european journal of economics, finance and administrative sciences

1 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)














EUROPEAN JOURNAL OF ECONOMICS, FINANCE
AND ADMINISTRATIVE SCIENCES

ISSN: 1450-2275


Issue 8
October, 2007
2 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


EUROPEAN JOURNAL OF ECONOMICS, FINANCE AND ADMINISTRATIVE SCIENCES
/>

Editor-In-Chief
Adrian M. Steinberg, Wissenschaftlicher Forscher

Editorial Advisory Board
Bansi Sawhney, University of Baltimore
Jwyang Jiawen Yang, The George Washington University
Zhihong Shi, State University of New York
Zeljko Bogetic, The World Bank
Jatin Pancholi, Middlesex University
Christos Giannikos, Columbia University
Hector Lozada, Seton Hall University
Jan Dutta, Rutgers University
Chiaku Chukwuogor-Ndu, Eastern Connecticut State University
Neil Reid, University of Toledo
John Mylonakis, Hellenic Open University (Tutor)
M. Femi Ayadi, University of Houston-Clear Lake
Emmanuel Anoruo, Coppin State University
H. Young Baek, Nova Southeastern University
Jean-Luc Grosso, University of South Carolina Sumter
Richard Omotoye, Virginia State University
Mahdi Hadi, Kuwait University
Jean-Luc Grosso, University of South Carolina
Ali Argun Karacabey, Ankara University
Felix Ayadi, Texas Southern University
Bansi Sawhney, University of Baltimore
David Wang, Hsuan Chuang University
Cornelis A. Los, Kazakh-British Technical University
Leo V. Ryan, DePaul University
Richard J. Hunter, Seton Hall University
Said Elnashaie, Auburn University
Panayiotis Tahinakis, University of Macedonia
Mukhopadhyay Bappaditya, Management Development Institute

M. Carmen Guisan, University of Santiago de Compostela
Subrata Chowdhury, University of Rhode Island
Teresa Smith, University of South Carolina
Wassim Shahin, Lebanese American University
Mete Feridun, Cyprus International University
Teresa Smith, University of South Carolina Sumter
Ranjit Biswas, Philadelphia University
Katerina Lyroudi, University of Macedonia
Maria Elena Garcia-Ruiz, University of Cantabria
Zulkarnain Muhamad Sori, University Putra Malaysia


Indexing / Abstracting
European Journal of Social Sciences is indexed in Scopus, Elsevier Bibliographic Databases,
EMBASE, Ulrich, DOAJ, Cabell, Compendex, GEOBASE, and Mosby.

3 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Aims and Scope
The European Journal of Scientific Research is a quarterly, peer-reviewed international research journal
that addresses both applied and theoretical issues. The scope of the journal encompasses research
articles, original research reports, reviews, short communications and scientific commentaries in the
fields of economics, finance and administrative sciences.

Editorial Policies
1) The journal realizes the meaning of fast publication to researchers, particularly to those working in
competitive & dynamic fields. Hence, it offers an exceptionally fast publication schedule including
prompt peer-review by the experts in the field and immediate publication upon acceptance. It is the
major editorial policy to review the submitted articles as fast as possible and promptly include them in

the forthcoming issues should they pass the evaluation process.
2) All research and reviews published in the journal have been fully peer-reviewed by two, and in some
cases, three internal or external reviewers. Unless they are out of scope for the journal, or are of an
unacceptably low standard of presentation, submitted articles will be sent to peer reviewers. They will
generally be reviewed by two experts with the aim of reaching a first decision within a two-month
period. Suggested reviewers will be considered alongside potential reviewers identified by their
publication record or recommended by Editorial Board members. Reviewers are asked whether the
manuscript is scientifically sound and coherent, how interesting it is and whether the quality of the
writing is acceptable. Where possible, the final decision is made on the basis that the peer reviewers are
in accordance with one another, or that at least there is no strong dissenting view.
3) In cases where there is strong disagreement either among peer reviewers or between the authors and
peer reviewers, advice is sought from a member of the journal's Editorial Board. The journal allows a
maximum of three revisions of any manuscripts. The ultimate responsibility for any decision lies with
the Editor-in-Chief. Reviewers are also asked to indicate which articles they consider to be especially
interesting or significant. These articles may be given greater prominence and greater external
publicity.
4) Any manuscript submitted to the journals must not already have been published in another journal or
be under consideration by any other journal. Manuscripts that are derived from papers presented at
conferences can be submitted even if they have been published as part of the conference proceedings in
a peer reviewed journal. Authors are required to ensure that no material submitted as part of a
manuscript infringes existing copyrights, or the rights of a third party. Contributing authors retain
copyright to their work.
5) The journal makes all published original research immediately accessible through
www.EuroJournals.com without subscription charges or registration barriers. Through its open access
policy, the journal is committed permanently to maintaining this policy. This process is streamlined
thanks to a user-friendly, web-based system for submission and for referees to view manuscripts and
return their reviews. The journal does not have page charges, color figure charges or submission fees.
However, there is an article-processing and publication fee payable only if the article is accepted for
publication.

