1 1 2 A 1 2 1 E1 3 5 6
E2 2 0 1 × B 2 1 1 = E2 2 5 4
E3 1 1 2 C 0 1 2 E3 3 5 6
Note: the order of the matrix is fundamental.

Interpretation of the obtained matrix:

To deliver 1 A, 2 B, and 0 C at the beginning of January, it is necessary to plan for 3 E1, 2 E2, and 3 E3 at the beginning of October.

2. Calculation of needs in sub-assemblies: M2× Matrix obtained previously

E1 E2 E3 O N D A S O
SE1 1 2 1 E1 3 5 6 SE110 20 20
SE2 1 1 1 × E2 2 5 4 = SE2 8 15 16
SE3 0 1 2 E3 3 5 6 SE3 8 15 16

Interpretation of the obtained matrix:

To have 3 E1, 2 E2, and 3 E3 at the beginning of October, it is necessary to plan for 10 SE1, 8 SE2, and 8 SE3 at the beginning.
August.

[Link] of parts requirements: M3× Matrix obtained previously

SE1 SE2 SE3 A S O J A S
P1 1 1 1 SE1 10 20 20 P1 26 50 52
P2 1 1 2 × SE2 8 15 16 = P2 34 65 68
P3 0 1 1 SE3 8 15 16 P3 16 30 32

Interpretation of the obtained matrix:

To have 10 SE1, 8 SE2, and 8 SE3 at the beginning of August, it is necessary to plan for 26 P1, 34 P2, and 16 P3 at the beginning.
July.

[Link] of raw material requirements: M4× Matrix obtained previously

P1 P2 P3 J A S J J A
MP1 2 0 2 P1 26 50 52 MP1 84 160 168
MP2 1 1 1 × P2 34 65 68 = MP2 76 145 152
MP3 1 1 0 P3 16 30 32 MP3 60 115 120

Interpretation of the obtained matrix:

 75

To have 26 P1, 34 P2 and 16 P3 at the beginning of July, it is necessary to plan for 84 MP1, 76 MP2 and 60 MP3.
early June.

2. Downstream control: see paragraph 3-5 of the handout

Just in time is more than a production management model: it is a philosophy that is rooted
in the total quality approach.

The concept of total quality corresponds to a global vision of the company, at all levels, in
all services and all functions. It is a management system.

B. PRODUCTION MANAGEMENT TECHNIQUES

Linear programming
Linear programming aims to solve an economic problem:
– Optimize (search for a maximum or minimum) an economic function of the form
linear,
Considering constraints (linear equations and/or inequalities).

EXAMPLE 2maximization problem involving two variables.

The company ALPHA manufactures, among other things, two products: the BET and the NUM.

The manufacturing of these products requires going through a workshop for which we have
following information for one month of activity:
Capacity expressed in Number of work units Number of units of work Variable cost of a
work units for a Bet product for a product Num work unit
Workshop 900 6 5 40 €
MOD 720 6 3 45 €

To make this activity profitable, the company must produce at least 80 products.
The maximum monthly request is for 100 Bet products and 150 Num products.
The unit selling prices are €770 for product Bet and €535 for product Num.

a. Canonical form
This is the most important step in the resolution. It involves expressing it in a form
mathematics a problem stated in a literary way. For this, it is necessary to:
Define the variables precisely.
Express the constraints.
Express the economic function.

Two errors are frequently encountered when it comes to maximizing a result:

The first is to confuse 'turnover' and 'result'.
The second is in the 'result' to remember: it is necessary to retain themargins on costs
unitary variablesto avoid experiencing fluctuations in results related to expense allocation
fixes).

Let it be:xthe
number of Bet products to be produced each month,
ythe number of Num products to be produced each month.
 76

Calculation of unit margins on variable costs

⎧ x≥ 0 Bet Num
⎪ y≥ 0
Workshop 240 200
⎪ 6 x+ 5 and≤ 900
⎪ MOD 270 135
⎨ 6 x+ 3 and≤ 720 Charges variables 510 335
⎪ x+ y≥ 80
⎪ x 2. Selling price 770 535
≤ 100
⎪ y≤ 150 2.1 Margin on variable costs 260 200
⎩
MAX F = 260x + 200y

The brace means that all constraints must be met.

This program is called linear because the constraints and the economic function are of the first degree.

b. Graphic resolution
Graphical resolution should be preferred for linear programs with two variables.

To define the domain of acceptable solutions, one must outline the constraints, then notice that
Each constraint divides the plane into two parts.
The study of the point (0, 0) allows us to define:

• The half-planewho respectsthe constraint. • The half-planewho does not respectthe constraint.
If the inequality is not satisfied for (0, 0), we
If the inequality is satisfied for (0, 0), the half-
shade this part that is not part of the
planning is part of the solutions domain
acceptable solutions domain.
acceptable.

When inequality is taken broadly, the points of

the right is part of the DSA
6x+ 5y≤ 900 x+ y≥ 80

y y
200 100
150
50
100
50 0 x
0 x
0 50 100
0 50 100 150 200

The domain of acceptable solutions (DAS) is the set of combinations (x, ywho respect the
constraints of the linear program to be solved.

To plot an economic function (ax + by), we must study atx+ by= k, by assigning a value to k
(generally 0).
a
This is a family of parallel lines with a slope coefficient of - x.
b
 77

STUDY of function 260x+ 200y

k=0 k = 26,000 k = 39,000
Function 260x + 200y = 0 260x + 200y = 26000 260x + 200y = 39000
Go through (0 ; 0) (0 ; 130) (0 ; 195)
and a couple (100; -130) (100; 0) (150 ; 0)
Value of F F=0 F = 26,000 F = 39,000

y
400
300
It should be noted that the higher the value k
200
increases (decreases), more the parallel of F
100 k = 39,000 is far (close) from the origin and more
0 x the value of the objective increases (decreases).
k = 26,000
-150 - 100 - 50 0 50
- 100
k=0
- 200

To search for the optimum, the following result will be accepted: if there is at least one optimal solution,he
yat a vertex of the D.S.A that corresponds to an optimal solution.

Two research methods are possible:

• Analytical method
MINIMIZATION
MAXIMISATION
The smaller the value of k, the better it is
The greater the value of k, the better the
result.
result.
The minimal solution is therefore the top by
The maximum solution is therefore the summit by
which passes the parallel of Fthe closest
which passes the parallel ofF the furthest
from the starting point(0 ; 0).
from the starting point(0 ; 0).

The analytical method therefore consists of outlining the constraints, defining the DSA, outlining the objective, and then doing
a translation (by moving a ruler for example) of the objective function.
 78

y
300

250

200

150
B C

100 A
D
F 50
E
G F
0 x
- 50 0 50 100 150
- 50

-100

The optimal solution (point D) is such that:

⎧ 6 x+ 5 y = 900
⎨ 6 x+ 3 y = 720 x= 75 y= 90
⎩
F = (75× 260) + (90× = 20037,500

The optimal program is therefore the monthly production of 75 Bet products and 90 Num products.
for a maximum contribution margin on variable costs equal to €37,500.

NOTE
The constraintx+ y≥ 80 is superfluous, meaning it can be removed without changing the solution.
about this problem.

• Enumerative method
The enumerative method consists of calculating the value of F at each of the vertices.
realm of acceptable solutions and to retain the optimal solution.
This method should be used in addition to the previous one, when the translation of the
economic function leaves a doubt between close peaks.
A B C D E F G
Coordinates (0 ; 80) (0 ; 150) (25 ; 150) (75 ; 90) (100 ; 40) (100 ; 0) (80 ; 0)
Value of F 16,000 30,000 36,500 34,000 26,00037,500 20,800

c. Simplex method
Graphical resolution is inapplicable beyond two variables. It is also necessary to resort to
another method: the simplex method, also known as the table method or method of
Dantzig.

This method, applicable regardless of the number of variables, will only be presented in this course when
for somemaximization problems where the constraints (other than those of positivity) are of
type≤ .
 79

EXAMPLE3 :maximization problem involving three variables.

The BETA company manufactures three models of furniture: classic, rustic, modern.
The standard unit production standards are summarized in the following table:
Model Model Model Capabilities
Classic Rustic Modern maximum
Wood 5 8 5 900
Labor 1 2 3 516
Finishing center 2 2 0 200
Margins on variable costs 1,000 960 1,200

The BETA company wishes to determine the quantities to produce in order to maximize its profit.

Canonical form of this program:

Let it be:x, name of classic models to produce,
ynumber of rustic models to produce,
zname of modern models to produce.
⎧ x≥ 0;y≥ 0;z≥ 0
⎪ 5x+ 8 y + 5z≤ 900
⎪
⎨ 1x+ 2 and+ 3z≤ 516
⎪ 2x+ 2 and+ 0z≤ 200
⎪ F= 1000 x+ 960 and+ 1200z(MAX)
⎩

The simplex method requires a conversion to standard form:inequalities are transformed into
equalitiesthanks to the introduction ofvariables of positive or null deviations, notées ei .
Theya slack variable for each constraint other than the non-negativity constraint.

STUDY of the constraint related to the wood factor:

Use of wood factor for Factor capacity
production levels x, y, z 5x + 8y + 5z + e1 = 900

Gap between the capacity and consumption of wood as a factor for production
dex,y,z. Thisgap allows equalitybetween the two members.

Canonical form: Standard form:

⎧ x≥ 0;y≥ 0;z≥ 0e≥ 0;e≥10;e≥ 02

⎧ x≥ 0;y≥ 0;z≥ 0 ⎪
3
⎪ 5x+ 8 y + 5z≤ 900 ⎪⎪
5x+ 8y+ 5z+ e1 = 900
⎪
⎨ 1x+ 2 y + 3z≤ 516 ⎨ 1x+ 2 y + 3z+ e2 = 516
⎪ 2x+ 2 y + 0z≤ 200 ⎪ 2x+ 2 y + 0z+ e3 = 200
⎪ F= 1000x+ 960y+ 1200z(MAX) ⎪
⎩ ⎪⎩ F= 1000x+ 960y+ 1200z(MAX)

NOTE
The objective function in canonical form remains unchanged and could be noted Max (1,000x + 960
y + 1 200z + 0 e1 + 0 e2 + 0 e3) and especially not Max (1 000x + 960y + 1 200z + e1 + e2 + e3)

For searching the optimal solution, the calculations are presented in tables using the
Gauss pivot method.
 80

Initial solution: first table

The first table shows the coefficients of the standard form.

Reading the table:
Outside Base x y z . . . B Variables Variables
In Base outside the base in the database
e1 5 8 5 1 0 0 900 x=0 e1= 900
e2 1 2 3 0 1 0 516 y=0 e2= 516
e3 2 2 0 0 0 1 200 z=0 e3200
F 1,000 960 1,200 0 0 0 0 F=0

WHATEVER THE PAINTING:

• The off-base variables are equal to zero.
• The value of the variables in the database is read from column B,
• The optimum is reached if all the coefficients in the last row are negative or zero.

Interpretation of the table:

This is the admissible starting solution that satisfies all constraints: produce nothing.
The production is therefore zero (x= 0; y= 0; z= 0) and the value of the objective function is equal to 0.
The available capacities of the factors are intact (thus e1= 900 means that there are 900 units remaining of
wood).
This solution can be improved since the coefficients in row F are not negative or zero.

Determination of the pivot

To improve the basic solution, we need to start producing.
The study of margins on variable costs indicates that it is better to start with the products.
(margin equal to 1,200, compared to 1,000 for x and 960 for y).

Given the constraints to be respected simultaneously, the maximum production of products.

constraint to 172:
Constraints to be respected: Maximum production of ten
5x + 8y + 5z ≤ 900 900 / 5 = 180 Division by 0 being impossible, let us set c
1x + 2y + 3z ≤ 516 516 / 3 = 172 which tends to 0.
2x + 2y + 0z ≤ 200 200 / c =∞ Division by 0 being impossible, let’s set c
which tends towards 0.

To meet the constraints, it is therefore only possible to produce 172z.

A resulting column (R) is added to the previous table to determine the pivot. This column
is to be presentedafter noticing that the optimum is not reachedand express the reasoning
which has just been presented.

