See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228317486

Have IFRS Standards Improved the Information Content of Intangibles?: The

Case of French Listed Companies

Article · April 2010

CITATIONS READS

2 153

2 authors, including:

Jean-Michel Sahut
Ecole de Commerce IDRAC
164 PUBLICATIONS   727 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

CSR, financial performance and risks View project

Facteurs et moyens informels pour garantir le remboursement des crédits View project

All content following this page was uploaded by Jean-Michel Sahut on 26 February 2014.

The user has requested enhancement of the downloaded file.

 Have IFRS improved the information content of intangibles? :
The case of French listed companies

Sandrine Boulerne
Cerege – University of Poitiers

Jean-Michel Sahut
Amiens School of Management
Cerege – University of Poitiers

Article presented at the 32nd Annual Congress of the European Accounting Association
Tampere, Finland, 12-15 May 2009

Abstract
IAS/IFRS are presumed to provide more uniform and value-relevant information which better
reflects a company’s financial situation and assets and, in particular, provide greater
transparency in the area of intangible assets. In this article, we study the information content of
these assets under IAS/IFS when compared to French GAAP for French companies listed on
the SBF 250 index. We show that the transition to IAS/IFRS did not change the overall amount
of intangible assets, even though it operated substitution effects in favour of goodwill.
However, total intangible assets and goodwill gain greater accounting relevance when they are
valued according to IAS/IFRS. By applying these standards, financial markets can better
integrate such contributions into share prices and returns, especially for companies with a high
intensity of intangible assets.

Key words: IFRS, IAS, GAAP, intangible asset, goodwill, performance, return
Introduction

The nature of investments made by companies has changed remarkably over the past few years.
New trends in business practices and changes to the economic environment have brought about
the emergence of new production function variables in which the role of intangibles, as both
factor and product, is increasing (de Montmorillon, 2001). The most frequently cited factors
driving the changes are the globalisation of financial markets, the development of the
knowledge-based economy and the growing number of mergers and acquisitions. This new
context has contributed to the rising importance of intangible assets such as brands, patents,
training costs, R&D costs, organisational competences, etc. Faced with this growing wave of
intangibles, traditional accounting standards systems, based on a ‘transactional principle’, are
finding it increasingly difficult to fulfil their informative role in decision-making (Lev, 1999).
Furthermore, there was a problem of comparability and transparency of financial information
particularly in Europe before 2005. Given the diversity of choices and conditions of financial
reporting of intangible investments offered by different accounting systems, there was a lack of
clarity as to whether intangibles were to be capitalized or expensed.

IAS/IFRS, mandatory since 1st January 2005 for listed companies, aim to give more uniform,
value-relevant information which better reflects a company’s financial situation and asset.
However, recent studies show the difficulty of forecasting the impact that changing some
accounting rules will have on the quality of financial data due to the fact that the latter is
influenced by several complex institutional factors (Ball et al., 2003; Ding et al., 2007).
Accounting regulations exist within a mosaic of other institutional rules. Changing one element
of this mosaic is not always the best solution when the other elements remain invariable (Hope
et al., 2006; Ding et al., 2007).
Though studies already carried out on the French market have shown a relatively modest
impact of IAS/IFRS on equity capital (Schatt and Gross, 2007), or fixed assets (Benabdellah
and Teller, 2006), many questions remain unanswered about their effect on intangible assets
which determine both the performance and the valuation of companies and, in particular, of
firms which are R&D intensive (Lantz and Sahut, 2005). Moreover, the small size of samples
looked at in these studies, in general CAC 40 companies (Schatt and Gross, 2007; Bessieux-
Ollier, 2007), limits the scope of their conclusions.

This study looks at the transition to IAS/IFRS by French companies and, in particular, at the
impact of the adoption of IAS 38 and IFRS 3 on the quality of financial information on
intangibles. We complete the Bessieux-Ollier and Walliser study (2007) in two specific ways.
We first look at the effect of the transition to IFRS on net income, equity capital and different
sorts of intangible assets on a bigger sample of firms, then we analyse the impact of these
standards on the share price and return of the firms concerned.
After a reminder of the main differences in regulations concerning intangibles between French
GAAP and international standards, we point out the results of previous work on the
information content of intangibles. Then, we present our research hypothesis concerning the
issues involved when first applying the standards. Finally, we test our empirical models on a
sample of French firms listed on the SBF 250 index, before presenting our conclusions.

2
1- The regulatory framework
In the international system of reference, intangible assets are governed by IAS 38i and IFRS 3
in the event of a business combination.

IAS 38 defines an intangible asset as being ‘an identifiable non-monetary asset which is
without physical substance but is separableii’. It imposes the reporting of all intangible
expenditure as intangible assets if, and only if, (IAS 38, § 22):
(a) it is probable that future economic benefits attributable to the asset will flow to the entity,
(b) the cost of this asset can be measured reliably.
An intangible expenditure must therefore be either expensed or capitalized. Optional treatments
no longer exist. If an intangible expense results from a business combination and can not be
reported as an intangible asset, then it is incorporated into the amount attributed to goodwill on
the date of acquisition.
Thus several intangible expenses (such as advertising, research costs, staff training costs…),
which provide companies with future economic benefits, can not be activated due to the
restrictions on capitalization (characteristics which identify the asset, control procedures…)
except if they are acquired as part of a business combination. After the initial recognition, IAS
38 specifies that amortizable intangible assets must be depreciated over their useful life with
the amortization method reflecting the pattern of consumption of future economic benefits.
By comparison, French standard accounting practices do not expand greatly on the criteria for
capitalizing intangible items or their accounting treatment after their initial recognition.
Generally they are treated on a case-by-case basis. Their method of reporting varies between
expensing, optional capitalization and obligatory capitalization. For amortizable assets, the
duration and the method of amortization are not strictly defined and allow for a margin of
discretion in how the accounting rules are applied.

Business combinations highlight, in the majority of cases, a difference between the cost of
acquisition and the acquiree’s proportionate interest in the fair valueiii of identifiable assets and
liabilities at the date of taking control (Martory and Verdier, 2000). Governed by IFRS 3, the
accounting treatment of goodwill is one of the most difficult problems in accounting. The
difficulty lies in how best to identify and measure goodwill. Indeed, goodwill is an asset
which, in practice, encompasses factors which do not possess the essential characteristics of an
asset (Johnson and Kimberley, 1998) such as overvaluing the purchased company. On the other
hand, goodwill generated in-house is not capitalized due to the fact that costs incurred during
its creation are not, in practice, identifiable from regular expenses or those which are necessary
to maintaining its value. This creates a distortion in the comparability of companies which have
different growth methods.