Submissions
All papers are subjected to a blind peer review process. Manuscripts are invited from academicians,
research students, and scientists for publication consideration. The journal welcomes submissions in all
areas related to science. Each manuscript must include a 200 word abstract. Authors should list their
contact information on a separate paper. Electronic submissions are acceptable. The journal publishes
both applied and conceptual research.
Articles for consideration are to be directed to the editor through In the
subject line of your e-mail please write "EJEFAS submission"
4 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


 Articles are accepted in MS-Word or pdf formats
 Contributors should adhere to the format of the journal.
 All correspondence should be directed to the editor
 There is no submission fee
 Publication fee for each accepted article is $150 USD

European Journal of Economics, Finance and Administrative Sciences is published in the United States

of America at Lulu Press, Inc (Morrisville, North Carolina) by EuroJournals, Inc.

5 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


European Journal of Economics, Finance and Administrative Sciences

Issue 8
October, 2007


Contents

Stock Market Trends, Day of the Week Effect and Investor’s Behavior after
the September’s 2001 Attacks 6-17

Dimitris Balios and Sophia Stavraki

Readability of Financial Statement Footnotes of Kuwaiti Corporations 18-28
Aly M. Hewaidy

An Application of Optimum Currency Area (OCA) Analysis to Egypt 29-38
Sherine El Hag

Openness-Inflation Puzzle for Pakistan: Under Two Alternative Approaches 39-50
Khalil Ahmad and Muhammad Shahbaz

The Institutions and Banking Performance in the OECD 51-59
Mahmoud Khalil, Shereef Ellaboudy and Arthur Denzau

Are Share Issue Privatisations Fairly Priced? 60-68
Panayotis Alexakis

Trends in the Market Growth for Proton Exchange Membrane Fuel Cells (PEMFC):
A Review of the Market Dynamics 69-92

Cihat Polat and Nurcan Kılınç

Highly Leveraged Firms and Corporate Performance in Distressed Industries 93-101
Anna Merika, Theodore Syriopoulos and Christos Negakis

Reengineering America- PART I: The Socioeconomic Costs of Urban Sprawl:
A Socioeconomic Analysis of Exploitative Cross-Industry Diversification Strategies 102-113

Syrous K. Kooros and George Alexakis

Economic Changes that Effect the Probability of Success in R&D 114-127
Nissim Ben-David

Young Adult Perception Towards Celebrity Endorsement:
A Comparative Study of Single Celebrity and Multiple Celebrities Endorsement 128-139

Farida Saleem

European Journal of Economics, Finance and Administrative Sciences
ISSN 1450-2887 Issue 8 (2007)
© EuroJournals, Inc. 2007

Stock Market Trends, Day of the Week Effect and Investor’s
Behavior after the September’s 2001 Attacks


Dimitris Balios
Department of Economics
University of Piraeus 80, Karaoli and Dimitriou st
185 34 Piraeus, Greece
E-mail:

Sophia Stavraki
Department of Economic Sciences
National and Kapodistrian University of Athens
5, Stadiou Str., 2nd floor 105 62 Athens, Greece
E-mail:


Abstract
In this study we examine the existence of the phenomenon «Day of the Week Effect» in six
European Indices after the September’s 2001 attacks in USA, dividing the whole period
according to the stock market’s trend. We conclude that the phenomenon of the «Day of the
Week Effect» seems to be weaker than it was in previous decades as a result of investor’s
behavior. Investors are more mature, well educated, with more professional attitude,
characteristics that help stock markets to become more efficient.