DETERMINATION OF THE PIVOT

The variable that enters the base is the one with the highest positive coefficient in the last row.
grand.
2. The variable that exits the base is the one with the smallest positive resultant (R).
The pivot is located at the intersection of the entering variable and the leaving variable.
 81

Outside Base x y z . . . B R
In Base
L1 e1 5 8 5 1 0 0 900 900 / 5 = 180
L2 e2 1 2 3 0 1 0 516516 / 3 = 172 e2go out of
the base
L3 e3 2 2 0 0 0 1 200 200 /ε = ∞
L4 F 1,000,9601,200 0 0 0 0
center in the database

Second table:
The variable entering the base has taken the place of the one that has exited. It is necessary to divide the pivot row.
from the preceding table by the pivot: L'2 = L'p = L1 / 3
Off Base x y . . e2 . B
In Base

L'2 z 1/3 2/3 1 0 1/3 0 172

The other lines can then be determined:

L'1 = L1 - 5 L'p
L'3 = L3 - 0 L'p
L'4 = L4 - 1 200 L'p
Off Base x y . . e2 . B
In Base
The 1 e1 10 / 3 14/3 0 1 -5 / 3 0 40
L'2 z 1/3 2/3 1 0 1/3 0 172
L'3 e3 2 2 0 0 0 1 200
The 4 F 600 160 0 0 -400 0 - 206 400

Detail of L'1 and L'4:

10/3 = 5 - (5× 1/3) 600 = 1,000 - (1,200× 1/3
14/3 = 8 - (5× 2/3 160= 960 - (1 200× 2/3
0 = 5 - (5× 1) 0 = 1,200 - (1,200×
1 = 1 - (5× 0) 0 = 0 – (1 200× 0)
-5/3 = 0 - (5× 1/3) - 400 = 0 - (1 200× 1/3
0 = 0 - (5× 0) 0 = 0 - (1 200× 0)
40 = 900 - (5× 172) - 206 400 = 0 - (1 200× 172)

The out-of-base variables are null: x = y = e 2= 0. The base variables are: e1= 40, z = 172 and
e3= 200.
The production is therefore equal to 172 products.
The second constraint is saturated.2= 0), the available capacity of constraint 1 is 40
units (e1= 40) and the available capacity of constraint 3 is 200 units (e3= 200).
The objective is equal to €206,400.
This solution can be improved since the coefficients in row F are not negative or zero.
 82

⎧ 900− (5× 0+ 8× 0+ 5 × 172)= 40

⎪ 516− (1× 0+ 2× 0+ 3 × 172)= 0
It is possible to verify this solution: ⎨
200− (2× 0+ 2× 0+ 0× 172)= 200
⎪
⎪⎩ F= 1000× 0+ 960× 0+ 1200× 172= 206 400

Outside Base x y . . e2 . B R
In Base
The 1 e1 10 / 3 14/3 0 1 -5 / 3 0 40 12
L’2 z 1/3 2/3 1 0 1/3 0 172 516
The 3 e3 2 2 0 0 0 1 200 100
The 4 F 600 160 0 0 - 400 0 – 206 400
Value of Fto the nearest sign.
Third table:
The entry variable has taken the place of the one that has exited. It is necessary to divide the pivot row.
from the preceding table by the pivot: L''1 = L''p = L'1 / (10/3)
Off Base . y . e1 e2 . B
In Base
L'1 x 1 1.4 0 0.3 - 0.5 0 12

The other lines can then be determined:

L''2 = L'2 - (1/3) L''p
L’3 = L'3 - 2 L’’p
L''4 = L'4 - 600 L''p

Outside Base . y . e1 e2 . B
In Base
The 1 x 1 1.4 0 0.3 0.5 0 12
L’2 z 0 0.2 1 -0.1 0.5 0 168
L'3 e3 0 -0.8 0 -0.6 1 1 176
L''4 F 0 -680 0 -180 -100 0 -213 600

The off-base variables are null: y= e1= e2= 0. The base variables are:x= 12z= 168 and
e3= 176.
This is the optimal solution since all the coefficients in the last row (or marginal rates of
substitutions) are negative or zero.
The optimal solution is producing 12 classic models, 0 rustic models, and 168 models.
modern for a maximum contribution margin on variable costs of €213,600
The constraints related to labor and wood are saturated, and there is a remaining capacity of 176.
units for the finishing center.
⎧ 900− (5× 12+ 8× 0+ 5 × 168)= 0
⎪ 516− (1× 12+ 2× 0+ 3 × 168)= 0
– It is possible to verify this solution: ⎨ 200− (2× 12+ 2× 0+ 0× 168)= 176
⎪
⎪⎩ F= 1,000× 12+ 960× 0+ 1 200× 168= 213,600
 83

d. Special case: rare factor

EXAMPLE4rare factor
The company Braque manufactures three products A, B, C with the following characteristics:

A B C
Contribution margin 100 300 360
Workshop (Capacity 8,500 units) 2 up 4 you 6 uo
Maximum sales 1,000 1,000 500
Note: uo = work unit.

A≥ 0 ; B≥ 0; C≥ 0
⎧ A≤ 1000
⎪⎪ B≤ 1000
Canonical form: ⎨
C≤ 500
⎪
⎩⎪ 2 A+ 4 B+ 6 C≤ 8 5 0 0
MAX (100 A + 300 B + 360 C)
This is a 'rare factor' problem: the workshop is the only common resource that constrains the
production.
The problem can be solved by the simplex algorithm. However, it is faster in this case.
In particular, to classify products based on the margin generated per unit of work:
A B C
Contribution margin on variable costs (1) 100 300 360
Workshop (2) 2 up 4 you 6 uo
Margins per unit of work (1) / (2) 50 75 60
Product ranking 3 1 2

First of all, we need to produce the maximum amount of product B, which is B = 1,000.
The consumption of the rare factor is equal to: 4× 1,000 = 4,000 units of work.
The available capacity of the rare factor becomes equal to: 8,500 - 4,000 = 4,500 units of work.
In a second phase, it is necessary to produce the maximum amount of product C, taking into account the capacity.
available, either C = 500.
The consumption of the rare factor is equal to: 6× 500 = 3,000 units of work.
The available capacity of the scarce factor becomes equal to: 4,500 – 3,000 = 1,500 units of work.
Finally, we need to produce the maximum amount of product A, considering the available capacity.
A = 1,500 / 2 = 750.
The consumption of the rare factor is equal to: 2× 750 = 1,500 units of work.
The rare factor is then saturated.

The optimal combination is therefore: A = 750, B = 1,000 and C = 500.

The corresponding contribution margin on variable costs is equal to: 75,000 + 300,000 + 180,000 = 555,000 €.
It is possible to verify the solution using the simplex algorithm...

C. SCHEDULING PROBLEMS

The scheduling aims to:

To organize over time ('to arrange') a set of tasks, subject to constraints, which
contribute to achieving a goal,
To determine the best time to achieve the objective,
 84

To indicate the tasks that cannot be delayed without compromising the total duration of
project.

1. Main representations of a graph

Sagittal representation (from Latinarrow, "arrow")

C E A pathis a sequence of arcs whose terminal end of

A each arc (except for the last one) is the starting point of the next.

Thus A, B, D is a path while A, B, C is not.

B D A circuitis a path that closes in on itself: the end
the terminal of the last arc is the origin of the first.

Thus C, B, D is a circuit.

Representation in the form of a boolean matrix (that is to say composed of 1s and 0s)
The matrix presentation allows for the use of scheduling and transportation software.

Arrival summits
A B C D E
The number 1 signifies the existence of a relationship.
A 0 1 0 0 0
The matrix can be read in both directions:
Summits B 0 0 0 1 0
W Online: vertex C is followed by vertices B and E.
origins C 0 1 0 0 1 W In column: vertex E is preceded by vertices C and D.
D 0 0 1 0 1
E 0 0 0 0 0

Representation in the form of dictionaries

Dictionary of Precedents: Dictionary of the following:
Tops Previous Summits Following
X P(X) X S(X)
A / A B
B A;C B D
C D C B;E
D B D E
E D;C E /

2. Search for levels of a graph without a circuit

This step aims to make the sagittal representation of the graph clearer. It consists of
Sort by ascending level the peaks from left to right.
A B C

The level of a vertex is the number of edges. D

who separates it from the origin by the way the
longer. Level 0: A, D
Level 1: B
Level 2: C
 85

EXAMPLE5

The execution of a project requires the completion of 10 tasks, including the prerequisites and the
The durations are specified in the following table:

Tasks (X) Previous tasks P(X) Duration in weeks

A None 6
B A 10
C A 6
D None 3
E A, C 14
F A, C, E 5
G A, B, C, J 9
H D 15
I A, B, C, E, F, G, J 7
J A, C 7

The approach is iterative.

Step one: the vertices without predecessors are at level 0: Level 0 = A, D
2. Second step: the peaks classified at level 0 are crossed out in both columns.
Tasks (X) Previous tasks P(X)
A
B A
C A
D
E A, C
F A, C, E
G A, B, C, J
H D
I A, B, C, E, F, G, J
J A, C
Level 1 are the peaks whose column P(Xis completely barred: Level 1 = B, C, H

3. Third step: the procedure from step 2 is repeated to obtain the following levels.
Tasks (X) Previous tasks P(X)
A
B A
C A
D
E A,C
F A, C, E
G A, B, C, J E, J
H D
I A,B,C, E, F, G, J
J A,C

Tasks (X) Previous tasks P(X)

A
B A
C A
D Level 3 = F, G
 86

E A,C
F A,C,E
G A,B,C,J
H D
I A A ,B,C,E, F, G, J
J/ A,C

Tasks (X) Previous tasks P(X)

A
B A
C A
D Level 4 = I
E A,C
F A,C,E
G A
H D
I A,B,C,E,F,G,J
J A,C

3. Elimination of redundancies
The removal of redundancies should only be done if the dictionary and/or the matrix of vertices
"immediately preceding" (that is, adjacent) are not communicated.

The term 'anterior' (or previous) means before (that is, not necessarily adjacent).

An excerpt from the precedent dictionary of example 2 illustrates this notion:

Tasks Previous Tasks

The arcA, Eis redundant andmust not appear
C A A C E
on the sagittal representation.
E A, C

The sagittal representation of the graph thus requires the removal of constraints.
redundant to obtain the dictionary (or the matrix) of the "immediate" vertices
previous" (that is to say adjacent).

The approach consists of examining the background of previous tasks.

Tasks Previous tasks Background Immediately preceding tasks(8)
A / /
B A / (1) A
C A / Common task not to A
not to remember
D / /
E A, C A(2) C
F A, C, E A, C(3) E
G A, B, C, J A, C(4) B, J
H D / (5) D
I A, B, C, E, F, G, J A, C, E, B, J(6) F, G
J A, C A(7) C

(1)
Task B has the preceding task: - task A which has no preceding task so /
(2)
Task E has the preceding task: - task A which has no preceding task therefore /
 87

– task C which has predecessor A

(3)
Task F has as its predecessor: - task A which has no predecessor therefore /
– task C which has antecedent A
– task E which has antecedent A and C
(4)
Task G has the preceding task: - task A which has no preceding task therefore /
– task B which has predecessor A
the task E which has predecessor A
the task J which has predecessors A and C
(5)
Task H has the predecessor task: - task D which has no predecessor task therefore /
(6)
Task I has the preceding task: - task A which has no preceding task therefore /
task B which has antecedent A
– task C which has antecedent A
– task E which has antecedents A and C
Task F which has antecedents A, C, and E
– task G which has predecessors A, B, C, and J
the task J that has predecessors A and C
(7)
Task E has the predecessor task: - task A which has no predecessor task therefore /
– task C which has predecessor A
(8)
The column of immediately preceding (or adjacent) tasks is obtained by removing the
common tasks from the previous tasks column and the background column.

NOTE:
It is also possible to determine the levels after removing redundancies.
 88

Sagittal presentation of the graph without redundancies:

D H

A C E F I

J
B
G

Level 0 Level 1 Level 2 Level 3 Level 4

4. METRA Potential Method (MPM)

The main scheduling methods are:
The PERT method (Program Evaluation Research Task) developed in the late 1950s in
United States by NASA for the development of the Polaris rockets. This is the most
known for its origin.
- The MPM method (METRA Potential Method) developed in France during the same period by
the B. ROY team for the establishment of a nuclear power plant.

Only the MPM method will be presented.

a. Representation conventions of the MPM method

The vertices of the graph represent the tasks and the edges represent the succession constraints.
realized and are symbolized by squares. They are valued and correspond to the deadline from
from which the following task can begin.
Start as early as possible Start at the latest

T duration of X t T duration of Y
X Y

Interpretation: task X precedes task Y and must be completed before the start of Y

b. Graph plotting
The vertices are arranged by levels and in a way that limits crossings between the arcs, and the duration of
tasks are postponed on the arcs.
A FIN vertex, the culmination of tasks without subsequently finishing the graph.
 89

3 15
D H

6 6 14 5 7
A C E F I END
6
6
7 9
J
Legend
a b
B 10 G a = start at the earliest
X
b = start at the latest

NOTEDo not forget to link task H to the final task.

c. Determination of the critical path

The search for the critical path requires the prior calculation of the earliest dates and the latest dates.
delayed for each task.

Calculation of earliest start dates

The earliest start date is the date on which a task can begin.
0 to begin.
6 A task cannot start until the previous tasks are
X 6
not all finished.
Z As a result, when heya convergence towards a task, it is
0 It is necessary to remember the longest path.
4
Y The earliest dates are [Link] marking the peaks of the
left to right.

0 3 3 15
D H

0 6 6 6 12 14 26 5 31 7 38
A C E F I END
6
6
12
7 9
J
6 19 Legend
a b
B 10 G a = start at the earliest
X
b = start at the latest

NOTE:
Tasks A and D can start immediately, hence an earliest date equal to 0.

The project can therefore be completed at best 38 weeks after the start of the work.
 90

Calculation of deadlines at the latest

The start date at the latest is the deadline by which a task
15 5 20 can start without delaying the project.
The deadline for the start of a task must not be postponed.
U V
3 due to the total duration of the project.
As a result, when heya convergence towards a task, it is
20 - 5 = 15
19 necessary to remember the shortest path.
19 - 3 = 16
W The start dates at the latestare calculated starting from the end
of the project.