The accounting treatment of goodwill which has been acquired after its initial recognition
complicates the situation further. According to French GAAP, it should be amortized whereas
IFRS recommends carrying out impairment tests.
Amortization allows companies to both apportion the cost of purchasing the goodwill over its
useful life (reflecting consumption of future economic benefits) and to make its value
progressively disappear from the balance sheet. This results in a value which is identical to that
of internally generated goodwill. However, the amortization of goodwill entails, in particular:
· a systematic depreciation of goodwill as well as a finite lifespan.
· a book value for goodwill which has no relation to the economic value of the
company (Jennings et al., 1996)

3
 · goodwill depreciation which does not really represent the loss of value of the latter
(Henning et al., 2000).

Non-amortization of goodwill avoids these problems, but impairment testing obliges managers
to make choices about numerous parameters which create possible sources of manipulation.

Finally, having considered the difficulties encountered when trying to define or assess
intangibles, one may question the capacity of IAS/IFRS to provide all stakeholders with the
most relevant information about intangibles in general, and goodwill in particular.

2- Information content of intangibles

Over the last three decades, researchers have tried to prove that intangible investments
contribute to a company’s future performance, that they should therefore be considered as
assets, and thus necessitate some information content. Most studies focus either on R&D
expenditure or on goodwill.

2.1 Impact of R&D expenditure

The first research carried out by Johnson (1967) and Newman (1968) on the information
content of R&D did not unmask any significant relationship between future returns and
investment in R&D. These first results have been contradicted by numerous subsequent studies
which have highlighted the relevance of entering this type of expenditure as an asset. Thus,
Hirschey (1982) and Hirschey et al (1984) show that, on average, advertising and R&D costs
are positively linked to the share prices of American companies. Sougiannis (1994), starting
from the hypothesis that the market value is the result of discounting the company’s future
performance, confirms previous results. He came to the conclusion that the increase in R&D
expenditure brought higher profits over a period of at least seven years. Furthermore, he
suggested that the non-conclusive results of the first studies carried out in the field are due to
the econometric techniques used and to the quality of the data on R&D at that time.
Lev and Sougiannis (1996), following the same lines as Sougiannis (1994), established there is
a significant link between equity capital and earnings and stock market price and returns when
R&D expenditure is capitalized. Similarly, Aboody and Lev (1998) show that software
development costs are seen as assets in companies’ share price and are significantly correlated
to future returns.
These results suggest, therefore, that R&D expenditure is, on average, considered by investors
as amortizable assets rather than immediate costs. They contradict the main reason on which
the FASB based its ban on their capitalization, namely, the lack of proof in the existence of a
direct relation between R&D expenditure and future income.

However, these results were obtained from empirical studies based on American data. Nothing
guarantees that the results and the relationships identified are transferrable to other
environments (Ding et al., 2007) such as European countries and in particular France (Casta et
al, 2007) because of the regulatory and cultural differences, as well as differences in the
structure of the markets (Ding et al., 2005; Ball et al., 2003; Hope, 2003, Hope et al, 2006; Ali
et Hwang, 2000; Pope and Walker, 1999). Issues concerning the appropriate accounting
treatment of R&D expenditure can, alone, illustrate this difficulty (Casta et al., 2007) for three
main reasons:

4
 · In the United States, the market is informed of the total amount of expenditure on R&D
(non-capitalized R&D expenditure is clearly disclosed in the income statement)
whereas in France, the information is generally only disclosed for the amount
capitalized.
· Inversely, in France (unlike the USA), the distinction is made between the cost of basic
research and the expenditure on R&D which has a high chance of success. Therefore
American studies investigating the relevance of capitalizing R&D expenditure take the
totality of this cost.
· American studies reprocess the companies’ accounting data to show their relevance in
explaining stock market data where R&D expenditure has been capitalized. The scope
of the conclusions remains rather limited. This is because, even if it really were
possible to capitalize R&D expenditure, it is not sure that companies would adopt the
same behaviour in terms of earnings management and how they disclose financial
information, and that investors would react in the same way to information published
about intangibles.

In fact, studies looking at the international context tend, overall, to confirm the American
results. Thus, Zhao (2002), in a comparative study of 4 countries (France, Great Britain,
Germany and the United States) concludes that: (1) the disclosure of information about total
R&D expenditure increases the association between the market value and financial data such as
the income statement and the equity capital in countries where the capitalization of costs is
banned (Germany and the United States), (2) the allocation of R&D expenditure between assets
and costs increases the relevance of financial information in countries which allow the practice
(Great Britain and France).

Whilst studies carried out in the French context differ from the majority of previous results.
Ding et al. (2004) tried to analyse the determinants in the capitalization of R&D expenditure
and the impact of this choice on the relevance of financial data from companies included in the
SBF 250 index. Their results stipulate that companies which capitalize their R&D expenditure
are those which are only listed on the French stock market, belong to the high-tech sector and
have a higher Beta risk. However, their statistical analyse does not confirm if this decision
improves the value-relevance of the financial data. On the other hand, Cazavan-Jeny (2004)
shows that over the period of 1994 to 1999 intangible expenditure expensed or capitalized was
not associated with the book-to-market ratio. Thus, it suggests that this information is not
useful in explaining the gap between companies’ book value and their market value. Moreover,
the Cazavan-Jeny and Jeanjean study (2005) of 93 listed companies over the 1998-2000 period
highlights a negative association between capitalized R&D costs and stock market returns. This
result contradicts most American and international research. The authors explain these contrary
results by the fact that corporate executives choose to capitalize R&D expenditure for
opportunist reasons or are incapable of discriminating R&D projects which are cost effective
from those which are not. Casavan-Jeny and Jeanjean (2006) confirm these results over a
longer period by establishing that firms which capitalize their R&D expenditure have lower
returns and are under-valued in comparison to those which expense their R&D costs.