Keywords: Day of the week effect, market efficiency, international financial markets,
stock market trends, investor’s behavior
Jel Classification Codes: G10, G14, G15

I. Introduction
According to Fama (1976), a market is efficient if prices fully and instantaneously reflect all available
information and no profit opportunities are left unexploited. In an efficient situation, new information is
unpredictable, so stock market returns cannot be predicted and there is therefore no trading pattern,
which an investor can follow in order to make unexpected profits.
The day of the week effect refers to the existence of a pattern of stock returns during the week,
a seasonal «anomaly», which contradicts the «Efficient Market Hypothesis». Cross (1973) and French
(1980) were the first to observe a specific seasonality in stock returns during the week, that was named
«Day of the Week Effect». According to this phenomenon, the average stock market return on the last
trading day of the week (Friday) is positive and is the highest across all days of the week and the return
on the first trading day of the week (Monday) is negative and is the lowest across the same period.
As we report in the literature review in the next session, all these studies examine the existence
of the phenomenon in a specific period. This study comes to examine the phenomenon, of course
during a specific period, but in addition, we divide this period in parts according to the market trend.
7 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Balios and Xanthakis examined the existence of the phenomenon for the period January 1994 – August
2001, dividing this period in a sideway period, in an uptrend period and a downtrend one. They found
that the week effect tends to fall away from the classical «Day of the week effect» possibly because
investors adjust their behavior in the previous findings in a way balancing the pattern of returns. In the
sideway period, they find that there is no particular pattern of returns for the indices and the disinterest
in dealing in the stock market to be a reason for this efficient pattern of returns. They supported that in
the uptrend period, investors are very optimistic for the price walk and the companies try to boost this
trend. That’s why they find positive returns (in most of the indices) on Fridays and Mondays something
that maybe is happening because companies announce or are expected to announce their good news
during the week and investors try to take positions from Friday to Monday so that they could profit
from the difference from the next week’s higher expected price. Finally, in the downtrend period, they
find a strong Wednesday negative effect in all indices. They support that in downtrend periods there is

a waiting attitude from investors for good news like the rate of interest which is determined by the
European Central Bank (ECB). They conclude that Wednesday is one day before the bank meeting and
a day that investors could incorporate in their behavior the ECB ‘s expected action which has to do
with the denial to reduce the interest rates for the examined period.
In this study we examine the phenomenon for the period after the September’s 2001 attacks in
the USA for six European Stock Indices. We are interested in finding how investors react after the
attacks, how their trading behavior influences the «Day of the week effect», and finally we compare
these results with the Balios and Xanthakis results.
The rest of the paper is organized as follows. Section II provides the literature review. Section
III contains the data and methodology. Section IV contains the empirical results and in Section V the
conclusions of the paper are reported.


II. Literature Review
The first studies that indicated the day of the week effect were written for the major stock markets, like
the US markets. For the US markets, except from Cross (1973) and French (1980), day effects or daily
anomalies have also been mentioned by Gibbons and Hess (1981), Lakonishok and Levi (1982),
Rogalski (1984), Keim and Stambaugh (1984), Smirlock and Starks (1986), Harris (1986), Lakonishok
and Smidt (1988). Since then, similar results have been found in several stock markets inside and
outside the United States. In this literature review we emphasize in studies after 2001.
In recent studies, several researchers have investigated the seasonality of the day of the week
effect. Chen et al. (2001) examine the day of the week effect in the stock markets of China for the
recent years. They find negative returns on Tuesday. Even though the results of this paper support the
evidence of the day of the week effect, they content that this evidence depends on the estimation
method and sample period. When the transaction costs are taken into account, the opportunity for
arbitrage profits from the day of the week effect seems to be very small. The conclusion is consistent
with the efficient market; there is no clear evidence of the day of the week effect.
Kiymaz and Berument (2003) investigate the day of the week effect on the volatility and return
of major stock markets (German, Japan, US, Canada and United Kingdom) for the time period from
1998 to 2002. Their findings are consistent with the day of the week effect both for returns and

volatility.
In addition, there are many studies that examine the day of the week effect at the emerging
markets. Cabello and Ortiz (2004) investigate the day of the week and month of the year effect for
Latin America stock markets. The paper supports the existence of calendar anomalies. They find the
lowest and negative returns on Mondays and high returns on Fridays.
Hui (2005) examines the day of the week effect at Asian-Pacific markets during the period of
Asian financial crisis and also tests the presence of weekend effect in developed stock markets of US
8 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