3 3 23 15 The end date at

0 20
earlier in the project
D H is postponed.

0 0 6 6 6 6 12 12 14 26 26 5 31 31 7 38 38
A C E F I END
6
6
12 15
7 9
J
6 12 19 22
B 10 G

Critical path: a b Legend

Let: A-C-E-F-I
X a = start as soon as possible
Verification: 6 + 6 + 14 + 15 + 7 = 38

The critical path is the longest path between level 0 and the end of the project.
It consists of all the critical tasks, that is to say, tasks whose completion cannot be delayed.
no delay.
A task is critical when its earliest start date is equal to its latest start date.
The length of the critical path is therefore equal to the sum of the durations of the critical tasks.

d. Calculation and interpretation of margins

TOTAL MARGIN FREE MARGIN

The total margin is the maximum delay that The total margin is the maximum delay that
can take on the completion of a task without can take the completion of a task without
delay the total duration of the project. questioning the earliest dates of
next tasks, and therefore without delaying the
total duration of the project.

tx Tx x dx y

X X Y

Total margin of X = Tx-x Total margin of X = ty– dx-x

NOTEThe margins of a critical task are therefore zero.

 91

Calculation of the free margin of a task in the event of convergence towards the task.

2 Free margin of task R:

18 22
Date at the latest Date at the latest Date at the latest
R S
early R early S early in T
2
18 22 25
Duration of R
25
M. L. = 22 - 2 - 18 = 2
T
M. L. = 25 - 2 - 18 = 5

Retain the minimum: free margin of R = 2

Calculation of total margins and free margins:

Dates at Dates to Margins Margins Detail of the calculation of free margins:
Tasks
earlier later totals free
A 0 0 0 0 6 - 6 - 0 = 0; 6 - 6 - 0 = 0
B 6 12 6 3 19 - 10 - 6 = 3
C 6 6 0 0 12 - 6 - 6 = 0
D 0 20 20 0 3-3-0=0
E 12 12 0 0 26 - 14 - 12 = 0
F 26 26 0 0 31 - 5 - 26 = 0
G 19 22 3 3 31 - 9 - 19 = 3
H 3 23 20 20 38 - 15 - 3 = 20
I 31 31 0 0 0
J 12 15 3 0 19 - 7 - 12 = 0
Note: the free float of a task is always less than or equal to its total float.

Interpretation of the margins of task J:

It is possible to have a delay of 3 weeks (total margin = 3) without changing the duration of the project.
On the other hand, if one wishes not to call into question the earliest start date of the tasks
following, a delay is impossible (free margin = 0).

VII. ESTIMATION OF A MEAN AND A FREQUENCY

The study of the characteristics of all elements of a population is often impossible to carry out.
due to constraints of cost and time.
This impossibility leads to studying a subset derived from the parent population: the sample.

Probabilistic sampling methods involve randomly selecting samples from

elements of the population and are the only ones to comply with statistical 1laws. The sampling of
sample elements can be performed:
– with release: the collected item is immediately returned to the parent population before collecting
the following. An element that can potentially be drawn multiple times, the draws are
independent and the sample is said to be non-exhaustive.

1 In practice, probabilistic methods are often neglected in favor of empirical methods (quota method, method
less expensive and less difficult to implement routes, etc.
 92

– without replacement: the sample is exhaustive, but the draws are not independent since the
the composition of the mother population is modified at each draw.
In the following, in order to apply the rules of probability calculation, the samples will be
supposed to be made with replacement, or be samples without replacement whose size is negligible
in relation to that of the population which is large or infinite (the sampling is then akin to
a draw with replacement.

Let it beamother populationΩ of size N, a random variableXfor which the hope

mathematicsm, the proportionpand the standard deviationσ are unknown.
This involves estimating, based on the characteristics calculated from a sample, the value of
characteristics of the mother population.
Mother population:Ω
Size: N Representative sample
Average: munknown Size: n
Proportion: punknown Hope: x(known)
Standard deviation:σ (unknown) Proportion: f (known)
Standard deviation:σ(known)

Induce the characteristics of the parent population

from the knowledge of the characteristics of a sample.

The larger the sample, the better the estimation.

Moreover, to reduce the risk of making a mistake, rather than giving a number called estimate
punctual, it is possible to give a "range" called confidence interval, in which
is the parameter to be estimated.

A. POINT ESTIMATIONS

Point estimation of a mean

Let m be the unknown mean of the variableXdefined on the parent population, and x the calculated mean
on a sample of size n. The mean x varying from one sample to another is an estimate
punctual of m:

The number x is a point estimate of m.

2. Point estimation of a proportion

Let p be the unknown proportion of the variableXdefined on the mother population, and f the proportion
calculated on a sample of size n. The proportion f varies from one sample to anothera
point estimation of p:

The number f is a point estimate of p.

3. Point estimation of a variance and a standard deviation

2 2
Let it beσ the variance andσ the unknown standard deviation of the variableXdefined on the mother population, andσthe
variance calculated on a sample of size n.
 93

2
Contrary to previous estimates,σis not a good estimate of the variance of the
2
mother population becauseσmeasures the dispersion around the sample mean x and not
around the mean m of the parent population.
2
Takeσwould underestimate the variance of the parent population, especially as the size n of
the sample is small.
2
On the other hand, the number s is a point estimate of the variance:

2 n 2 2
The number is = σis a point estimate ofσ
n− 1

n
The number s = s2= σis a point estimate ofσ
n− 1

NOTE:the key [σ A calculator

N–1 directly gives the point estimate ofσ.

Example
In order to better manage the credit requests of its clients, the director of a bank branch is making
a study regarding the duration of processing files. A random non-exhaustive sample of 30
processed files gave:
Duration in minutes [0, 10[ [10, 20[ [20, 30[ [30, 40[ [40, 50[ [50, 60[
Name 3 6 10 7 3 1

Calculate the average and standard deviation of the processing times of this sample.
[Link] point estimates of the mean m and the standard deviation.σ of the total population
processed files.
[Link] a point estimate of the proportion of requests from the total population that the
processing time is over 40 minutes.

Elements of response
1. The use of a calculator gives (take the class centers for calculations):
x= 26.33 minutes.
σ12.31 minutes.

2. A point estimate of the population mean m is 26.33 minutes.

30
A point estimate of the standard deviationσ the population is s = 12.31 12.52 minutes.
30− 1

3.A point estimate of the proportion p of requests from the total population whose duration of
The treatment lasts more than 40 minutes and 4 / 30 = 0.1333.

B. ESTIMATIONS BY CONFIDENCE INTERVALS

The objective is to determine a symmetric interval [a, b] that will contain the parameter.θ (readtheta) à
estimate with a probabilityα called threshold or confidence coefficient (the number 1 -α being called
threshold or risk coefficient) :
P{θ ∈ [a, b]}α
 94

NOTE:The estimation of a variance and the estimation of a standard deviation will not be studied.

1. Confidence interval estimation of a mean

Let m be the unknown average of the [Link] on the mother population, and Xnthe variable
random that associates with every sample of size n the mean of that sample.
Two cases must be considered, depending on whether the standard deviationσ from the random variableXis known or not.

a. The standard deviationσ of the mother population is known.

It will be accepted that after taking a sample of size n, one of the realizations of the interval of
trust I am:
⎡ α α ⎤ With tα = 1.96 forα 0.95
I =x⎢- t α ; x+tα ⎥
⎣ n & n⎦ tα = 2.575 forα = 0.99

REMARKS
- The values tα are obtained by setting t such that: p(-t < T < t) =α, i.e. 2π(t) - 1 =α
Thus, forα = 0.95, we pose 2π(t) - 1 = 0.95 and we are looking for t such thatπ(t) = 1.95 / 2 = 0.975. The
reading the table of the standard normal distribution (appendix of the previous series of the course by
correspondence) allows to obtain t = 1.96.
The average m of the population belongs to the interval calculated inα percent of cases, which means
also that it may not belong to the confidence interval in 100 -α % of cases.
If it belongs to the interval, the mean m can be located anywhere within the interval and has
no more reasons to be near the center.
The higher the confidence coefficientα increases, the greater the amplitude of the interval increases: which is
Gained in certainty is lost in precision.

b. The standard deviationσ the mother population is unknown.

1

This is the most common case.

If the sample size is large enough (greater than 30), 2 according to the theorem of the
central limit, it is possible to take s as an estimate ofσ to calculate the interval of
confidence

⎡ s s ⎤
If n≥30 I =x- t ⎢ α < m <x+tα ⎥
⎣ n n⎦

2. Confidence interval estimation of a proportion

It will be accepted that after taking a sample of size n, the value f can serve as an estimate
punctual of p (and of q = 1 - f), and thatathe achievements of the confidence interval I is :

⎡ f(1− f) f(1 Š f) ⎤
⎢ t
I = f− α ;f+tα ⎥
⎢⎣ n translatedText
⎥⎦

1Only the case of sufficiently large samples (n≥ will be presented.

2The condition n≥ 30 is essential. Indeed, if n < 30, even if X follows a normal distribution, the variable X does not follow a normal distribution, but
a Student's law.
 95

EXAMPLE(continued)

Knowing that the standard deviation of the processing time of requests from the total population is 13
minutes, provide a 95% and a 99% confidence interval for the mean.
[Link] standard deviation of the processing time for requests from the total population being unknown, provide a
95% and 99% confidence interval of the mean.
[Link] a 95% confidence interval for the proportion of requests of the total population
whose processing time is more than 40 minutes.

Response elements
The mean m of the total population belongs in 95% of cases to the interval:
⎡ 13 13 ⎤
⎢ 26,33 − 1,96× ; 26,33+ 1,96× ⎥ [21.68; 30.98]
⎣ 30 30 ⎦
The average m of the total population belongs in 99% of cases to the interval:
[26,33 – 2.575× 2,373 ; 26.33 + 2.575× 2,373] = [20,22 ; 32,44]

[Link] sample size is sufficiently large. The estimation of the standard deviation s = 12.52 can be
retention:
The average m of the total population belongs in 95% of cases to the interval:
⎡ 12.52 12.52 ⎤
⎢ 26,33 − 1,96× ; 26,33+ 1,96× ⎥ [21,85 ; 30,81]
⎣ 30 30 ⎦
The mean m of the total population belongs in 99% of cases to the interval:
⎡ 12.52 12.52 ⎤
⎢ 26,33 − 2,575× ; 26,33+ 2,575× ⎥ [20,44 ; 32,22]
⎣ 30 30 ⎦

3.A point estimate of the proportion is 4 / 30 = 0.1333.

The proportion p of requests from the total population whose processing time is more than 40
minutes belong in 95% of cases to the interval:
[0,1333 – 1,96× 0.06205; 0.1333 + 1.96× [0.0117 ; 0.2549]

VIII. CONTROL OF THE PAYROLL

The control of labor costs is a fundamental challenge for organizations:
Compensation is one of the factors ofmotivationThe compensation system must also
to allowstabilizerthe staff and ofto attractnew employees.
Wage costs represent, depending on the organization's activity,25 to 90% of
chargesThey have astrong impact on cash flow.

Personnel-related costs include:

– Compensation and various benefits: bonuses, overtime, profit-sharing, vehicle
function...
The corresponding social, fiscal, and parafiscal charges.
– The costs of temporary staff, the portion of subcontracting corresponding to expenses of
labor.
 96

The administrative costs of personnel management, training expenses, subsidies to

works council...

Thepayrollis a more restricted concept that coversthe actual cost of salariesverses

to a given population for a given period, usually the year.
The 'loaded' payroll includes all gross compensation (salaries
nets + employee contributions) and employer contributions (social, fiscal, and parafiscal)
corresponding.
Payroll "loaded" = gross salaries× (1 + charge rate)

The evolution of the payroll from one year to another results fromfour factors
who interact with one another:
General augmentations
1 salaries: collective measures applying to the whole of
personnel or to certain categories.
Individual augmentationssalaries.
Variations ofeffectivehiring, retirements, resignations, dismissals, deaths.
Variations ofstructurequalification, seniority.

The management of the payroll assumes that the organization knows how to anticipate and act on these factors.

A. THE FORECAST OF THE PAYROLL

The calculation of a projected payroll consists of evolving the existing situation at the end of
the year preceding the projection based on social policy (general increases and
individual salary levels) and anticipated changes in staffing levels

PERSONNEL EXITING

STABLE WORKFORCE

RECRUITMENT

Outgoing staff report Management System for Incoming Staff

MS effective
MS = + (including internal movements) + (including internal movements)
stable
until release from the date of hire

A second method of calculation is possible:

Staff entry MS
MS effective Salaries not paid to the staff
MS = – + (including internal movements)
present at 1/1 emerging
from the date of hire

EXAMPLE1

The information related to the payroll for the year N is as follows:

Categories Effective Average gross annual salary Average gross salary of December
Frames 8 43,384 € 3,650 €
Technicians 16 26,197 € 2,205 €
Employees 26 17,723 € 1,498 €

1Ilya a ratchet effect: downward rigidity of wages, except in special cases (to preserve employment).
 97

It has been planned to grant in year N+1 two general and uniform salary increases: the
first of 1% on April 1, the second of 1.5% on October 1.
An employee W whose gross salary for December is €1,800 is expected to be promoted to technician.
on June 1st at a gross salary of 2,000 €.
Table of planned retirements
Categories Departure date Average gross salary for December
Employee X February 28 N+1 2,250 €
Technician Y June 30, N+1 3,310 €
Employee Z August 31, N+1 2,170 €

Planned Recruitment Table

Categories Date of hiring
Gross hiring salary
Employee A March 1st N+1 1,130 €
Employee B September 1 N+1 1,130 €

Work to be done:
Calculate the forecasted payroll for N+1.
2. Calculate the relative change in the wage bill and briefly comment on the reasons for it.
variation.