Similarly, Dufour and Zemzem (2005), while analysing how French companies listed on the
new technologies exchange (Nouveau Marché) report their R&D expenditures, stipulate that
financial information concerning R&D can not be used for external analysis. However, they
state that the only independent variable of R&D cost capitalization is profitability. Indeed, the
propensity to capitalize R&D costs is more important in companies which have low
profitability. This suggests that capitalization is motivated by the desire to improve financial

5
results, confirming therefore the findings of Cazavan-Jeny and Jeanjean (2005) about the
opportunism of corporate executives. These results question the merit of capitalising R&D
expenditure as advised in IAS 38, in countries such as France, where the level of shareholder
protection is relatively weak compared to the USA (Cazavan-Jeny and Jeanjean, 2006).

2.2 Impact of goodwill

Concerning goodwill, studies focussing on the USA dominate because changes to reporting
methods far precede those observed in Europe. For many authors, the impairment of goodwill
(according to SFAS 142 standards or IFRS 3) should reveal better information than systematic
amortization as the latter can underestimate the loss in real value of goodwill (David, 2005).
Indeed, he observes that companies having implemented SFAS 142 did not all depreciate their
goodwill. For those companies that did record a depreciation of their goodwill, it was higher
than the amortization which they used previously.
Recent studies by Henning et al. (2000), Hirschey and Richardson (2002), Duangploy et al.
(2005), Schultze (2005) also show the relevance of impairment tests on capitalized goodwill
when it is not amortized. The use of impairment tests enable goodwill paid without
consideration to be written off as a loss. Thus, equity capital and the income statement convey
respectively better information on the value of the company and its fluctuating value.
Similarly, if a company announces the depreciation of its goodwill the result will be a fall in its
trading price as investors interpret it as negative information about the future economic benefits
that this asset was supposed to bring.

However, the implementation of these impairment tests does have limitations. Indeed, it
obliges corporate executives to make discretionary choices such as the rate of discounting, the
evaluation of future cash flows, etc. (Massoud and Raiborng, 2003; Schevin, 2005). In
particular, it becomes possible to produce ‘revaluation reserves’ against reporting loss of value
to acquired goodwill. These reserves enable the capitalization of internally generated goodwill
up to the level of initially recognized goodwill.

Inversely, the tests can be used for big bath accounting following a strategic error or a change
of management (Sevin and Schroeder, 2005). This practice was already possible with French
GAAP via exceptional amortization. IFR 3 will surely have little impact on this aspect. Indeed,
when the speculative dot.com bubble burst, Vivendi recorded an exceptional amortization of
15.7 billion euros in 2001, against 12.8 billion euros for France Telecom between 2002 and
2004 in order to depreciate their goodwill from numerous acquisitions carried out during the
technological bubble in the late 1990s.

To summarize, these studies point out that IAS 38 and IFRS 3 have, overall, increased the
information content of intangibles despite substantial sectoral and geographic differences. It is
important therefore to check if their information content is more value-relevant in the French
context considering the contrasting results observed for this country.

6
3- Hypotheses and data collection

3.1 Hypotheses

The conditions for entering intangible items under assets on the balance as defined by the IAS
38 standard are stricter than those under French GAAP, in particular regulation 99-02iv. IAS 38
standard imposes that an intangible asset must be identifiable to clearly distinguish it from
goodwill, as well as defining its future economic benefitsv.
These more restrictive conditions for capitalization as defined by this international standard
should, on one hand, incite companies who adopt them to minimise the intangible assets on the
balance sheet (Gatet P. and Tassin H, 1998) and, on another hand, generate a shift of
unidentifiable intangible items towards goodwill. Consequently, accounting models to measure
goodwill should reflect a combination of synergies, those of initial consolidation difference and
those of non-homogenous items which could not be qualified as intangibles.

From these reflections we formulated the following hypotheses:

Hypothesis 1a: IAS 38 incites companies to minimise intangible assets, other than goodwill, on
their balance sheet.

Hypothesis 1b: The transition to IFRS increases goodwill under the combined effect of IFRS 3
and IAS 38. In particular, IAS 38 induces companies to subsume unidentifiable intangible
items in goodwill.

Hypothesis 1c: The transition to IFRS has not made the overall amount of intangible assets
vary in any significant way (effect of substitution between goodwill and other intangible
assetsvi).

Moreover, data on intangible assets is often used by investors and financial analysts as forecast
indicators of a firm’s value and performance. Therefore, the financial statements under
IAS/IFRS present, contrary to those under French GAAP, detailed information about the
totality of intangible expenditure capitalized or expensed in the footnotes. Moreover, the
banning of optional treatments and of derogation methods in the IAS/IFRS system of reference
should not only reduce the risks of manipulative accounting practices but, by increasing the
transparency and comparability of financial data between firms, should make such behaviour
more easily detectable. This should reduce information asymmetry between corporate
executives and investors, consequently relieve the problem of undervaluing R&D-intensive
companies and thus increase the correlation between a company’s accounting and stock market
data.

However, some researchers point out that international standards offer a wide margin of
discretion allowing corporate executives to appreciate capitalizable intangible expenditure, and
define the useful life of intangible assets in order to carry out goodwill impairment tests. This
discretion makes it easier for corporate executives to ‘manage’ the profit and loss statement
(Stolowy and Breton, 200; Cazavan-jeny and Jeanjean, 2005). It should be highlighted that
managerial latitude has not been curtailed, at least for intangibles, under French GAAP. Firstly,
optional treatment for several intangible costs offers the corporate executives the choice as to
whether or not to capitalize the expenditure. Then, on choosing capitalization, they can
manipulate the amount to capitalize. In particular, the propensity to capitalize R&D
expenditure is higher for companies which have low profitability (Dufour and Zemzem, 2005).

7
Finally, the amortization expense can equally be manipulated by under-estimating or over-
estimating the useful life of assets as French GAAP gives no guidelines for determining the
length of amortization of R&D expenditure capitalized. It only stipulates a maximum delay of
5 years.
Moreover, the restrictive conditions for capitalization specified under IAS/IFRS create a certain
discrepancy in the treatment of some expenses (such as brands, market share and research
costs) if they have been acquired or have been produced internally. Thus, companies which are
experiencing internal growth must expense these costs and their accounting data becomes less
value-relevant in comparison to companies which are growing through mergers and acquisition
operations. This situation does little to improve the transparency and comparability of financial
data.
Despite these different contrary effects, we anticipate that the changes brought about by
IAS/IFRS will improve the information content of intangible assets.

Hypothesis 2a: The informational content of intangibles under IFRS is even higher when
goodwill is disassociated from other intangible items. In other words, the explanatory capacity
of the model will be improved if goodwill is distinguished from other intangible items under
IFRS rather than if an overall level of intangible assets is considered.