and Japan. The paper supports no evidence of the day of the week effect in capital markets for the
recent years, in both Asian Pacific and US capital markets. Holden et al (2005) examine the day of the
week effect for the daily returns of stock market index of Thailand during the period of Asian financial
crisis. This study uses a general model for the test of anomalies included not only the day of the week
effect but also the month of the year and days after holidays and within month effects. The conclusion
of the test performed is that the calendar effects improve the forecast accuracy for daily returns.
Joshi and Bahadur (2006) study the day of the week effect empirically in the Nepalese stock
market for daily data of Nepal Stock Exchange Index from 1995 to 2004. They find persistent evidence
of day-of-the-week anomaly, and they also document no evidence of month-of-the-year anomaly and
half-month effect. Their result for the month-of-the-year anomaly is consistent to the finding observed
for the Jordanian stock market and that for the day-of-the-week anomaly to the Greek stock market.
Patev et al (2003) examine the presence of the day-of-the-week effect anomaly in the Central
European stock markets during the period 1997 to 2002. Their results indicated that the Czech and
Romanian markets have significant negative Monday returns while the Slovenian market has
significant positive Wednesday returns and has non-significant negative returns on Fridays. The Polish
and Slovak markets have no day-of-the week effect anomaly. Lyroudi et al (2002) examine empirically
the day of the week effect anomaly in the Athens Stock Exchange (ASE) for the period 1994 to 1999.
Their results indicated that the day of the week effect is strongly observed in the Greek Stock market
during the second sub period (1996-1999). Chukwuogor-Ndu (2006) reported the existence of the
phenomenon in the BVSP index (Brazil), but not for MRVE (Argentina) and MXSE (Mexico) indices.

III. Data and Methodology
1. Data
The data set used in this study consists of six European Index values. In particular, the six indices used
are, FTSE 100, GDAX 30, CAC 40, Madrid General, MibTel and ASE General corresponding to,
United Kingdom, Germany, France, Spain, Italy and Greece, respectively. The data used in this study
concerns the period from Monday 1st October of 2001 to Friday 2nd February of 2007 and is obtained
directly from their stock exchanges. The examined period consists of 279 weeks, so the created data
series has 1395 observations. For econometric reasons, for working days that the stock markets did not
open and of course the indices did not change, the value of the previous day has been used.
The whole period from October 2001 to February 2007 has been divided into 2 subperiods
based on a different market trend. The trend of an index or a value can be upward, downward or
sideway. An uptrend movement of stock market indices is characterized by an uptrend pattern of prices
that, in our case, can be easily observed in their graphs. Because of the almost same date reversal of
European stock market’s trend, we chose the same sub-periods for the six European indices, so the
observations used for the econometric analysis are the same. The time period and the number of
observations used for each market are shown in Table 1.

Table 1

Period FTSE 100 CAC 40 Dax Mibtel Madrid ASE
Downtrend
1/10/2001-
28/3/2003
1/10/2001-
28/3/2003
1/10/2001-
28/3/2003
1/10/2001-

28/3/2003
1/10/2001-
28/3/2003
1/10/2001-
28/3/2003

(390) (390) (390) (390) (390) (390)
Uptrend
31/3/2003-
2/2/2007
31/3/2003-
2/2/2007
31/3/2003-
2/2/2007
31/3/2003-
2/2/2007
31/3/2003-
2/2/2007
31/3/2003-
2/2/2007
(1395) (1395) (1395) (1395) (1395) (1395)
Notes: Numbers in parentheses depict observations used in each period
9 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


The returns used in each of the time series are computed as follows:
1t
t
t
P

P
nr
−
= l
r
t
: the day return
P
t
: the value of the index
P
t-1
: the value of the index the previous working day

2. Methodology
A test of a day of the week effect is a market efficiency test, which is based on return data series. A
stock value is changing because of the arrival of new information. The way new information is coming
to the market is not constant, so the conditional variance of the daily price change is considered to be
an increasing function of the rate at which new information enters the market. What one must take into
account is that the variance is time dependent and in order to produce the appropriate asymptotic
distributions, one must correct the time series from heteroskedasticity. The best way to correct a model
from the non-constant heteroskedasticity is the GARCH(p,q) models, where p refers to the squared
error terms and q to the lag on variance (Bollerslev, 1986).
The model employed can be written as:
r
t
= b
1
Mo + b
2

Tu + b
3
We + b
4
Th + b
5
Fr + c
i
r
t-i
+ u
t

with u
t
= σ
t
z
t
, z
t
~ i.i.d. with E(z
t
) = 0 and Var(z
t
) = 1 and σ
2
t
= Var(r
t