Elements of response
[Link] calculation retained will be based on stable populations.
Preliminary calculation of projected headcount for N+1
Effective N Outputs N+1 Stable workforce N+1 Entries N+1 Staff N+1
Frames 8 0 8 0 8
Technicians 16 1 15 1 16
Employees 26 2+1=3 23 2 25

Calculation of the payroll for the stable workforce

To save time, it isfaster to work with indices(base 100 December N) rather
that on amounts in euros.
Indices of stable staff:
J F M A M J J A S O N D
Indices 100 100 100 101 101 101 101 101 101 102,515 102,515 102,515
Cumul 100 200 300 401 502 603 704 805 906 1,008,515 1,111.03 1 213,545

REMARKS
It should be noted that if several increases are applied,the effects are multiplicative.

These are percentages. Thus: 1,213.545 %

Calculation of the payroll of the stable workforce Amount
Frames (8× 3 650× 1 213,545 % 354 355.14
Technicians (16× -1 105) × 1,213,545 % 387,970.34
Employees (26× 1 498 – 2 250 – 2 170 – 1 800) × 1,213,545 % 397 169.01
 98

1,139,494.49

Payroll of the outgoing staff until their departure

Calculation of the payroll of outgoing staff Amount
Frames 0
Technicians 1× 3,310× 603 % 19,959.30
Employees 1× 2,250× 200 % + 1× 2 170× 805 % + 1× 1 800× 502 % 31,004.50
50,963.80

Payroll of the hired workforce

Indices of hired staff:
J F M A M J J A S O N D Cumul
A 100 101 101 101 101 101 101 102,515 102,515 102,515 1 013,545
B 100 101.5 101.5 101.5 404.50
W 100 100 100 100 101.5 101.5 101.5 704.50

Calculation of the payroll of the hired staff Amount

Employee A 1 130× 1,013.545% 11,453.06
Employee B 1 130× 404.50% 4,570.85
Technician W 2,000× 704.50 % 14,090.00
30,113.91

Projected payroll for N+1

Amount
Salary mass of the stable workforce 1,139,494.49
Payroll of the outgoing staff 50,963.80
Salary mass of the hired workforce 30,113.91
1,220,572.20

Stable workforce N+1 Outputs N+1 Entries N+1 MS N+1

Frames 8 354 355.14 0 0 0 0 8 354 355.14
Technicians 15 387,970.34 1 19,959.30 1 14,090.00 16 422,019.64
Employees 23 397 169.01 3 31,004.50 2 16,023.91 25 444 197.42
1,139,494.49 50,963.80 30,113.91 49 1,220,572.20

2. Calculation of the payroll of N:

Categories Effective Average gross annual salary Mass of N
Frames 8 43 384 347,072.00
Technicians 16 26 197 419,152.00
Employees 26 17 723 460,798.00
1,227,022.00

Absolute variation = 1,220,572.20 – 1,227,022.00 = –6,449.80

Relative variation = (1,220,572.20 - 1,227,022.00) / 1,227,022 = -0.5256 %.

Despite the salary increases of N+1, the total decreases due to:
at the retirement of two employees replaced by younger and lower-paid employees
(noria effect)
 99

– the replacement of a technician by an internal promotion who is paid less than the one who left
retirement (professional structural effect).

B. CONSEQUENCES OF THE SALARY POLICY ON THE PAYROLL

The interest of a predictive approach is to grasp the consequences of increases

wage trends on the evolution of the wage bill for the year and for the following years.

Impact of collective salary variations

In order to isolate the impact of compensation from other factors affecting the payroll mass,the study
is done with constant staff.

a. The level effect

The level effect measuresthe evolution of remunerationof an employee or a group of employees
between two given datesin general the closing dates of the financial years:
Month m of N+1 Salary of month m of N+1
Level effect= =
MS of the month m of N Salary for the month of m N

NOTES
It should be noted that if several increases are applied,the effects are multiplicative:
1.02515 = 1.01× 1,015
If the calculation had been made using the payroll, it would have been necessary to reason with constant staffing.
This effect is also called apparent effect becauseIt is independent of the dates on which they are.
applied increasesduring the year.
An employee is sensitive to the date of the raise: it is preferable to obtain a raise in
January rather than at the end of the year...

b. The mass effect

The mass effecttranslate the impact of time and express the real increaseof the evolution of the
remuneration of an employee or a group of employees over a year:
Annual MS with increase Annual MS N
Mass effect= =
Annual MS without salary increase January N× 12

c. The carryover effect

The carryover effect measuresthe impact of the increases that occurred over the course of a year on the
progression of the payroll for the following year:
Annual MS of N+1 Salary January N+1× 12
Carryover effect from year N to N+1= =
Annual MS N Annual MS N

This effect reflects the fact that staggered increases in a given year N will fully impact
the year N+1.
 100

d. Relations between the three effects

(1+ Level N effect
(1+ Carrying effect from N to N+1 =
(1+ Mass effect N)

LEVEL EFFECT Carried Forward Effect

MASS EFFECT
MS YEAR N MS YEAR N+1 MS YEAR N+2

J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D

e. It is possible to verify the following relation:

Variation of the payroll N+1 = Mass effect N+1× Carryover effect from N to N+1

EXAMPLE1 (continued)
1. Calculate the level effect between December N+1 and December N.
2. Calculate the mass effect of the category executives for the year N+1.
3. Calculate the mass effect of the employee category for the year N+1.
4. What can be concluded between the date of a raise and the mass effect?
5. What can we conclude if the level effect is equal to the mass effect?
6. Calculate the carryover effect of the executive category for the year N+2. Interpret the result obtained.
7. Check the relationship between the three effects calculated for the executive category.

Elements of response
Level effect:
With the stable staffing indices 102.515 / 100 = 1.02515 So be it + 2.515 %.

2. Mass effect of the managerial category

With the indices: 1 213,545 / (100× 1.0112875 So be + 1.12875%
The workforce being stable, the calculation can also be done based on the MS:
354 355.14 / (3 650× 12× 8) = 354 355.14 / 350 400 = 1.0112875 + 1.12875 %

[Link] effect of the employee category

It is not possible to retain the payroll since the workforce is not stable.
It is therefore necessary to remember the calculation with the indices: 1,213.545 / (100× 12) Let it be + 1.12875 %

[Link] mass effect depends on the date of the increases. For the same increase in level,
the later the increase, the weaker the mass effect.

5. If the mass effect and the level effect are equal, it means that ityhad only one increase on the 01
January.

[Link] effect
With the indices: (102,515× 12) / 1 213.545 = 1.013707 It is +1.37%
The workforce being stable, the calculation can also be done based on the MS:
3,650× 102,515 % = 3,741.7975
Carrying effect = (3,741.7975× 12× 8) / 354 355.14 = 1.013707 So + 1.37%
 101

Even if no increase is granted in N+2, the payroll for N+2 will increase by
1.37% compared to that of N+1 due to the measures decided in N+1.

[Link]: 1.02515 / 1.0112875 = 1.013707 It is +1.37 %

2. Individual variations in salaries

Individual variations lead to distinguishing the GVT effect and the noria effect.

a. The sliding, aging, technicality effect (GVT)

The GTV effect refers to structural factors other than the workforce.
– Sliding
The sliding represents the increase in [Link] to merit, that is to say to promotions
individuals not related to seniority.
– Aging
Aging isautomatic increasesalaries based on a grid
advancement related to theseniority.

– Technicality
Technicality is the increase in wages due to the advancement of technical qualifications, that is to say...
to say the access to ahigher qualification.
Note: it is often difficult to separate the sliding effect from the technicality effect.

b. The effect of noria

The reference to the noria, a hydraulic machine with buckets that brings up water, is a metaphor for
illustratethe replacement of older employees with younger ones of the same category but
generally lower paiddue to their lack of seniority and experience.
The noria effect is calculated based on the total number of outgoing employees replaced position by position by incoming employees.

EXAMPLE2

A. Graduates end N Average annual salary at the end of year N for entrants at the beginning of year N+1 Average annual salary
Frames 4 50,000 3 40,000
Employees 5 30,000 7 22,000

Calculate the effect of noria

[Link] employee earning €30,000 per year retires on March 31st of year N. He is replaced on April 1st of year N.
by a new hire at €20,000 per year.
Calculate the effect of noria.

[Link] employee earning €30,000 per year retires on March 31st of year N. He is replaced on July 1st of year N.
by a new hire at €20,000 per year.
Calculate the effect of noria

Elements of response
 102

[Link] noria effect is calculated based on all the [Link] position by position by
entrantstwo incoming employees and an unfilled managerial position do not affect the effect of noria. Their
compensation will appear in the effective effect.
Noria effect = 3× (40,000 - 50,000) + 5× (22,000 - 30,000) = -70,000 favorable

[Link] noria effect represents the savings that will be made on the annual payroll: the new one will be
paid 9 months based on an annual salary of €20,000 compared to €30,000 for the outgoing.

Noria effect: 9/12× (20,000 - 30,000) = -7,500 (fav.)

Salary of the outgoing
3/12× 30,000 Incoming salary
September 12× 20,000 = 15,000

[Link] noria effect is calculated from the replacement. The payroll cost savings from the 31
March 30 corresponds to an effective effect.

Noria effect: 9/12× (20,000 - 30,000) = -7,500 (fav.)

Effective effect
Salary of the outgoing
3/12× 30,000
3/12× 30,000 Entry salary
- 7,500 (fav) 6/12× 20,000 = 10,000

C. ANALYZE THE VARIATION OF THE WAGE BILL

The analysis of the variation in the wage bill highlights the change due to the wage effect, the variation
due to the overall workforce, and the variation due to the structure.

Variation in payroll

Effective effect
Wage effect Professional structure effect Noria effect
total

Slippage, aging, technicality

Sliding + technicality Aging

Structural effect
Noria effect
professional

The decomposition of the GVT effect is only possible if

seniority is given

Several analyses are possible. Only the most common one will be presented.
 103

EXAMPLE3

Year N Year N+1

Categories Effective Gross salary MS of N Effective Gross salary MS of N+1
average annual average annual
Frames 8 43 384 347,072.00 8 44 294 354 355.14
Technicians 16 26 197 419,152.00 15 25,473.40 382 101.04
Employees 26 17,723 460,798.00 22 17,372.20 382 188.42
50 24,540.44 1,227,022.00 45 24,858.77 1,118,644.60

Deviation on payroll: 1,118,644.60 - 1,227,022 = -108,377.40 € (favorable)

-108,377.40 / 1,227,022 = -8.83%

This difference is due to a variation:

staff numbers
salaries
from the professional structure
It is not possible to calculate the effect of noria since the seniority structure is not
communication for the categories. Only the GVT effect can be calculated, without being able to distinguish.

2. Difference on total workforce: (En+1 - En) × S

As its name suggests, this gap involves comparing 'total numbers'. By convention, a
The gap on quantity will be analyzed at a constant average salary to avoid having a factor simultaneously.
quantity and a price factor.
Difference on total workforce: (45 – 50)× 24,540.44 = -122,702.20 (favorable)
45 / 50 = 0.9 - 10% favorable

Most works present the calculation of this gap using an intermediate budget. This
the approach is based on the development of the formula:
Difference on total workforce: (En+1 E) × Sn
En+1 × SnEn× Sn
MSn+1 at constant overall salary – MSn

MSn+1 at constant overall salary MSn

Effective Sn Amount Effective Sn Amount

Total general 45 24,540.44 1,104,319.80 50 24,540.44 1,227,022

[Link] structure gap

It is about highlighting the impact of the professional structure, that is to say the impact of
the number of staff in the different categories, on the expected average salary.

To calculate this gap, three approaches are possible.

 104

– Compare the N+1 workforce to the N+1 workforce with a constant professional structure. To avoid having a
salary effect and a structural effect, one must reason with the constant salaries of the categories.
N+1 structure effectiveness
Effective N+1 SC Gap
constant professional
Frames 8 45× (8 / 50) = 7.2 43 384 34,707.20 Unfavorable
Technicians 15 45× (16 /50) = 14.4 26 197 15,718.20 Unfavorable
Employees 22 45× (26 /50) = 23.4 17 723 – 24,812.20 Favorable
45 Total 25,613.20 Unfavorable

There is a shift from the employee category to the better categories of management and technicians.
paid.
The gap is generally unfavorable, the expected average salary is higher with the N+1 structure.
with the structure of N, as shown by the following approach.