Hypothesis 2b: Goodwill and other intangible items under IFRS are positively associated with
trading prices.

Hypothesis 2c: Goodwill and other intangible items under IFRS are positively associated with
higher returns.

Considering the chosen sample, the company profiles are very disparate both in size and in
volume of intangible assets. The latter represent 3% (Michelin) to 70% (Sanaofi-Aventis) of
the companies’ balance sheet. Thus, there is a risk that the informational impact of IFRS 3 and
IAS 38 on share prices and returns depends on how intangible-intensive the firms are. In order
to verify this hypothesis, we differentiate firms which are intangible-intensive from the rest. To
be qualified as such, their balance sheet must show an above average level of intangible assets
compared to the average level for the sample group.

Hypothesis 2d: A high intensity of total intangibles under IFRS has a positive impact on share
price and returns.

In France, the Accounting Standards Council (Conseil National de la Comptabilité) encouraged

companies to adopt international standards before 2005 in parallel to French standards, though
without authorizing an anticipated adoptionvii. Since then, French companies listed on British
or American stock markets simultaneously publish their consolidated financial statements
under US GAAP or IAS/IFRS and under French GAAP. However, international accounting
standards have been strongly inspired by British and American accounting regulations, in
particular where the capitalising of intangible assets is concerned (Obert R. 2003).
Consequently, to ensure the comparability and coherence of financial information, French
companies listed on British or American stock markets tend to expense some unidentifiable
intangibles, such as R&D costs, even on their accounts prepared according to French GAAP.

8
Hypothesis 3: A listing on the British and American stock markets would discourage
companies earlier from capitalizing unidentifiable intangible items and would have a positive
impact on their stock price and returns.

3.2 Data collection and selection of sample group

Our sample is made up of companies listed on the SBF 250 index in October 2005. The
accounting data under both French and international standards were obtained by direct
consultation of published annual reports and a study of the consolidated financial statements
published in the BALO (French official legal announcements publication)viii. The main
difficulty was to identify firms which communicated their financial statements under
IAS/IFRS. Firms whose financial year ended after the 31st December 2004, were not subjected
to AMF recommendations to also publish their financial statements to international standards
during the transition period. In order to keep the information relevant, only companies which
had published their consolidated accounts according to IAS/IFRS were maintained in our
sample group. Stock market data was obtained from the DataStream database.

The composition of the sample is shown in table 1 below:

Table 1: Composition of the preliminary sample

Number of companies from the SBF index 250
- financial and real estate companies - 38
- companies which ended their fiscal period after the 31/12/04 - 50
- companies publishing their financial statement according to IFRS before 2004 -4
- companies which do not distinguish goodwill from other intangibles even in the footnotes -3
- companies not listed on the stock market in 2003 -2
- companies which have not published their opening balance sheet on the 01/01/2004 to IFRS - 29
- companies which have not published their consolidated accounts to 2004 IFRS -4
Companies making up the preliminary sample for statistical study 120

4-Methodology and results

4.1 Univariate tests

Before testing our research hypotheses 2 and 3 to determine the degree of relevance of
accounting data under IAS/IFRS, it is essential to study the eventual change to the value of this
data, expressed simultaneously under two different accounting systems of reference, on the
same date, (the 31st December 2004), across the whole of our sample (hypotheses 1a, b and c).
The descriptive analysis and the univariate results for all the firms included in the study are
indicated in table 2.

9
 Tableau 2 : Impact of the French GAAP to IFRS transition in 2004

Wilcoxon test
French GAAP IFRS Mean test
(French GAAP versus IFRS)
% of total Assets N Mean Median Mean Median T test P value Rank N Z test P value
a
GW 120 11,98% 8,60% 15,96% 13,96% 4,790 0,000 positive 86 6,003 0,000

OINT 120 9,13% 3,52% 5,99% 2,31% -3,641 0,000 negativeb 74 -3,106 0,002

INTTOT 120 21,11% 19,36% 21,95% 20,05% 1,825 0,071 positive c 70 2,959 0,003

NI 120 3,95% 3,84% 4,59% 4,18% 4,124 0,000 positived 86 4,675 0,000
e
EQUI 120 37,58% 36,78% 37,25% 34,71% -0,764 0,446 negative 69 -1,390 0,165

Variable definitions (data source) :

Sample consists of French listed firms where transition to IFRS was mandatory in 2004. GW is goodwill, OINT are other intangible assets,
INTTOT are total intangible assets, NI is net income and EQUI is owners' Equity. All data were collected from the Datastream database.
Notes :
a. GW IFRS > GW French GAAP, b. OINT IFRS < OINT French GAAP, c. INTTOT IFRS > INTTOT French GAAP,
d. NI IFRS > NI French GAAP, e. EQUI IFRS < EQUI French GAAP

10
The adoption of IAS/IFRS for drawing up financial statements has really brought about
modifications in the value of accounting data. Over the whole sample, the net income (NI) has
increased on average by 3.95% of total assets using French standards to 4.59% of total assets
using IFRS (the median rising from 3.84% to 4.18%) This 16% rise is statistically significant.
The impact of IAS 38 on intangibles appears more pronounced. In fact, total intangible assetsix
(INTTOT) have increased, on average, by nearly 4% during the transition to the new
accounting standards. More preciselyx, goodwill (GW) has increased from 11.98% to 15.96%
of total assets (the median has increased from 8.6% to 13.96%), in other words a difference of
33.2% which is statistically significant. Inversely, the average of other intangible assets
(OINC) has decreased from 9.13% to 5.99% of total assets (the median from 3.53% to 2.31%).
This decrease of 34.4% is also statistically significant.

According to the Wilcoxon testxi, more than 71% of companies in our sample (86 firms out of
120) have seen an increase in the value of goodwill when changing over to IFRS and more than
61% of them have diminished the book value of other intangible assets. The Wilcoxon test has
a significance threshold of 1%, except for equity.

A first interpretation of these results suggests that companies had transferred intangible assets
that couldn’t be separated over to goodwill. Under the more restrictive IFRS reporting
conditions, intangible assets should no longer include unidentifiable intangibles. Only
separable assets can be qualified as intangible items. Thus, in this initial analysis of the results,
the book value of intangible assets which are not individualised seem to have been integrated
into goodwill. This shifting of unidentifiable intangibles to goodwill tends to confirm
hypotheses 1a and 1b.
We specify that when applying the revised IRFS 3 and IAS 38 standards, goodwill and
intangible assets with an indefinite useful life also undergo a supplementary revaluation
because of the suppression of obligatory amortization which applied to them. This
supplementary revaluation of goodwill and other intangible assets justifies an average rise of
only 4% in total intangible assets (INTTOT) of our sample and confirms our hypothesis 1c.