/I
t-1
) the conditional variance
based on the information set I
t-1
. The use of a lagged observation of the return (if it is useful) happens
in order to correct the model from autocorrelation.
The method used for the estimations was the OLS (Ordinary Least Squares) method and
whenever heteroskedasticity was present, the method used was the maximum likelihood and the
GARCH(p,q) models (the econometric software used for the model estimation was E-views).
In the model, the variables Mo, Tu, We, Th, Fr are the dummy variables of Monday, Tuesday,
Wednesday, Thursday and Friday, respectively. For example, the Mo dummy variable, takes the value
of 1 if the observation falls on Monday and zero in all other cases. The disturbance error term is
depicted by u
t
. The coefficients b
1
to b
5
are the mean returns for the five trading days.
As it has been mentioned before, in an efficient market there will not be a pattern in returns
during the week, so the unexpected return for each day of the week will be the same and close to zero
and the z-statistic measuring the significance of the dummy variables will be small enough for the
dummy variables to be statistically significant. If an estimator is statistically significant (5% level of
significance), we conclude the existence of a day of the week effect which is evidence for inefficiency.


IV. Descriptive Statistics
Table 2 presents descriptive statistics for the whole period data series. FTSE 100 index has the lowest
return (0.018%) and the General Madrid index has the biggest one (0.058%), while the MibTel index

has the lowest standard deviation (1.025%) and the GDAX the biggest standard deviation (1.579%)
among all indices. The kurtosis measures indicate that the return series are leptokurtic compared to the
normal distribution. Table 2 also reports the Augmented Dickey - Fuller statistics for both the
logarithm of the stock prices and the logarithmic first differences (returns). The hypothesis of a single
unit root in the logarithm of the stock prices is accepted, but is strongly rejected in the logarithmic first
differences.
10 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Table 2

Whole period FTSE 100 CAC 40 DAX MIBTEL MADRID ASE
Mean
0.018% 0.024% 0.034% 0.033% 0.058% 0.055%
Median
0.017% 0.025% 0.057% 0.053% 0.071% 0.007%
Maximum
5.904% 7.002% 7.553% 5.292% 4.857% 4.974%
Minimum
-5.589% -6.045% -6.336% -4.416% -4.339% -6.107%
Std. Dev.
1.068% 1.381% 1.579% 1.025% 1.080% 1.046%
Skewness
-0.135 0.056 0.022 -0.026 0.028 -0.081
Kurtosis
7.755 6.794 6.135 6.041 5.496 5.042
Jarque – Bera
1317.714 837.095 571.263 537.411 362.052 243.823
ADF (levels)
-0.178 -0.032 0.377 0.531 1.858 1.237

ADF (returns)
-18.617** -19.183** -18.309** -17.327** -17.237** -15.441**
** denotes significance at the 1% level of significance

In Table 2A we can study descriptive statistics for the downtrend period data series. GDAX
index has the lowest return (-0.137%) and the biggest standard deviation (2.432%) among all indices.
For this subperiod too, the hypothesis of a single unit root in the logarithm of the stock prices is
accepted, but is strongly rejected in the logarithmic first differences.

Table 2A

Downtrend period FTSE 100 CAC 40 DAX MIBTEL MADRID ASE
Mean
-0.072% -0.103% -0.137% -0.059% -0.035% -0.098%
Median
-0.068% -0.107% -0.059% -0.071% 0.000% -0.072%
Maximum
5.904% 7.002% 7.553% 5.292% 4.857% 4.021%
Minimum
-5.589% -6.045% -6.336% -4.416% -4.339% -3.007%
Std. Dev.
1.660% 2.147% 2.432% 1.579% 1.642% 1.109%
Skewness
0.041 0.250 0.195 0.258 0.277 0.365
Kurtosis
4.366 3.773 3.393 3.333 3.140 3.898
Jarque – Bera
30.371 13.772 4.985 6.132 5.304 21.708
ADF (levels)
-0.592 -0.438 -0.043 -0.662 -0.929 0.122

ADF (returns)
-10.216** -10.438** -9.905** -9.456** -9.645** -7.788**
** denotes significance at the 1% level of significance

Table 2B presents mean returns, standard deviations and other descriptive statistics for the
uptrend period data series. ASE index has the biggest return (0.114%) and GDAX has the biggest
standard deviation (1.075%) among all indices. For the uptrend period too, the hypothesis of a single
unit root in the logarithm of the stock prices is accepted, but is strongly rejected in the logarithmic first
differences.