A second approach is possible to highlight the influence of structural change.

professional on the expected average salary:
− Salary N at headcount
Effective N+1 S Cn N+1
Frames 8 43 384 347,072.00
Technicians 15 26 197 392,955.00
Employees 22 17,723 389,906.00
45 25 109.6222 1,129,933.00

With the structure of N, the expected average salary is €24,540.44, whereas with the new ...
the composition of the workforce, the expected average salary is €25,109.62.
Professional structure gap = 45 (25 109.6222 - 24 540.44) = 25 613.20 Unfavorable

The third approach, with a medium budget, is based on the development of calculations of
the first approach:
⎡ 8 ⎤ ⎡ 16 ⎤ ⎡ 26 ⎤
⎢ 8 × 43 384− 4× 50 × 43,384+⎥ 15⎢ ×⎥ 26⎢ 197− 45× × 26 197+ 22× 17 723− 45×
50 50
× 17,723 ⎥
⎣ ⎦ ⎣⎦ ⎣ ⎦
⎡ 8 × 43 384+ 16× 26 197+ 26× 17 723+ ⎤
⎡⎣ 8 × 43 384+ 15× 26 197+ 22× 17,723−⎤⎦45× ⎢ ⎥
⎣ 50 ⎦
MSn
[8× 43 384 + 15 × 26 197 + 22× 17 723 - 45×
En

−
Σ (ECn+1 × SCn) - En+1 × S n
Σ (ECn+1 × SCn+1 at a constant overall salary
MSnN+1 structure - MSn+1 at constant overall salary
[8× 43 384 + 15 × 48 197× 17,723 - 1,104,319.80
25,613.20
 105

[Link] on nominal rate

This is about comparing the salary of N+1 to the salary of N for each type of workforce of N+1.
According to the convention, this price difference will be calculated with the workforce N+1:

Effective N+1 Salary N+1 Salary N Gap

Frames 8 44,294.00 43 384 7,283.14 Unfavorable
Technicians 15 25,473.40 26 197 -10,853.96 Favorable
Employees 22 17,372.20 17 723 - 7,717.58 Favorable
- 11,288.40 Favorable

The development of the calculations gives:Σ (ECn+1 × SCn+1) - (ECn+1 × SC)

MSn+1MSnN+1 structure

[Link]: the sum of the deviations must allow for the recovery of the total deviation.
Deviation from total workforce -122,702.20 Favorable
Professional structure gap 25,613.20 Unfavorable
Difference on nominal rate - 11,288.40 Favorable
Total - 108,377.40 Favorable
 [Link] can be made with intermediate fictitious masses, as previously stated:

MS n with structure MSn+1 to salary

MSn+1 MSn
N+1 global constant
1,118,644.60 1,129,933.00 1,104,319.80 1,227,022.00
Gap on rate Gap on structure Deviation on workforce
professional total
- 11,288.40 25,613.20 - 122,702.20

MSn+1 MS n with N+1 structure MSn+1 at constant overall salary MS

Effective Salary Amount Effective Salary N Amount Effective − Amount Effective Salary N Amount
N+1 N+1 N+1 N+1 Sn N
Frames 8 44,294.00 8 43 384 347 072 8 43 384 347,072.00
Technicians 15 25,473.40 15 26 197 392 955 16 26 197 419,152.00
Employees 22 17,372.20 22 17,723 389 906 26 17,723 460,798.00
45 24,858.77 45 25,109.62 1,129,933.00 45 24,540.44 1,104,319.80 50 24,540.44 1,227,022.00

The payroll has decreased by – 108,377.40 -8.8325% be multiplied by 0.91167

Nominal wages decreased from 1,129,933 to 1,118,644 -0.9965% be multiplied by 0.99003
The workforce has decreased by 5. minus 10% be multiplied by 0.9

[Link] effect = structural factors other than the workforce.

Percentage variations being multiplicative:
0.91167 = 0.99003× 0.9× GVT effect
GVT effect = 1.02316 let it be + 2.316

It is not possible to calculate the effect of noria.

IX. CORRECTED EXERCISES

Exercise 1
The Alpha company would like to regularly procure raw materials to avoid any
stock shortages and excessive price fluctuations. For the coming year (a year of 360 days, for
simplifying), the projected production would be 2,800 productsX9,000 productsYand 2 150
productsZ.
The launch cost has been estimated at €45 per order and the inventory holding rate at 10%.
[Link] the optimal supply program. Deduce the total annual management cost.
this stock.
[Link] company would also like to know the impact of setting a safety stock of 3.
000 kg on all the previous parameters.
3. If the supply lead time is set to 30 days, what would the alert stock be?
It is possible to admit that the annual sales of productsX, YandZare variables
independent variables that follow a normal law with the following parameters:
X follows the law N(3000; 200)
Y follows the law N(9,000; 100)
Z follows the law N(2000; 50)
[Link] the law and parameters of total consumption.
[Link] service level would a safety stock of 3,000 kg correspond to?
[Link] what level should the safety stock be set in order to limit the stockout rate to 5%?

Appendix: consumption of raw materials by product.

X Y Z
Unit consumption of raw materials 10 kg 15 kg 15 kg

The purchase cost of one kilogram of raw material is equal to €1.5.

Elements of response
1. The annual consumption = 10× 2 800 + 15 × 9,000 + 15× 2,150 = 195,250 kg.
195,250
cl= 45
cs equals 1.5× 10% = €0.15 per year (therefore P = 1)

Q
CG = 45 n + × 0.15× 1
2

195 250
CG(Q) = 45 × + 0.075 Q
Q

8,786,250
CG'(Q) = - + 0.075
Q2

107
Cancels for Q* = 10,823.58 kg
n* = 195 250 / 10 823.58 = 18.04 orders
T* = 360 / 18.04 = 20 days
CG (10 823.58) = 811.77 + 811.77 = 1,623.54 €

NOTEIt is possible to round n* to 18 orders, hence Q* = 195,250 / 18 = 10,847.22

Theywould have 17 orders of 10,847 and a last one of 10,851.

[Link] function would be

⎛Q ⎞
CG = 45 n + ⎜ + 3,000⎟× 0.15× 1
⎝ 2 ⎠
195 250
CG(Q) = 45 × + 0.075 Q + 450
Q

It does notywould not impact the optimal supply program because it is a

constant whose derivative is zero.
On the other hand, the cost of managing the inventory would increase by 450 to reach €2,073.54.

[Link] the supply lead time (30 days) is greater than the supply period (20
days), heyan order that will be delivered.
195 250
SA = × 30 + 3,000 - 10,823.58 = 8,447.24 kg
360

[Link] consumption (QT = 10 X + 15 Y + 15 Z) follows a normal distribution since it is the sum

of normal random variables.
E(QT) = E(10X) + E(15Y) + E(15Z)
10 E(X) + 15 E(Y) + 15 E(Z)
10× 3 000 + 15 × 9,000 + 15× 2,000 = 195,000

As the random variables are independent:

V(QT) = V(10X) + V(15Y) + V(15Z)
V(QT) = 102V(X) + 152V(Y) + 152V(Z)
V(QT) = 102× 2002+ 152× 1002+ 152× 5026,812,500
σ= 6 812 500≈ 2,610

QT follows the law N(195,000; 2,610)

[Link] service rate TS is the probability that demand is met given the stock of
security of 3,000 kg:
TS = p(QT < 195,000 + 3,000) = p(QT < 198,000)
⎛ 198,000− 195,000 ⎞
TS = p ⎜T< ⎟⎠ = p(T < 1.15) = 0.8749 or 87.49%
⎝ 2,610

108
[Link] safety stock S is such that:
p(QT > 195,000 + S) = 0.05
1 - p(QT < 195,000 + S) = 0.05
p(QT < 195,000 + S) = 0.95
⎛ (195 000+ S) − 195,000 ⎞
< p ⎜T < ⎟⎠ = 0.95
⎝ 2,610
⎛ ⎞ S
p ⎜ T< = 0.95 let's place t =
⎝ 2,610 ⎟⎠ 2,610
S
π(t) = 0.95 t = 1.645 from where = 1,645 et S = 4 293,45 kg
2 610

Exercise 2
The company M is asking you to analyze the inventory management problem of a new material.
first M1 necessary for its production.
Working hypotheses:
The production of the new product is regular and takes place every day.
The consumption of raw material M1 is regular and a necessary quantity is planned.
5,760 units for a reference period of 360 days.
The supplier's delivery must be made when the available stock becomes zero, none
safety stock is not necessary.
The cost of launching each order is equal to 224 €, this amount being independent of the
ordered quantity and total purchase price.
The cost of storage is directly proportional to the purchase price of the material and it represents
0.525 € per unit stored and per month of storage when the unit purchase price is 75 €.
The volume ordered from the supplier must be constant.
No delay is possible in the supplier's delivery.
No stockouts can be accepted.

Supplier conditions:
The gross selling price is €75 per unit. A discount is granted or not depending on the volume.x
ordered:
Ordered volume Unit price
x≤ 1,200 no discount
1 200 <x≤ 2,400 2% discount
2,400 < x≤ 4,800 2.5% discount
x > 4,800 4% discount

NOTE
The supplier has committed to deliver on the scheduled date, guaranteeing that he will notywould have neither advance nor delay
in the delivery.

Express the total procurement cost of raw material M1 for the period of 360 days.
and in the event that the supplier does not grant a discount, depending on the constant volume of each
order.
[Link] the optimal volume of each order and deduce the minimal total purchase cost.
the M1 material in the case where the supplier does not provide a discount.

109
 [Link] order to study the interest of the supplier's commercial proposals regarding the prices.
degressive:
[Link] the (integer) order quantities for which the pricing conditions are distinct.
the rate without discount.
[Link] the optimal number of orders.

Elements of response
Total procurement cost: CA
Q
CA = 5 760 × 75 + 224 n + × 0.525× 12
2
5760 1290 000
CA (Q) = 432,000 + 224 + 3.15 Q = 432,000 + + 3.15 Q
Q Q

1290 000
[Link]’ (Q) = - + 3.15
Q2
Cancels for Q* = 640
n* = 5,760 / 640 = 9 orders
T* = 360 / 9 = 40 days
CA (640) = 432,000 + 2,016 + 2,016 = 436,032 €

[Link] tiered pricing only applies from 1,200 units. So if n is greater than 5 orders,
he does notyno discount and in this case Q* = 640 and n* = 9 orders. It is therefore possible to limit
the study of the function for n = 1, 2, 3, and 4.

The storage cost is proportional to the price. For €75, it is equal to €0.525. It is such that:
0.525 = ts× 75
Q
It therefore represents 0.7% of the price p and CA = 5,760.× p + 224 n + × (0.7% p)× 12
2
Purchase price Cost of Cost of
n Q = 5,760 / n CA
5,760 p launch storage
1 5,760 5 760× 72 224 17,418.24 432,362.24
2 2,880 5 760× 73,125 448 8,845.20 430,493.20
3 1,920 5,760× 73.5 448 5,927.04 429,735.04
4 1,440 5 760× 73.5 896 4,445.28 428,701.28
(1) 5,760× 75
9 640 2 106 2016 436,032.00
(1)
Last line unnecessary but presented for reference

The optimal solution is therefore to place 4 orders of 1,440 units each, which is
Order every 90 days.

110
Exercise 3
Company M plans to sell 1,600 [Link] the coming year.
The cost of launching an order is €78
The storage cost of an item is €0.65 per month,
The cost of shortage is estimated at €2.6 per month per item.

1. Calculate the optimal supply program and stock level.

2. Verify the solution from the Wilson model without shortage.

Elements of response
1. Study of the function

1,600 S S Q− S Q-S
CG (S, Q) = 78× + × 0.65× × 12 + × 2.6× × 12
Q 2 Q 2 Q

124,800 S2 (Q-S)2
CG (S, Q) = +3.9 + 15.6
Q Q Q

124,800 S2 2
(Q-2SQ+S) 2
CG (S, Q) = + 3.9 + 15.6
Q Q Q

124,800 S2 S2
CG (S, Q) = + 3.9 + 15.6 Q - 31.2 S + 15.6
Q Q Q

124,800 S2
CG (S, Q) = + 19.5 + 15.6 Q - 31.2 S
Q Q

S
• CGS(S, Q) = 39 - 31.2
Q
S
The derivative is zero for = 31.2 / 39 = 0.8
Q

124800 S2
• CG'Q(S, Q) = - - 19.5 + 15.6
Q2 Q2

124800 124800
CGQ(S, Q) = - 19.5 (0.8)2+ 15.6 = - + 3.12
2
Q Q2

The derivative is zero at Q2= 40,000 let for Q* = 200

S* = 0.8× 200 = 160
n* = 1 600 / 200 = 8 orders
T* = 360 / 8 = 45 days
Ts= 45× 0.8 = 36
Tp= 45× 0.2 = 9 days
CG (160, 200) = 1,248 €

111
2. Wilson model without shortage
1,600 Q 124,800
CG (Q) = 78× + × 0.65× 12 = + 3.9 Q
Q 2 Q
124,800
CG’ (Q) = - + 3.9
Q
The derivative is zero for Q2= 32,000 let it be for Q* = 178.88
Shortage rate = 2.6 / 3.25 = 0.8

Wilson Model Model with shortage

1
Q* = 178,88 × = 200
Q≈ 178.88 0.8

n≈ 8.94 n* = 8.94× 0.8 = 8 orders

1
T* = 40.25× 45 days
T≈ 40.25 0.8

CG = 1,395.30 € CG = 1,248 €

Exercise 4
The company STOCAL manufactures a product P from a material M. The consumption
Forecasts of M for the months of year N+1 are as follows (in kilograms):
janvier : 240 July: 250
février : 250 August 0 (paid leave)
mars : 240 September: 250
April 260 October 400
I 320 novembre : 530
June 200 December: 340
The stock of material M at the end of year N is 320 kilograms. The delivery time is
10 days, and STOCAL wants to always hold a minimum safety stock of 40 kilograms.
The purchase price is 16.40 euros per kilogram, the variable cost of placing an order is
90 euros and holding the stock incurs a financial cost of 10% of the average stock value.
Determine the optimal quantity to order and establish the purchasing budget using the
principle of constant quantity.
To simplify, we will reason in months of 30 days.