These results back up those of Bessieux-Ollier and Walliser (2007) obtained from studying
CAC 40 companies. The relative overall stability of intangible assets in total long-term assets
was counterbalanced by substitution effects; some companies reclassified intangible assets as
goodwill even if they no longer fit the definition given by IFRS such as market share.

Besides observing these reclassifications of unidentifiable intangible items, originally recorded

in other intangible assets, as goodwill, it is important to question how investors perceive the
practice. What impact does such a transposition have on the firm’s share price and returns? The
accounting measurement of goodwill reflects not only the value of first consolidation spread
but also the value of unidentified intangible assets in a more consistent way under international
standards. However, do investors see it as providing them with more value-relevant
information? A multivariate analysis can be used to test hypotheses 2 and 3 and suggest some
answers.

11
 4.2 Multivariate tests

4.2.1 Association between intangibles and share prices

To establish the relevance of accounting information on intangibles by examining their impact

on the financial market, we use a first model, frequently used in empirical research, which
studies the relationship between the price of securities (P) and the book value of equity capital
per ordinary share and the net income per share (NIPS)xii. The book value of equity capital is
broken down into book value per adjusted share of capitalized intangible assets (EQUIPSA)
and into book value per action of total intangible assets (INTTOTPS). Moreover, to isolate the
relevance of the book value of goodwill and of other intangibles, the book value per share of
total intangible assets (INTTOTPS) is broken down into book value per goodwill share
(GWPS) and into book value per share of other intangibles (OINTPS). Firms whose total
intangible assets are higher than the averagexiii for the sample are considered as having a high
intensity of total intangibles (HDTI). Equally, the listing (C) of a company on British and
American stock markets is integrated into the model. The unknown factor e represents the
unexplained element of the share price. Equations (1) and (2) of this first model are as follows:

P bi ,t
=
0
+ b NIPS 1 i ,t
+ b EQUIPSA
2 i ,t
+ b INTTOTPS + b 3 HDTI + b C + e
3 i ,t
[1]
5 i ,t i ,t

P =b + b NIPS + b EQUIPSA + b GWPS + b OINTPS + b HDTI + b C + e [2]

i ,t 0 1 i ,t 2 i ,t 3 i ,t 4 5 i ,t 6 i ,t i ,t

Pi,t
= price of a share in firm i 3 months after the end date of fiscal period t.
NIPS i, t
= Net Income per share of firm i in fiscal period t.
EQUIPSA i,t
= book value of equity per share in firm i for fiscal period t adjusted of total intangible assets.
INTTOT i, t
= book value of total infantile assets per share in firm i for fiscal period t.
GWPS i, t
= Book value of the goodwill per share of firm i for the fiscal period t.
OINCPS i, t
= Book value of other intangible assets per share in firm i for the fiscal period t.
HDTI i, t = high intensity of total intangibles of firm i at the end of the fiscal period t: .
C i,t
= Listing on the British and American stock market of firm i at the end of the fiscal period t:

The information on the book value of goodwill and of other intangibles is available only when
the financial statements have been published, in other words three months after the end date of
the fiscal period. Like Aboody and Lev (1998), we consider that the dependant variable will be
the share price three months after the end date of the fiscal period.

This assessment model has the advantage of using the accounting data as an approximation of
the discounted future cash flow hoped for by investors and of the market value of the firm.
The equations [1] and [2] of the first model will be subjected to two regressions: a first with the
accounting data using French standards and a second with the financial data using IFRS.

According to hypotheses 2b and 2c, valuing goodwill and other intangibles under the new
accounting standards should facilitate forecasting the price of securities. If the overall quality
of the model with accounting data using IFRS, measured by the R², is better than the same
model with accounting data using French standards, the hypotheses 2b and 2c will be validated.
The coefficient associated with OINTPS should be positive if the amount of intangible items
capitalized using IFRS has a higher predictive value for investors. The coefficient associated
with GWPS should also be positive if investors have perceived that, under IFRS, goodwill can
integrate unidentifiable intangibles which have future economic benefitsxiv. On the contrary, it
should not be significant if they perceive these unidentifiable intangible items as a source of
information which has little relevance.

12
 4.2.2 Results of model 1

The statistical results of linear regressions on model (1) are presented in table 3.

The quality of the adjustment and the overall significance of the model using IFRS are higher
than those of the model using French GAAP. The results of the model show the existence of a
positive and significant relationship at the threshold of 1% between goodwill per share and the
price of the share. Thus, the financial information conveyed by capitalized goodwill appears to
be as weak using IFRS as French GAAP. Indeed, the coefficient associated with GWPS is
positive and statistically significant under the two accounting standards. Even if unidentifiable
intangible items are lost within the heterogeneous whole which makes up goodwill, the
accounting measurement of the latter under international standards is always a relevant source
of information for investors.

When valued according to international standards, other intangible assets do not provide
investors with more value-relevant information. The coefficient associated with INCPS is
positive but does not show any significance either under French GAAP or under IFRS
(p>0.05). Therefore, identified intangible assets capitalized on the companies’ balance sheet do
not provide any more value-relevant information for shareholders than unidentified intangible
assets which have been transferred into goodwill. The standard-setters do not seem to have
achieved their objectives with the application of IAS 38, in giving more importance to the
reliability of information by banning the capitalization of several unidentifiable intangible
items. These results partially disprove hypothesis 2b.

However, shareholders consider that a high intensity of total intangible assets is a source of
future economic benefits, but only if they have been valued under international standards. The
coefficient associated with HDTI is positive and significant at a threshold of 5%, this result
validates hypothesis 2d.

We should also underline that shareholders consider the informational content of total
intangible assets (INTTOTPS) as being value-relevant, without making a distinction between
goodwill and other intangible items. The coefficient associated with INTTOTPS is positive and
significant (p>0.01). Hypothesis 2a is disproved as the explanatory power of the model is low
with the distinction of two intangible assets rather than considering an overall level of
intangible assets.

Hypothesis 3 is also disproved (coefficient is not significant). The fact that a firm is listed on
the British or American stock market has no resulting impact on the share price.

