Ø BUSINESS CONTROL
Ø
Business control, lesson controlling,
free courses, direct costing, margin on variable costs, Margin variance
analysis, Bruno THIONNET
Home : http://controlegestion.pagespersoorange.fr/
©
Copyright 2011 – Bruno THIONNET – Le contrôle de gestion opérationnel
3. Margin analysis
B 
efore
discussing the principles relating to the margins variances analysis, we need
first to consider how information should be organized in order to produce all
required data. We are going to take the example of a processing company which
sells its products to retail customers. However, this method could be applied
to other economic models.
3.1. Data sourcing
As shown in the chart below, statistics are
provided by the data flows recorded on IT systems, from purchasing to sales
delivery. Information is collected and organized into files in order to perform
margin variances analysis. Joint Marketing efforts (JME = sales deals) with
distributors must be integrated into the analysis process.
Before
starting margin analysis, we need to know the right levels on which we are
going to carry out the calculations. Most of the time, they are product
families, brands, sales networks or countries i.e. the best way to describe the
business of a company. This stage is really important in order to have a clear
view of the margin analysis. In that way, some mix effects are not really
relevant. Particularly we will try to avoid consolidate dissimilar activities.
The example below is based on a company organized around two different sales
networks, two product families and for each one: company brand, private labels
and first prices:
¨ Sales Network A
·
Products family 1

Branded products

Private labels

First prices
·
Products family 2

Branded products

Private labels

First prices
¨ Sales Network B
·
Products family 1

Branded products

Private labels

First prices
·
Products family 2

Branded products

Private labels

First prices

¨ And so on…
3.2. Margin
variances analysis
Margin variances come from a
comparison between actual figures and budget or previous year data. They are
divided in four levels:
·
Sales volumes variances
·
Net selling
price variances
·
Mix effects
·
Standard costs
variances
In the following development
the budget will be used for comparison. But we could as well use the previous
fiscal year, and even better. However a comparison with previous year needs
some IT adjustments and particularly the introduction of ‘master references’ in
order to link new articles with ‘old’ items still in use.
3.2.1. Sales volume variances
Sales volume variances must be
calculated for each selected group (network / products family / brand). The
formula is really easy to implement:
Δ Sales volume = (Actual Qties – Budget Qties) * €/kg
margin of selected Group
3.2.2. Net selling price variances
The net selling price (NSP)
variance (price net of rebates and JME = sales deals) must be calculated item per
item, using the following formula:
Δ NSP =
Most of the
time, some references are not budgeted. For these items the NSP variance should
be at a zero value. It is then necessary to perform the calculations by using a
similar article. This adjustment (arbitration) must be done manually.
3.2.3. Mix effect
The mix effect is the most
difficult variance to calculate. Actually mix comes both from changes in
products proportion inside one aggregate (a product family for instance) and
different profitability levels of references.
Δ mix = [(Actual volume of article  budget
volume article / budget volume of the family of the article * actual volume of the family)
* (€/Kg budget margin of article  €/kg budget
margin of the family of article)]
The first member of the equation (Actual volume of article  budget volume article / budget
volume of the family of the article x actual volume of the family) is called adapted volume on budget proportions. The
second member of the equation (€/kg
budget margin budget of article  €/kg budget margin of the family of the
article) is related to the various profitabilities of articles.
The following table should give a better understanding of the formula
above:
In fact, a mix calculation is
not just a simple formula to use. Most of the time some articles are not budgeted
(new items), it is then necessary to proceed to manual adjustments. These new
items have to be linked to a similar reference (raw in our example). But sometime, it is not
possible at all to link new articles to any ‘old’ item. In that case it is
possible to use the actual or estimated margin of these new references. This
way to proceed can introduce a slight approximation in the calculation but
anyway this will give a better appreciation of the products mix effect.
The ‘link’ principle is also
possible for the net selling prices (NSP) variance calculation as well as the
costs of products variance of which the formula will be detailed further down.
Under normal conditions, the
mix effect will be calculated only when a detailed analysis is necessary. Thus
for each item, negative and positive effects will be determined with accuracy.
But generally, the mix variance will be calculated by difference as shown
below:
Δ Mix = Δ Total Margin – Δ Sales Volume  Δ Net selling
prices – Δ COSTS
Some people could be amazed, but mix calculations can be time consuming
both for development and analysis, particularly when a product family is
composed with several hundreds of articles. We always try to shorten the
closing time, but of course without undermine the quality level of analysis.
3.2.3. Costs variances
The costs variance is broken down as follows:

Purchase price differences

Yield and productivity variances

Hourly rate
variations

Standards actualisation effects

Temporary standard costs adjustments
In practice, only purchase prices variations can be incorporated into the
stock entries valuation. For other variances, in one hand it depends on their
significance level and in the other hand on the possibilities of the computer data
processing. The variances, which are not included into production batches
valuation, will be taken into account later in the reconciliations margins
‘controlling <> accounting' (cf. § 3.4)
The calculation formula for the
variance of costs is the same like sales price ones. She must be calculated
item per item:
Δ Costs=
Arbitrations have to be done for
new references created during the current period because it could not be
budgeted (same approach versus N1)
3.3. Table of
Margin variances
When calculations and
adjustments have been done, margin variances can be synthesized in a table, as shown
below:
In this example, the analysis
is organized along three axes:
 The first
one for product family
 The second
one for sales network
 The third
one related to brands
Generally speaking, it is
better when the variances are calculated in YTD position and the month deducted
by difference. Thus the emergence of a referent product on a certain month
allows the possibility for a global recalculation. Moreover, the sum of the
month variances is not equal to the calculation in cumulative terms. Some
transfers exist between the different margin compounds.
3.4. Margin
reconciliation between cost accounting and statistical position
Some differences are not taken
into account by statistical approach. It is therefore necessary to make
reconciliation with the cost accounting:
BUDGET
MARGIN ACTUAL STATISTICAL MARGIN
Δ MARGIN
STATS
RECONCILIATION
Δ MARGIN
ACCOUNTING
BUDGET
MARGIN ACTUAL ACCOUNTING MARGIN
They are two types of variances which
are not integrated into statistical margins approach:
Ø
The differences that we decided not to include because
they are considered not really significant,
Ø
The differences that we do not know integrate (or
difficult to integrate in the stats for technical reasons)
Among the differences that we decided
to include or not in the statistical models, we can mention:
§ differences
in productivity (labor) and yields (material)
§ hourly rate variances
§ differences
of raw material hedging
§ exchange
rate differences
§ discounts on purchases
§ subcontracting supplementary costs
§ …
§
Variances, which are difficult
to integrate in the statistical models, are mostly related to accounting
entries not reflecting an actual physical flow:
§ accruals
for credit notes
§ accruals
for credit notes receivable not related to physical flows
§ invoices to
be issued
§ Differences
of stock inventory
§ invoicing
for various products and services (called ‘out of stats’)
§ …
These lists are of course not
exhaustive.
The table below allows you to
group the various kinds of margin differences. In the first part, we can find
differences related to statistical margin analysis (Δ sales volume, sales
prices, mix and costs) and in the second part, the complementary variances to
ensure the reconciliation of the margin difference of the management P & L
between 'Actual' and 'Referent' (the budget in our example):
It is very
difficult to ensure a complete reconciliation of the margin difference. Most of
the time, there remains a residual gap but that must still be the lowest
possible. This reconciliation difference is also called ‘system gap’. This
variance reflects the efficiency of your model, the lower it is, and the better
your system is performing.
© Copyright 2011 – Bruno THIONNET – Le contrôle de gestion opérationnel