Elements of response
The total procurement cost is written as:
Fixed costs + Transaction costs + Holding costs
⎛ 3280 ⎞ ⎡⎛ X ⎞ 10 ⎤
CF = FF +⎜ × 90⎟ + ⎢ ⎜ + 40×
⎟⎠ 16.40× ⎥
⎝ X ⎠ ⎣ ⎝ 2 100 ⎦
295 200
= FF + + 0.82X + 65.6
X

112
The optimum is reached if:
295 200
CT'(X) = 0− + 0.82 + 0 = 0
X2
295 200
from where : 0.82
X2
295 200
X2=
0.82

295 200
X= = 600
0.82

The optimal order quantity according to the Wilson model is 600 kilograms.

113
 Purchasing Program:

Cumulative outputs Initial stock

Outputs Projected date Estimated date
Month + + Necessary purchases?
forecasted of delivery of order
Safety stock Cumulative purchases

January 240 240 + 40 = 280 320 320 > 280 → NO

February 250 280 + 250 = 530 320 530 > 320 → YESFebruary 4 January 24 (2)
Mars 240 770 = 770 320 + 600 = 920 920 > 770 → NO -
April 260 770 + 260 = 1,030 920 1 030 > 920 → YESApril 17 (3) April 7
May 320 1,030 + 320 = 1,350 920 + 600 = 1,520 1,520 > 1,350 → NO -
June 200 1,350 + 200 = 1,550 1,520 1,550 > 1,520 → YESJune 25 June 15
July 250 1 550 + 250 = 1 800 1,520 + 600 = 2,120 → NO - -
August 0 1 800 + 0 = 1 800 2 120 2 120 > 1 800 → NO - -
September 250 1 800 + 250 = 2 050 2 120 2 120 > 2050 → NO - -
October 400 2,050 + 400 = 2,450 2,120 2,450 > 2,120 → YESOctober 5 September 25
November 530 2,450 + 530 = 2,980 2,120 + 600 = 2,720 → YESNovember 15 November 5
December 340 2980 + 340 = 3,320 2,720 + 600 = 3,320 3,320 = 3,320 → YESDecember 30 December 20

114
Calculation by linear interpolation of the delivery date:
In February, the curve "Cumulative Outputs + Safety Stock" rises from 280 to 530. It will reach
so 320 at the datex1as such:

x1 320± 280 40× 30

= ⇒ x1=
30 530± 280 250

x1≈
from where: 4

Graphically:

We anticipate the order date 10 days in advance (delivery time).

x2 920− 770
(3) In April: = → x217
30 1 0 3 0 − 770

x3 1520 − 1 3 5 0
(4) In June: = → x3= 25
30 1550 − 1 3 5 0

x4 2 1 2 0 − 2,050
In October = → x4= 5
30 2,450− 2,050

116
 x5 2,720− 2,450
In November: = → x5= 15
30 2,980− 2,450

x6 3 3 2 0 − 2,980
(7) In December: = → x6= 30
30 3 3 2 0 − 2,980

We will therefore have a budget of 600.× 16.40 = 9,840 euros for the months of February, April, June,
October, November, December and nothing for the other months, especially in summer, due to seasonality.
of the activity.

Exercise 5
A company manufactures and sells, among other things, products X and Y. The variable elements of
the costs are as follows:
Product X Product Y
Unit weight 11.8 kg 2.5 kg
Raw materials:
MP A 12 kg at 1 € = 12 € 3 kg at 1 € = 3 €
MP B 0.5 kg at 2 € = 1 € 0.1 kg at 2 F = 0.2 €
Direct labor 0.5 hours at €60 = €30 5 minutes at 60 €/h = 5 €
Production centers:
Center 1 3 units at 1 € = 3 € 0.5 units at 1 € = 0.5 €
Center 2 8 units at 2 € = 16 € 1 up to 2 € = 2 €
Distribution center 13 € 2.3 €

Given the supply difficulties, the company will not be able to procure more than 24,000.
tons of MP A. In contrast, MP B is available in unlimited quantity.
Considering the staff employed in the company, the total number of hours of labor
directly cannot exceed 700,000 hours.
Center 1 will only be able to provide 4,000,000 units of work (uw) and center 2 only 10,000,000.
work units.
Given the market, the company plans to produce at least 3,540 tons of productX and
-12,800 tonnes of product Y.
The company can sell them at a price of €105 excluding tax for product X and €23 excluding tax for product Y.

You need to help the team make a decision regarding the quantities to be manufactured of
each of these two products to maximize the result.

1. Present the corresponding linear program in canonical form (the reasoning will be done)
in the unit used to define the constraint: ton and hour).

2. Provide a graphical solution.

Elements of response
1.Définition des variables :x= nombre produitsxà fabriquer,
y= number of products to be manufactured.

• Constraint of the MP A expressed in tons: 0.012x + 0.003y≤ 24,000

Let: 4x+y≤ 8,000,000

117
• Labor constraint: 0.5x + 1/12y≤ 700,000
Let: 6x + y≤ 8,400,000
• Constraint of center 1: 3x + 0.5y≤ 4,000,000
Let: 6x+y≤ 8,000,000
• Constraint of center 2: 8x + 1y≤ 10,000,000
• Commercial constraints in tons: 0.0118x≥ 3,540 and 0.0025y≥ 12,820
Let :x≥ 300,000≥ 5,120,000
• Calculation of margins on unit variable costs (MVC):
x y
Selling price 105 23
Charges variables 75 13
Economic function: MAX F = 30x + 10y
MCV 30 10
x≥ 0 ;y≥ 0
⎧ 4 x+ 1 and≤ 8,000,000
⎪ 6 x+ 1 and≤ 8,400,000
⎪⎪ 6 x+ 1 and≤ 8,000,000
Canonical form: ⎨ 8 x+ 1 and≤ 10 million
⎪
⎪ x ≥ 300,000
⎪⎩ y≥ 5120 000

MAX F = 30x+ 10y

Graphical resolution:

y The margin on variable costs is

maximum at the summit A of which the
10,000,000
the coordinates are such that:
x = 300,000
8,000,000 6x + y = 8,000,000
Let it bex= 300,000 andy= 6,200,000

6,000,000 A
We need to produce 300,000 productsx
and 6,200,000 productsyfor a margin
4,000,000 on maximum variable costs equal to:
D. S. A F = 300,000× 30 + 6 200 000× 10
2,000,000 71,000,000

0 x
- 500,000 0 500 000 1 000 000 1 500 000 2 000 000

20,000,000

Exercise 6
Company V is engaged in the extraction and distribution of quarry materials.
She must ensure, for road works, the supply of gravel in various forms to the Bridges & Roads.
calibers.

118
A contract concerning the following quantities:
Gravel grade 1: 13,500 tons
Gravel caliber 2: 11,200 tons,
Gravel grade 3: 5,000 tons.
was awarded for a total billing price.
The Company operates two quarries P1and P2leased to a civil society that receives a fee by
ton of stone extracted. This is as follows:
- For P119.40 € per ton
For P220 € per ton.
After extraction, the stone is crushed. The resulting gravel is sorted by size.
Each ton of stone provides the following quantities of gravel (expressed in tons):
Pierre de P1:
WGravel grade 1: 0.36 tons,
WGravel grade 2: 0.40 ton,
WGravel grade 3: 0.16 tons.
Pierre de P2:
WGravel grade 1: 0.45 tons,
WGravel grade 2: 0.20 tons,
WGravel grade 3: 0.10 ton.
(The supplement to a ton represents sand, currently considered waste without value)
merchant)
Management wishes to define its stone extraction program from P.1and P2in order to minimize
the cost of royalties to civil society.
1. Present the corresponding linear program in canonical form.
2. Provide a graphical solution.
[Link] program optimization lead to producing gravel in excess compared to
awarded tonnages? Justify the answer.

Elements of response
1. Definition of variables:xnumber of tonnes of stone P1,
ynumber of tons of stone P2.

Canonical form:
x≥ 0 ; y≥ 0
⎧ 0.36 x+ 0.45 and≥ 13500
⎪
⎨ 0.40 x+ 0.20 and≥ 11200
⎪⎩ 0.16 x+ 0,10 and≥ 5 0 0 0
MIN F = 19.4x+ 20y

So let it be after simplification:

x≥ 0 ; y≥ 0
⎧ 4 x+ 5 and≥ 150,000
⎪
⎨ 2 x+ 1 and≥ 56,000
⎪⎩ 8 x+ 5 and≥ 250,000
MIN F = 19.4x+ 20y

119
 [Link] resolution:

y The fee is minimal at point C, which

are the coordinates only:
60,000
A ⎧ 4x+ 5y= 150,000
⎨ 8x+ 5 years= 250,000
50,000 ⎩
D.S.A Let it bex= 25,000 andy10,000
40,000 It is necessary to produce 25,000 tons of stone.
P1 and 10,000 tons of stones P2 for a
30,000 B minimum royalty equal to:
F = 25 000 × 19.4 + 10,000× 20 = 685,000 €
20,000

10,000 C
D
0 x
- 10,000 0 10,000 20,000 30,000 40,000
- 10,000

[Link] 25,000 andyadding 10,000 to each inequality of the first system allows
to study whether itywhether or not there is a production surplus:
Constraint 1: 0.36× 25,000 + 0.45× 10,000 = 13,500 so no surplus.
Constraint 2: 0.40× 25,000 + 0.20× 10,000 = 12,000 12,000 - 11,200 = 800 t
Constraint 3: 0.16× 25,000 + 0.10× 10,000 = 5,000 so no surplus.

Exercise 7
The company D manufactures stainless steel parts for hardware companies. These parts are
three types: A, B, C. They are produced in batches of 50 in a workshop where two are gathered.
machines for cutting stainless steel, one machine for stamping, two machines for the
polishing and finishing. Each machine operates 120 hours per month.
The variable manufacturing costs are compiled in the following table:
Cost of the hour Lot A Lot B Lot C
Cut 20 € 1h 1.5 hours 1.5 hours
Stamping 30 € 0.5 h 1h 1h
Polishing and finishing 40 € 2h 85 € 1h
Stainless steel 50 € 200 € 68 €
Selling price (excl. tax) 200 € 210 €

Determine the monthly production program that maximizes the result.

Elements of response
Definition of variables:xnumber of batches of 50 pieces A,
ynumber of batches of 50 pieces B,
znumber of batches of 50 pieces C.

120
Canonical form:
Positivity x≥ 0 ; y≥ 0 ; z≥ 0
Cut 1x+ 1.5y+ 1.5z≤ 240 (2 machines)
Stamping 0.5x+ 1z≤ 120 (1 machine)
Polishing and finishing 2x+ 1y+ 1z≤ 240 (2 machines)
MAX F = 35x+ 45y+ 42z

Calculation of the coefficients of the economic function:

Lot A Lot B Lot C

Cut 020 30 30
Stamping 15 30
Polishing and finishing 80 40 40
Stainless steel 50 85 68
1- Total of variable charges 165 155 168
2- Selling price 200 200 210
2 - 1 = Contribution margin 35 45 42

x≥ 0 ; y≥ 0 ; z≥ 0 ; e1≥ 0 ; e2≥ 0 ; e3≥ 0

⎧ 1x+ 1.5 y+ 1.5 z+ e1 = 240
⎪
Standard form: 0.5⎨ x + 1 z+ e 2= 120
⎪ 2 x+ 1 and+ 1 z + e3= 240
⎩
MAX F = 35x+ 45y+ 42z+ 0 e1+ 0 e2+ 0 and3

First table:
Outside Base x y z . . . B R
In Base
L1 e1 1 1.5 1.5 1 0 0 240 160
L2 e2 0.5 0 1 0 1 0 120 ∞
L3 e3 2 1 1 0 0 1 240 240
L4 F 35 45 42 0 0 0 0

Second table:
Off Base x . z e1 . . B R
In Base
L'p = L1 / 1.5 y 2/3 1 1 2/3 0 0 160 240
L’2 = L2 - 0 L’p e2 0.5 0 1 0 1 0 120 240
L3 = L3 - 1 Lp e3 4/3 0 0 -2/3 0 1 80 60
L4 = L4 - 35 Lp F 5 0 3 - 30 0 0 - 7 200

Third table:
Outside Base . . z e1 . e3 B
In Base
L'1 = L'1 - 2/3 L''p y 0 1 1 1 0 -0.5 120
L’’2 = L2 - 0 L’’p e2 0 0 1 0.25 1 -0.375 90
L''3 = L'p = L'3 / 4/3 x 1 0 0 -0.5 0 0.75 60
L''4 = L4 - 35 L''p F 0 0 -3 -27.5 0 -3.75 - 7 500

The optimum is reached since all the marginal rates of substitution (or coefficients of the last
lines) are negative or zero.