13
 Table 3: Regression analysis of stock prices: French GAPP & IFRS
PRICE MODEL
French GAAP IFRS
[1] [2] [1] [2]
NIPS 4,647** 4,921** 5,983** 5,99**
(4,342) (4,555) (6,843) (6,883)
EQUIPSA 1,236** 1,201** 1,025** 1,03**
(10,883) (10,477) (12,577) (12,552)
INTTOTPS 0,597** 0,524**
(3,499) (3,672)
GWPS 0,65** 0,596**
(2,591) (3,316)
OINTPS 0,468 0,314
(1,477) (0,901)
HDTI 15,297 15,54 15,718* 16,084*
(1,877) (1,823) (2,060) (2,097)
Cotation -3,594 -4,883 -5,296 -5,351
(-0,408) (-0,547) (-0,652) (-0657)
Constant 10,42 10,517 6,948 6,816
(1,54) (1,494) (1,095) (1,071)

Number of observations 120 120 120 120

R-square 0,879 0,875 0,899 0,900
Adjusted R-squared 0,762 0,753 0,801 0,800
F 77,128 61,448 96,596 80,172
Prob > F 0,000 0,000 0,000 0,000

Two regression models with P i,j as the dependent variable. Pi,j is the stock price for firm i 3 months after fiscal year-end t. The sample consists of
French listed firms under Local GAAP and IFRS in 2004. NIPS i,t is the net income per share for firm i at time t. EQUIPSAi,t is the book value of
owners' equity per share, adjusted of total intangible assets, for firm i at time t. INTTOTPSi,t are total intangible assets per share for firm i at time t.
GWPSi,t is the goodwill per share for firm i at time t. OINTPSi,t are other intangible assets per share for firm i at time t.
Notes :
*, ** and *** represent significance at 10%, 5% and 1% levels respectively. All data were collected frome the Datastream database.

14
 4.2.3 Association between intangibles and stock returns

In the continuity of this research and to confirm the robustness of our results, as Easton (1999)
suggests, we tested a second model which linked the stock returns (R) to the variations of book
value of unidentifiable intangibles by share (DOINTPS) and of goodwill by share (DGWPS),
besides those of Net Income per share (DNIPS) and adjusted equity capital per share
(DEQUIPSA).

R i ,t
= b 0
+ b D NIPS
1 i ,t
+ b D EQUIPSA
2 i ,t
+ b D INTTOTPS + b HDTI
3 i ,t 4 i ,t
+b C +e
5 i ,t i ,t
[3]

R i ,t
= b
0
+ b D NIPS
1 i ,t
+ b D EQUIPSA
2 i ,t
+ b D GWPS
3
+ b D OINTPS
i ,t 4 i ,t
+ b HDTI
5 i,t + b 6
C i ,t
+e i ,t
[4]

With
R i ,t
= Stock returns of firm i 3 months after the end date of the fiscal period t.
where P is the price of the share 3 months after the end date of the fiscal period
R = [(P + Dividendes )/ P ]- 1
t.
i ,t i ,t i ,t i ,t -1

DNIPS i,t
= Variation of the Net Income per share of firm i in year t.
DEQUIPSA i ,t
= Variation of the book value of equity per share of firm i at the end of year t adjusted for total intangible
assets.
DINTTOTPS i ,t
= Variation of the book value of total intangible assets per share of firm i at the end of year t.
DGWPS i ,t
= Variation of the book value of goodwill per share of firm i at the end of year t.
DOINTPSi,t = Variation of the book value of other intangible assets per share of firm i at the end of year t.
HDTI i ,t = high intensity of total intangibles of firm i at the end of the year t:
C i,t = Listing on the British and American stock market of firm i at the end of year t:

In order to have data which complies with the model to be evaluated, a preliminary analysis of
the residualsxv enabled us to delete 17 observations whose residuals exceeded two and a half
times the standard deviation estimated for the unknown factor as an absolute value. For the
second model, the definitive sample for a multivariate analysis is made up of 103 companies.

4.2.4 Results of model 2

The results of model (2), shown on table 4, enable us to partially corroborate hypothesis 2c.
The improvement in the book value of goodwill under international standards has informative
value for explaining stock market returns. The coefficients associated with DGWPS are
positive and significant (p<0.05).

However, this informative value is inexistent for investors when other intangible items are
valued using French GAAP (p>0.418) or IFRS. They do not perceive identifiable intangibles as
being a source of value for the firm. The overall quality of the model using IFRS is superior to
that of the model using French GAAP (the adjusted R-squared increases from 23% to 28.2%).

15
 Table 4: Regression analysis of stock returns: French GAAP & IFRS
RETURN MODEL
French GAAP IFRS
[3] [4] [3] [4]
ÀNIPS 0,02766** 0,02623** 0,02633** 0,02699**
(4,197) (4,175) (4,317) (4,253)
ÀEQUIPSA 0,006139** 0,005104* 0,005244** 0,005082*
(2,721) (2,359) (2,611) (2,415)
ÀINTTOTPS 0,006002* 0,006393**
(2,170) (2,710)
ÀGWPS 0,006975* 0,006908*
(2,019) (2,129)
˨INTPS -0,002643 0,004854
(-0,409) (0,868)
HDTI -0,04812 -0,04278 -0,07431* -0,09175*
(-1,273) (-1,196) (-2,077) (-2,495)
Cotation -0,113** -0,103* -0,07263 0,0832
(-2,618) (-2,503) (-1,782) (-1,927)
Constant 0,231** 0,244** 0,214** 0,237**
(8,586) (9,529) (8,078) (8,694)

Number of observations 103 103 103 103

R-square 0,529 0,525 0,564 ,569
R2 ajusté 0,243 0,230 0,283 0,282
F 7,536 6,083 9,069 7,737
Prob > F 0,000 0,000 0,000 0,000

Two regression models with Ri,j as the dependent variable. Ri,j is the stock return for firm i 3 months after fiscal year-end t. The sample consists
of French listed firms under Local GAAP and IFRS in 2004. ÀNIPSi,t is the variation of net income per share for firm i at time t. ÀEQUIPSAi,t is
the variation of the book value of owners' equity per share, adjusted of total intangible assets, for firm i at time t. ÀINTTOTPS i,t are the variation
ot the total intangible assets per share for firm i at time t. ÀGWPS i,t is the variation of the goodwill per share for firm i at time t. ÀOINTPS i,t are
the variation of the other intangible assets per share for firm i at time t.
Notes :
*, ** and *** represent significance at 10%, 5% and 1% levels respectively. All data were collected frome the Datastream database.