121
 The variablezis out of base, soz= 0. The value of variablesxandywho are in the database is read in the
Column B:x= 60 ;y= 120 and F = 7,500
Constraint 2 has an unused capacity of 90 work units.
It is therefore necessary to produce 60 batches of part A, 120 batches of part B, and no batches of part C for a margin.
with maximum variable costs equal to €7,500.

Exercise 8
The production manager of a product P has studied the scheduling of tasks for the order.
from a client to delivery.
The process requires the completion of 10 tasks, the prerequisites and duration of which are
specified in the following table:
Summits of arrival
A B C D E F G H I J
Duration in
days
A 02 1 1 1
B 7 1
C 5 1 1
D 21
Summits E 14 1
origins F 5 1
G 4 1 1
H 1 1
I 2 1
J 1
Example of interpretation: task A, lasting 2 days, precedes tasks D, E, and F.

Work to be done:
1. Present the dictionary of immediately preceding tasks.
[Link] the levels of the graph.
[Link] the graph specifying on the sagittal representation: the earliest start dates and of
start at the latest and the critical path.
[Link] the critical path.
[Link] the total margin and the free margin of the tasks.
[Link] duration of the tasks is fixed, except for task E which represents an average duration.
The duration of task E is a random variable XEwhich follows a normal law. The experience of
the production manager notes that:
The average duration of task E is 14 days,
HeyThere is a 47% chance that this task will take between 7 and 21 days.
Let it beZthe total duration of the longest path containing E.
[Link] simple relationship can we write betweenZand XE?
[Link] the standard deviation of XE.
6.3Calculate the probability of delivering product P more than 45 days after its order.

122
 Elements of response
Question 1
Reading the matrix column by column gives the task dictionary immediately
previous:
Tasks Immediately Prior Tasks
A /
B /
C /
D A
E A, B, C
F A, C
G E, F
H G
I G
J H, I

Question 2
Tasks TAI(1) Level 0 Level 1 Level 2 Level 3 Level 4
A / A
B / B
C / C
D A D
E A, B, C E
F A, C F
G E, F G
H G H
I G I
J H, I J
(1)
TAI = Immediately preceding tasks

Question 3
Sagittal representation of the graph according to the MPM method.

123
 0 5 2 2 7
A D 21
2

0 0 7 7 7 25 26 28 28
14
B E H 1 THE END
4
5 1
21 21 27 27
2 G 4 J
5 2
0 2 5 5 16 25 25 Legend
a b
C F I a = start as soon as possible
X
b = start at the latest

Critical path: B-E-G-I-J

Question 4
Critical path verification: 7 + 14 + 4 + 2 + 1 = 28
Critical tasks have zero slack. Their calculation is, however, recalled.
Tasks Total margins Free margins
A 05 Minimum = 2 – 2 – 0 = 0
B 0 7-7-0=0
C 2 Minimum = 5 – 5 – 0 = 0
D 5 5
E 0 21 - 14 - 7 = 0
F 11 21 - 5 - 5 = 11
G 0 Minimum = 25 - 4 - 21 = 0
H 1 27 - 1 - 25 = 1
I 0 27 - 2 - 25 = 0
J 0 28 - 1 - 27 = 0

Question 6
Z = 14 + XE

XEfollows a normal lawN(14,σ). According to the text :

P(7 < XE< 21) = 0.47
X E− 14
- 7/σ < T < 7/σ= 0.47 with T = according to the normal lawN(0, 1)
σ
P(-t < T < t) = 0.47 with t = 7/σ
2π(t) - 1 = 0.47
π(t) = 0.735
Reading the table of the normal lawN(0, 1) gives t = 0.63.
Where 0.63 = 7/σ andσ 11.11
It is about calculating P(Z45) let P(XE+ 14 > 45)

124
 ⎛ 31− 14 ⎞
P(XE> 31) = P T>
⎜⎝ = P(T > 1.53) = 1 - P(T < 1.53) = 1 - 0.937 = 0.063
11.11 ⎟⎠

Exercise 9
Company D produces new stainless steel parts whose demand follows a normal distribution.
For the entire market, we have observed the demand per week for two years:
Number of lots requested Number of weeks
[0 - 10[ 1
[10 – 20[ 2
[20 - 30[ 3
[30 – 40[ 8
[40 - 50[ 25
[50 – 60[ 27
[60 – 70[ 20
[70 – 80[ 12
[80 – 90[ 5
[90 – 100[ 1

Provide a point estimate of the parameters of the normal distribution of demand.

[Link] company decides, to evaluate the profitability of this project, to base itself on an overall demand.
annual that has a 60% chance of being exceeded. It estimates that in the first year, it can take
1/4 of the market then 1/3 in the following years. We will accept the previous estimate and the independence.
requests for each of the weeks.
Determine, in number of lots, the total potential market and the sales the company can hope for.
to accomplish.

Elements of response
1. Calculation of the mean and standard deviation of the sample of size 104 (by taking the center of the
classes) :
x5,720 divided by 104 equals 55.
σ‘ = 16.1125
A point estimate of m isxThat's 55 lots per week.
104
A point estimate ofσ is s = 16.1125 = 16.19 lots per week.
104− 1
The weekly demand follows a normal distribution.N(55 ; 16,19)

[Link] annual demand D is the sum of the 52 independent weekly demands and follows a distribution.
⎞
normalN⎜⎛55×
⎝ 52 ; 52× 16,19 = N(2,860
2
⎟⎠ ; 116.75)

The annual demand Q, which has a 60% chance of being exceeded, is such that:
⎛ Q− 2,860 ⎞
P(D > Q) = 1 – P T<
⎜⎝ = 0.6
116.75 ⎟⎠
⎛ Q− 2,860 ⎞
P ⎜ T< = π(t) = 0.4 (t < 0 car 0.4 < 0.5)
⎝ 116.75 ⎟⎠
The value t will be found using the property:π(– t) = 1 –π(t).

125
The problem to be solved becomesπ(– t) = 1 - 0.4, that is:π(– t) = 0.6.
The reading of the table gives: - t = 0.253 and therefore t = -0.253.
The annual demand that has a 60% chance of being exceeded is therefore such that:
Q− 2860
- 0.253 s let Q = 2,830, 4 2,831 lots
116.75
Expected sales for the first year are 2,831 / 4≈ 707 lots.
The expected sales for the following years are 2,831 / 3≈ 943 lots.

Exercise 10
To appreciate the quality of the production, a controller seeks to evaluate the percentage p of items.
non-marketable. To do this, it randomly selects samples of size n with replacement.
The controller takes a sample of 125 items and finds that 10 are not sellable.
sands.
Determine an estimate of p, using a centered confidence interval, with the confidence coefficient.
of 95%.

What should be the minimum sample size n (n is an integer) taken for that, with
the 95% confidence coefficient, the percentage p being 8% within 2%?

Elements of response
A point estimate of the proportion p is f = 10 / 125 = 0.08.
⎡ 0.08× 0.92 0.08× 0.92 ⎤
⎢ − 1,96
I = 0,08 0.08+ 1,96 ⎥ [0.0324; 0.1275]
⎢⎣ 125 125 ⎥⎦

[Link] that f is always 8%, the sought interval at a 95% confidence level is
[0,08 – 0,02 ; 0,08 + 0,02]

0.08× 0.92
The size n of the sample is such that: 1.96 0.02
n
0.08× 0.92
2 2
let it be 1.96 n 0.02
where n = 706.85≈ 707 articles

Exercise 11
In order to control its production, a company conducts a study on a sample of 65 products. The
The results obtained for the variable 'weight of a product' are as follows:
Weight in grams Number of observations
From 115 to < 117 04
From 117 to < 119 6
From 119 to < 121 12
From 121 to < 123 20
From 123 to < 125 13
From 125 to < 127 7

126
 From 127 to < 129 3

[Link] the mean and standard deviation of the weight distribution observed in the sample.
[Link] the value of the point estimators of the mean and the standard deviation of the weight of a
product of the entire production of the company.
[Link] a 95% confidence interval for an estimate of the average weight value.
of a product of the entire production of the company.

Elements of response
1. Calculation of the mean and standard deviation of the sample of size 65 (by taking the center of
classes)
x= 7 930 / 65 = 122 grams.
σ‘ = 2.9351 grams.
2.A point estimate of m is 122 grams.
65
A point estimate ofσ is s = 2.9351 = 2.958 grams.
65− 1

[Link] population variance is unknown, but since the sample size is greater than 30,
the central limit theorem it is possible to retain s as an estimate.
⎡ 2,958 2,958 ⎤
A 95% confidence interval is: 122− 1,96× ⎢ ; 122+ 1,96× ⎥
⎣ 65 65 ⎦
[121,28 ; 122,72]

Exercise 12
At the end of December N–1, the human resources director communicates to the employee representatives
the following negotiation basis:
Two cumulative increases in monthly salaries:
Increase from July 1 N Increase of November 01 N
Frames 1% 0.75%
Non-executives 1.25 % 1%

A retention of staff for the years N and N+1.

A regulated salary increase (non-cumulative) of 1.5% applicable to non-executives on July 1st.
exercises N and N+1.
The objectives of limiting categorized salary masses are as follows:
Objective of variation for N+1
Frames 2%
Non-executives 4%

The forecasted salaries in thousands of euros for the first semester are as follows:
First semester N
Frames 4,782,000 €

127
 Non-executives 13,692 K€

Work to do:
[Link] the projected payroll masses for the fiscal years N and N+1.
2. Calculate for each category the level effect, the mass effect, and the transfer effect from N to N+1.
[Link] the relationship between the three calculated effects.
[Link] the variation of the payroll mass of N+1.
[Link] to the human resources director a single negotiated salary increase by category
applicable from July 1 of year N+1, allowing to meet the goals of salary mass variation.
For non-executives, you will specify the portion outside of the regulated increase.
[Link] calculating, indicate what the consequences would be of changing the date of the
modifications of N+1 on the effects.

Elements of response
1. Let M be the monthly mass of January.
For the frames, we know that 6 M = 4,782
J F M A M J J A S O N D
Indices 100 100 100 100 100 100 101 101 101 101 101,7575 101,7575
Cumul 100 200 300 400 500 600 701 802 903 1,004 1,105.7575 1,207,515

MS exercise N = (4,782 / 6)× 1,207.515%≈ 9,624 K€

In N+1, as no increase is planned, the index will remain equal to 101.7575 for the twelve.
month, for a total annual of 1,221.09%
MS exercise N+1 = (4 782 / 6)× 101,7575%× 12= (4 782 / 6)× 1,221.09%≈ 9,732,000€

For non-executives, the regulated increase in July is not multiplicative but additive.
J F M A M J J A O N D
Indices 100 100 100 100 100 100102.75102.75 102.75 102,75 103,7775 103,7775
Cumul 100 200 300 400 500 600 702,75 805,5 908,25 1 011,00 1 114,7775 1 218,555

MS exercise N = (13,692 / 6)× 1,218.555%≈ 27,807 million euros

In N+1, the regulated increase in July is not multiplicative but additive.

J F M A M J J A S O N D
103.7775 103.7775 103.7775 103.7775 103.7775 103.7775 103.7925 103.7925 103.7925 103.7925 103.7925 103.7925
103,7775 207,555 311,3325 415,11 518,8875 622,665 726,4575 830,25 934,0425 1 037,8350 1 141,6275 1 245,42

MS exercise N+1 = (13,692 / 6)× 1,245.42%≈ 28,420,000 €

2.
Level effect Mass effect Carryover effect N to N+1
Frames 101.7575 / 100 1 207,515 / (100× 12) (101,7575× 12) / 1 207,515
1.017575 1.00626 = 1.01124
= + 1.7575 % = + 0.626 % = + 1,124 %
Non-executives 103.7775 1 218,555 / (100× 12) (103,7775× 12) / 1 218.555

128
 1.037775 1.01546 1.02197
= + 3.7775 % = + 1,546 % = + 2,197 %

[Link]: level effect N = mass effect N× Deferred effect N on N+1

1.01757 = 1.00626× 1.01124
1.037775 = 1.01546× 1.02197

4. To verify the variation of the payroll of N+1, it is necessary to calculate the mass effect.
of N+1.
N+1 mass effect
Frames 1,221.09 / (101.7575× 12) = 1
=0%
Non-executives 1,245.42 / (103.7775× 12) = 1.000072
= + 0.0072%

Variation of the payroll N+1 mass N+1× report N on N+1

Frames 1 221.09 / 1 207.515 = 1× 1,01124 1.01124 = 1× 1.01124
Non-executives 1 245.42 / 1 218.555 = 1.02204 1.02204 = 1.000072× 1.02197

[Link] executives
The sum of the indices of N+1 with the increase ofx% = Σ of the indices N × 1.02
101,7575 × 6 + 101,7575 × 6 × (1+x1207.515 × 1.02
x1.732 %

It is also possible to use the relation:

Variation of the payroll N+1 = mass effect N+1× effect report N on N+1
Let it bexthe negotiated increase on July 1
x
1.02 = 1.01124× (1 + )
2
x 1.732 %

For non-executives
Sum of indices of N+1 with the increase ofx% = ∑ of indices N × 1.04.
103.7775 × 6 + 103.7775 × 6 × (1 +x1 218.555 × 1.04
x= 3,527 % (yunderstood the regulated increase of 1.5%
The part outside of the regulated increase is therefore 3.527 - 1.5 = 2.027%. It was possible to pose
(1 + (x+ 1.5)) in the equation to be solved to obtain this percentage directly.

It is also possible to use the relationship:

Variation of the payroll N+1 = mass effect N+1× effect report N on N+1
Let it bex, the negotiated increase on July 1
x+ 0.015
1.04 = 1.02197× (1 + )
2
x2,027%
This increase adds to the regulated rise of 1.5%, which allows for the return of the increase.
total of 3.528 %.

[Link] the date of the N+1 increase is brought forward, the mass effect of N+1 increases and the effect of
The report from N+1 to N+2 is less significant.

129
Conversely, if the date of the increase for N+1 is delayed, the mass effect of N+1 is weaker.
and the carryover effect from N+1 to N+2 is more significant.

Exercise 13
At the end of December N, the human resources director informs you of the following information:
Payroll:
Average gross salaries for December N Effective Gross annual salary N
Management personnel 6,000 5 355,000
Engineers 5,000 16 960,000
Technicians 4,000 222 10,500,000
Workers 2,400 50 1,410,000
Employees 3,000 45 1,580,000
TOTAL 338 14,805,000

The N+1 salary agreement provides for a uniform increase applicable to all in the form of
two cumulative increases in monthly salaries:
Increase of February 1st N+1 Increase from September 1 N+1
1% 0.75%

An individual regulated salary increase (non-cumulative) of 0.95% applicable from July 1st.
N+1.
Neither the employees who will leave during the year, nor those hired during the year will benefit.
individual increases.
No internal promotion is planned for year N+1.
On the other hand, external movements are being considered. Only retirements will free up some
positions. Not all positions will be filled.

Table of planned departures and hires:

Gross salary Gross salary
Departure date Arrival date
monthly monthly
Worker 31/03/N+1 2,500 01/04/N+1 1,900
Worker 30/11/N+1 3,000 / /
Employee 31/07/N+1 3,000 / /
Employee 30/11/N+1 3 200 / /
Technician 31/08/N+1 3,500 01/06/N+1 2,500
Technician / / 01/11/N+1 2,500
Technician / / 01/11/N+1 2,500

Calculate the projected payroll for N+1.

[Link] for the management staff category the N+1 level effect, the mass effect of
N+1 and the carryover effect of N+1 on N+2. Interpret the carryover effect.

[Link] the relationship between the three calculated effects for the management personnel category.

130
 Calculate the carryover effect from N to N+1 for the management personnel category. Check the variation.
from the projected payroll of N+1 for this category based on this effect.

Elements of response
Indices of stable workforce:
J F M A M J J A S O N D
Indices 100 101 101 101 101 101 101,95 101,95 103,5812 103,5812 103.5812 103,5812
Cumulative100 201 302 403 504 605 706,95 808,9 912,4812 1 016,0624 1 119,6436 1 223,2248
The regulated increase in July is not multiplicative but additive: 101.95 = 101 + 0.95
The general increase in September is multiplicative: 103.5812 = 101.95× 1,016

– Payroll of the stable workforce:

Calculation Amount
Management staff 5× 1 223.2248 %× 6,000 366,967.44
Engineers 16× 1,223.2248%× 5,000 978,579.84
Technicians (222× 4,000 - 3,500) 1,223.2248 % 10,819,423.36
Workers (50× 2,400 - 2,500 - 3,000) 1,223.2248 % 1,400,592.40
Employees (45× 3,000 – 3,000 – 3,200) 1,223.2248 % 1,575,513.54
TOTAL 15,141,076.57

– Payroll of the outgoing staff until departure:

Indices (without individual increase)
J F M A M J J A S O N D
Indices 100 101 101 101 101 101 101 101 102,616 102,616 102,616 102,616
Total 100 201 302 403 504 605 706 807 909.616 1 012.232 1 114.848 1 217.464

Calculation Amount
Worker 302 %× 2,500 7,550.00
Worker 1 114,848 %× 3,000 33,445.44
Employee 706 %× 3,000 21,180.00
Employee 1 114,848 %× 3,200 35,675.14
Technician 807 %×3,500 28,245.00
TOTAL 126,095.58

– Payroll of the outgoing staff until departure:

Indices (without individual increase)
J F M A M J J A S O N D Total
Worker 100 100 100 100 100 101,6 101,6 101,6 101,6 906,4
Technician 100 100 100 101,6 101,6 101,6 101,6 706,4
Technician 100 100 200

Calculation Amount
Worker 1,900× 906.4% 17,221.60
Technician 2,500× 706.4% 17,660.00
2 Technicians 2,500× 2× 200 % 10,000.00
TOTAL 44,881.60

Projected mass N+1 = 15,141,076.57 + 126,095.58 + 44,881.60 = 15,312,053.75

131
[Link] effect = 103.5812 / 100 = 1.035812 soit + 3.5812 %
Mass effect = 1,233.2248 / 100× 12 = 1.019354 up + 1.9354 %
Carryover effect N+1 on N+2 = 103.5812× 12 / 1 233.2248 = 1.016145 or + 1.6145%
Even if no increase occurs in N+2, the payroll in N+2 will increase by 1.6145% in
reason for the measures taken in N+1.

[Link]: level effect N+1 = mass effect N+1× Carryover effect from N+1 to N+2
1.035812 = 1.019354× 1,016145

[Link] of the payroll N+1 = 366,967.44 / 355,000 = 1.03371 or +3.371%

Variation of the payroll N+1 = effect of payroll N+1× carryover effect N on N+1
Carryover effect N on N+1 = 6,000× 5× 12 / 355,000 = 1.014084 or +1.4084%
Variation of the payroll N+1 = 1.019354× 1.014084 = 1.03371 therefore + 3.371%

Exercise 14
The following information is communicated:
Year N Year N+1
Average salary Average salary
Effective MS N Effective MS N+1
annual annual
Frames
< 5 years 2 55,000 110,000 5 54,000 270,000
[5 –10] 5 65,000 325,000 3 65,500 196,500
10 3 75,000 225,000 2 76,000 152,000
Total 10 66,000 660,000 10 61,850 618 500
Technicians
< 5 years 5 20,000 100,000 10 19,000 190,000
[5 –10] 30 22,000 660,000 35 22,500 787 500
10 5 25,000 125,000 5 25,500 127,500
Total 40 22 125 885,000 50 22 100 1,105,000
Total general 50 30 900 1,545,000 60 28,725 1,723,500

1. Calculate the variance in payroll (absolute and relative variation). What are the factors that
Explain this variation?
[Link] this gap (the gap based on seniority will be calculated from the seniority structure at
constant composition for the category of managers and then the category of technicians.
[Link] the GVT effect, then break down this effect.

Elements of response
Discrepancy on payroll: 1,723,500 - 1,545,000 = +178,500 € (Unfavorable)
178,500 / 1,545,000 = + 11.5534 %

This difference is due to a variation: - in the numbers

- salaries
from the professional structure
from the seniority structure

132
[Link] on total workforce: (En+1 En) × Sn
(60 – 50)× 30,900 = + 309,000 (Unfavorable)
60 / 50 = 1.2 + 20 %
MSn+1 at constant total salary MSn
Average salary
Effective Amount Effective Sn Amount
annual
Total general 60 30,900 1,854,000 50 30,900 1,545,000

[Link] structure gap

Effective N+1 structure
Effective N+1 SCn Gap
constant professional
Frames 10 60× (10 / 50) = 12 66,000 132,000 Favorable
Technicians 50 60× (40 / 50) = 48 22 125 + 44 250 Unfavorable
Total 87,750 Favorable

Theya shift from the executive category to the lower-paid technician category like
shows the second possible approach:
Salary N at
Staff N+1 SCn
effective N+1
Frames 10 66,000 660,000.00
Technicians 50 22 125 1,106,250.00
60 29,437.5 1,766,250.00

Professional structure gap = 60 (29,437.5 – 30,900) = –87,750 Favorable

The development of the calculations gives: [10× 66,000 + 50× 22 125] - 1 854 000
-87,750
1,766,250 / 1,854,000 = 0.95266 -4,733 %

[Link] structure gap or noria effect

It is about highlighting the impact of seniority for each category.
Workforce N+1 with seniority
Effective N+1 SAn Gap
constant
Frames
< 5 years 05 10× (2 / 10) = 2 55,000 165,000 Unfavorable
[5 –10] 3 10× (5 / 10) = 5 65,000 – 130,000 Favorable
10 2 10× (3 / 10) = 3 75,000 – 75,000 Favorable
Total cadre - 40,000 Favorable
Technicians
less than 5 years 10 50× (5 / 40) = 6.25 20,000 75,000 Unfavorable
[5 –10] 35 50× (30 / 40) = 37.5 22,000 – 55,000 Favorable
10 5 50× (5 / 40) = 6.25 25,000 -31,250 Favorable
Total technicians - 11 250 Favorable
Overall total - 51 250 Favorable

Rejuvenation effect on categories [5-10] and more than 10 years of seniority among executives and among
the technicians, which translates to a decrease in the average salary of each of these two categories.

133
Second approach applied to frameworks:
Salary N at
Staff N+1 SAn
effective N+1
< 5 years 5 55,000 275,000.00
[5 –10] 3 65,000 195,000.00
10 2 75,000 150,000.00
10 62,000 620,000.00

Seniority structure discrepancy of executives = 10 (62,000 - 66,000) = -40,000 Favorable

Second approach applied to technicians:

Salary N to
Staff N+1 SAn
effective N+1
< 5 years 10 20,000 200,000
[5 –10] 35 22,000 770,000
10 5 25,000 125,000
50 21,900 1,095,000

Seniority structure gap of technicians = 50 (21,900 - 22,125) = -11,250 Favorable

The development of the calculations gives for the frameworks:

⎡ 2 5 3 ⎤
[5× 55,000 + 3× 65,000 + 2× 75,000 - 10×⎢ × 55,000+ 10× × 65,000+ 10× × 75,000 ⎥
⎣ 10 10 10 ⎦
⎡ 2× 55,000+ 5 × 65,000+ 3 × 75,000 ⎤
[5 × 55,000 + 3× 65,000 + 2× 75,000 - 10×⎢ ⎥
⎣ 10 ⎦
Σ (EAn+1 × SAn) - ECn+1 × S Cn
[5× 55,000 + 3× 65,000 + 2× 75,000 - 10× 66,000
-40,000

The development of calculations provides for technicians:

⎡ 5 30 5 ⎤
[10× 20,000 + 35× 22,000 + 5× 25,000 – 50×⎢ × 20,000+ 50× × 22,000+ 50× × 25,000 ⎥
⎣ 40 40 40 ⎦
⎡ 5 × 20,000+ 30× 22,000+ 5 × 25,000 ⎤
10× 20,000 + 35× 22,000 + 5× 25,000 - 50×⎢ ⎥
⎣ 40 ⎦
[10× 20,000 + 35× 22,000 + 5× 25,000 - 50× 22 125
-11,250

- Total: 1,715,000 - 1,766,250 = -51,250 0.97098 -2.9016%

[Link] on nominal rate

This is about comparing the salary of N+1 to the salary of N for each type of staff of N+1.
Headcount N+1 Salary N+1 Salary N Gap
Frames
less than 5 years 05 54,000 55,000 -5 000
[5 -10] 3 65,500 65,000 1 500
10 2 76,000 75,000 2,000
Technicians

134
< 5 years 10 19,000 20,000 -10,000
[5 –10] 35 22,500 22,000 17,500
10 5 25,500 25,000 2,500
1,723,500 1,715,000 8,500 Unfavorable

[Link]:
Deviation on total headcount + 309 000
Professional structure gap - 87 750
Seniority structure gap - 51 250
Spread on nominal rate 8,500
Total 178,500

7. Calculation of the GVT effect

The GTV effect refers to structural factors other than the number of employees. Here, it can be broken down into the effect of
professional structure and noria effect.

The payroll has increased by 178 500

up + 11.5534% be multiplied by 1.1155

Nominal wages have increased by:

1,723,500 / 1,715,000 soit 0.4956 % be multiplied by 1.00495

The number of staff has increased by: 10

it is 20% be multiplied by 1.2

GVT Effect
1,1155 = 1,00495× 1,2× GVT effect GVT effect = 0.925 -7.5%

This calculation can also be obtained by: Noria effect× Professional structure effect
0.97098× 0.9526 = 0.925 -7.5%

135