16
Considering the results of tables 3 and 4, the low inflation factors of the variance (VIF<2.0)
associated to low standard deviations from estimates of parameters shows the absence of
problems of colinearityxvi.

5- Conclusion and limits

The adoption of IAS/IFRS has really brought about changes in the value-relevance of financial
data on SBF 250 companies. The effect, which professionals and researchers know well
enough for net income, is even more marked for intangible assets. The relative stability of total
intangible assets (4% on average) hides strong substitution effects of other intangible assets
(drop of 34%) towards goodwill (rise of 33%). Numerous companies have reclassified as
goodwill their intangible assets which no longer fitted the definition given by IAS 38.
The empirical tests show that information conveyed by accounting data on goodwill and total
intangible assets valued according to IFRS is more value-relevant. The effect being that total
intangibles and goodwill are statistically significant under the two accounting reference
systems, as opposed to other intangible assets, but the quality of the adjustment and the overall
significance of the model using IFRS is superior to that of the model using French GAAP.
Despite the tightening of criteria for entering assets in the category of other intangible assets
(IAS 38), only goodwill explains the share price and the stock returns when the different types
of intangible assets are separated for the model using IFRS. Investors therefore continue to
only pay attention to goodwill. This phenomenon brings into question the investors’ perception
of goodwill considering the reclassification during the changeover to IAS/IFRS which made it
increase in a significant manner. It also casts doubts on the usefulness for investors of IAS 38
which bans the entering of several unidentifiable intangible items on the balance sheet.
Moreover, when we differentiate firms according to the intensity of their intangible assets, we
notice that this criterion explains their stock price and their return only when using
international standards. From this perspective, shareholders consider that a high intensity of
total intangible assets is a source of future economic benefits.
Finally, whether a company is listed on the British or American stock market has no resulting
impact on the share price.

In their annual report, the majority of companies have chosen the option offered by IFRS 1 to
not reprocess business combinations which took place earlier than the date of transition, in
other words, 1st January 2004. Therefore, our comparative analysis between the application of
IFRS and French GAAP has an imperfection. Indeed, the transition to IFRS will only be
completed for accounts consolidated on 31st December 2005. Accounts consolidated in 2004
for the majority of companies in our sample were drawn up without any pre-application of IAS
32 and 39 which concern financial instruments. The trends of stock prices and returns are not
affected by the fair value of these financial instruments, which introduces supplementary bias
in the veracity of our results.
Moreover, it should be underlined that the sample used is made up of only 120 companies on
the SBF 250 index using financial and stock market data from a single year (2004). Carrying
out the study of a larger sample group and with a period of at least three years to look back
over would enable us to more accurately capture the impact of international standards (a
revised IAS 38 and IFRS 3) on the stock prices and returns of listed companies.

Finally, we can conclude that goodwill under IAS/IFRS represents the synergies from business
combinations and also from unidentifiable intangible assets reclassified under IAS 38. Doesn’t
this phenomenon risk changing investors’ perception of goodwill?

17
Bibliography

Aboody, D. and Lev, B. (1998). The value relevance of intangibles: the case of software
capitalization. Journal of Accounting Research 36: 161-191.

Ali, A. and Hwang L. (2000). Country-specific factors related to financial reporting and the
value relevance of accounting data. Journal of Accounting Research 38: pp. 1-21.

Ball, R., Robin, A. and Wu, J.S. (2003). Incentives versus Standards: Properties of Accounting
Income in Four East Asian Countries and Implications for Acceptance of IAS. Journal of
Accounting and Economics 36: 235-270.

Benabdellah, S. and Teller, R. (2006). Immobilisations corporelles et immeubles de placement

en IFRS: le coût historique va-t-il résister à la juste valeur ? Revue Française de Comptabilité
March: 40-44.

Besley, D.A., Kuh, E. and Welsch, R.E. (1980), Regression Diagnostic, John Wiley & Sons.

Bessieux-Ollier, C. (2006). Les pratiques d'évaluation et de publication des entreprises

françaises, allemandes et américaines: le cas des éléments incorporels. Comptabilité-Contrôle,-
Audit 12 (2) : 167-189.

Bessieux-Ollier, C. and Walliser, E. (2007). La transition et le bilan de la première application

en France des normes IFRS: le cas des incorporels. Comptabilité-Contrôle-Audit 13 (special
edition): 219-246.

Casta, J.F., Escaffre, L. and Ramond, O. (2007). Investissement immatériel et utilité de

l’information comptable: Etude empirique sur les marchés financiers européens. Conférence
SIFF, Louvain-la-Neuve, 16-17 September.

Cazavan-Jeny, A. (2004). Le ratio market-to-book et la reconnaissance des immatériels- une

étude du marché français. Comptabilité-Contrôle-Audit 10 (2): 99-24.

Cazavan-Jeny, A. and Jeanjean, T. (2006). The Negative Impact of R&D Capitalization: A

Value Relevance Approach. European Accounting Review 15 (1): 37-61.

Cazavan-Jeny, A. and Jeanjean, T. (2005). Pertinence de l’inscription à l’actif des frais de

R&D: une étude empirique. Comptabilité-Contrôle-Audit 11 (1): 5-21.

Chan, K., Lakonishok, J. and Sougiannis, T. (2001). The Stock Market Valuation of Research
and Development Expenditures. Journal of finance 56 (6): 2431-2456.

Deng, Z. and Lev, B. (2006). In-process R&D: To capitalize or expense? Journal of

engineering and technology management 23: 18-32.

Ding, Y., Jeanjean, T. and Stolowy, H. (2005). Why do national GAAP differ from IAS? The
role of culture. The International Journal of Accounting 40: 325-350.

18
Ding, Y., Stolowy, H. and Tenenhaus, M. (2004). Les déterminants de la stratégie de
capitalization des frais de recherche et développement en France. Revue Finance Contrôle
Stratégie 7 (4): 87-106.

Ding, Y., Stolowy, H. and Tenenhaus, M. (2007). R&D productivity: an exploratory

international study. Review of Accounting and Finance 6 (1): 86-101.

Duangploy, O., Shelton, M. and Omer, K. (2005). The value relevance of goodwill impairment
loss. Bank Accounting & Finance: 23-28.

Dufour, D. and Zemzem, A. (2005). Activation des frais de R&D: le cas des entreprises
françaises cotées sur le nouveau marché. 26th annual congress of the Association Francophone
de Comptabilité, 11th-13th May.

Gatet, P. and Tassin, H., (1998). La norme IASC va conduire à minimiser les actifs
immatériels. Analyse Financière 116: 26-27.

Hennings, S. L., Lewis, B. L. and Shaw, W. H. (2000). Valuation of the components of

purchased goodwill. Journal of Accounting Research 38: 375-386.

Hirschey, M. (1982). Intangible Capital Aspects of Advertising and R & D Expenditures. The
Journal of Industrial Economics 30 (4): 375-390.

Hirschey, M. and Connolly, R. (1984). R & D, Market Structure and Profits: A Value-Based
Approach. The Review of Economics and Statistics 66 (4): 682-686.

Hirschey, M. and Vernon, J. R. (2002). Information content of accounting goodwill numbers.

Journal of Accounting and Public Policy 21: 173-191.

Hope, O. K. (2003). Disclosure practices, enforcement of accounting standards and analysts’

forecast accuracy: An international study. Journal of Accounting Research 41: 235-272.

Hope, O. K., Jin, J. and Kang, T. (2006). Empirical evidence on jurisdictions that adopt IFRS.
Journal of International Accounting Research 5 (2): 1-20

Jennings, R., Robinson J., Thompson, R. B. and Duvall, L. (1996). The relation between
accounting goodwill numbers and equity values. Journal of Business Finance & Accounting
23: 513-533.

Johnson, J. (1967). A consequential Approach to Accounting for R&D. Journal of Accounting

Research Autumn: 164-172.

Johnson, J. T and Kimberley, R. P. (1998). Is Goodwill an Asset? Accounting Horizons 12:

293-303.

Lantz, J. and Sahut, J M. (2005). R&D Investment and the Financial Performance of
Technological Firms. International journal of Business 8 (4): 365-384.

Lev, B. (1999). R&D and Capital Markets. Journal of Applied Corporate Finance 11 (4): 21-
35.

19
Lev, B., Bharat, S. and Sougiannis, T (2005). R&D Reporting Biases and Their Consequences.
Contemporary Accounting Research 22 (4): 977-1026.

Lev, B. and Sougiannis, T. (1996). The capitalization, amortization, and value-relevance of

R&D. Journal of Accounting and Economics 21 (1): 107-138.

Lev, B. and Zarowin, P. (1999). The Boundaries of Financial Reporting and How to Extend
Them. Journal of Accounting Research, 37 (2): 353-385.

Martory, B., and Verdier, F. (2000). Comment traiter le goodwill? Pratique d’une théorie,
théorie d’une pratique. Comptabilité-Contrôle-Audit 6 (2): 175-193.

De Montmorillon, B. (2001). L’investissement immatériel. In Images de l’investissement (Ed

Charreaux G.). Paris. Vuibert, 259-282.

Newman, M. (1968). Equating Return from R&D Expenditures. Financial Executive, 26-33.

Obert, R., (2003). Pratique des normes IAS/IFRS. Paris. Dunod.

Ohlson, J.A. (2001). Earnings, Book Values, and Dividends in Equity Valuation.
Contemporary Accounting Research 18 (1): 107-120.

Pope, P. and Walker, M. (1999). International differences in the timeliness, conservatism, and
classification of earnings. Journal of Accounting Research 37: 53–87.

Schatt, A. and Gross, E. (2007). Quelle est l’incidence des normes IAS/IFRS sur les capitaux
propres des entreprises françaises ? Revue Française de Comptabilité 396: 35-39.

Schultze, W. (2005). The information Content of Goodwill- Impairments under FAS 142:
Implications for External Analysis and internal control. Schmalenbach Business Review 57:
276-297.

Sevin, S. and Schroeder, R. (2005). Earnings Management: evidence from SFAS n°142
reporting. Management Auditing Journal 20 (1): 47–54.

Sougiannis, T. (1994). The accounting based valuation of corporate R&D. The Accounting
Review 69: 44-68.

Stolowy, H. & Breton, G. (2003). La gestion des données comptables: une revue de la
littérature. Comptabilité- Contrôle- Audit 9 (1): 125-152.

Zhao, R. (2002). Relative value relevance of R&D reporting: an international comparison.

Journal of International Financial Management and Accounting 13: 153–174.

i
This standard was approved in July 1998 and revised in March 2004. It prescribes the accounting treatment of
intangible assets which are not specifically treated by other standards and applies to expenditure on R&D,
advertising, training, etc.

20
 ii
Asset which could be sold without giving up other company assets.
iii
The fair value is the amount for which the asset could be exchanged, or a liability extinguished, between well-
informed consenting parties and respecting normal competitive conditions (NC 38, § 8).
iv
Regulation concerning accounting regulations and methods applicable to consolidated accounts.
v
This excludes fundamental research costs, training and advertising as well as brands.
vi
In the continuation of the article, ‘other intangible assets’ will be defined as total intangible assets less the
goodwill.
vii
The regulation concerning the publication of information on listed companies varies from one country to
another. Thus, German and Swiss accounting regulations authorize companies listed on the British and American
stock markets to apply solely the IAS/IFRS or US GAAP accounting standards (Bessieux Ollier, 2006).
viii
These are generally companies which have not published their financial statements under international
standards in their annual report, which necessitated research into their accounting data under IFRS by looking in
the BALO.
ix
We specify once again, that total intangible assets are made up of goodwill and other intangibles.
x
The interpretation of this analysis is identical if the accounting data is shown per share and not as a percentage of
total assets.
xi
When the rank-sum of positive differences is higher than the rank-sum of negative differences, the values of
financial data such as the net income, total intangible assets and goodwill, expressed as the percentage of total
assets and valued according to international standards, are higher than those evaluated using French standards.
xii
This model was inspired by theoretical work on evaluation models (Ohlso, 2001).
xiii
Total intangible assets, as a percentage of total assets, are on average 21.11% using French standards and
21.95% using IFRS standards. Companies which have total intangible assets higher than the average are
considered as having a high intensity of intangible assets and vice versa.
xiv
Unidentifiable intangibles such as brands, market share, etc.
xv
This procedure which diagnoses observations to identify atypical points was complemented by the study of
diagrams of standardized residual.
xvi
The indexes of conditioning are all below 5, in other words, well below the critical limit fixed at 30 (Besley,
Kuh and Welsch, 1980).

21

View publication stats