Table 2B

Uptrend period FTSE 100 CAC 40 DAX MIBTEL MADRID ASE
Mean
0.053% 0.073% 0.100% 0.068% 0.094% 0.114%
Median
0.056% 0.081% 0.110% 0.095% 0.114% 0.082%
Maximum
3.133% 4.047% 5.671% 2.778% 3.026% 4.974%
Minimum
-2.963% -4.353% -4.741% -3.879% -4.124% -6.107%
Std. Dev.
0.716% 0.925% 1.075% 0.698% 0.755% 1.016%
Skewness
-0.213 0.073% 0.100% 0.068% 0.094% 0.024%
Kurtosis
4.529 4.982 5.483 5.666 5.172 5.815
Jarque – Bera
105.02 176.27 257.04 370.46 233.37 342.45
ADF (levels)

-1.109 -1.105 -1.499 -0.817 0.037 -1.451
ADF (returns)
-14.936** -15.971** -15.361** -14.149** -13.585** -13.632**
** denotes significance at the 1% level of significance

11 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Very interesting are the descriptive statistics for each day of the week. Tables 3A and 3B
present mean returns and standard deviations respectively, for each day of the week for the whole
period data series. Mean returns are the highest on Thursday for all stock indices, while Monday has
the lowest returns for all indices, too. A summary of these results is displayed in table 3. From equality
tests for the mean returns across all days of the week, we don’t conclude the rejection of the hypothesis
that mean returns among the days of the week are equal, except from ASE General. Table 3B presents a
summary of the standard deviations. It is observed the highest standard deviation on Monday for all
indices and the lowest on Friday for all indices, too. The Levene test, which measures the equality of
variance in returns, shows that risk is distributed equally among the weekdays except for MibTel.

Table 3A: Mean returns on common stocks indices by day of the week for the whole examined period

Monday Tuesday Wednesday Thursday Friday Average F-stat p-value
FTSE 100
-0.077% -0.023% -0.075% 0.166% 0.102% 0.051% 1.528 0.191
CAC 40
-0.109% -0.025% -0.012% 0.153% 0.077% 0.072% 1.172 0.321
DAX
-0.138% -0.048% -0.007% 0.146% 0.044% 0.099% 0.566 0.687
MIBTEL
-0.070% 0.002% 0.046% 0.100% 0.071% 0.068% 0.961 0.427
MADRID

-0.046% 0.008% 0.087% 0.141% 0.072% 0.094% 0.954 0.431
ASE
-0.081% -0.006% 0.085% 0.166% 0.143% 0.113% 3.353 0.010
The F-Stat refers to the F-Statistic of the Equality of means test.
If p-value < 0.050, then the hypothesis of equal means is rejected

Table 3B: Standard deviation on common stocks indices by day of the week for the whole examined period

Monday Tuesday Wednesday Thursday Friday Average L-Value p-value
FTSE 100
1.662% 0.982% 1.066% 1.151% 0.945% 0.717% 1.478 0.206
CAC 40
2.159% 1.315% 1.302% 1.489% 1.226% 0.927% 2.309 0.056
DAX
2.352% 1.447% 1.569% 1.592% 1.457% 1.077% 2.371 0.050
MIBTEL
1.603% 0.968% 0.946% 1.089% 0.888% 0.699% 4.640 0.001
MADRID
1.700% 1.008% 1.070% 1.098% 1.003% 0.757% 1.978 0.095
ASE
1.131% 1.031% 1.031% 1.053% 0.926% 1.018% 1.082 0.363
The L-Value refers to the Levene Value of the Equality of variance test. If p-value < 0.050, then the hypothesis of equal variances is rejected

Tables 3C and 3D present the same descriptive statistics but for the downtrend period data
series this time. Mean returns are lowest on Wednesday for FTSE 100, while for the rest stock markets,
mean returns are lowest on Monday. Standard deviations are highest on Monday for FTSE 100, GDAX
and General Madrid index, on Thursday for CAC 40 and MibTel, while ASE has the highest standard
deviation on Wednesday. From equality tests across all days of the week we accept the null hypothesis
that means and standard deviations are equal in each subgroup.

Table 3C: Mean returns on common stocks indices by day of the week for the downtrend examined period

Monday Tuesday Wednesday Thursday Friday Average F-stat p-value
FTSE 100
-0.241% -0.135% -0.333% 0.270% 0.081% -0.072% 1.716 0.145
CAC 40
-0.398% -0.165% -0.270% 0.333% -0.014% -0.103% 1.345 0.252
DAX
-0.339% -0.202% -0.294% 0.232% -0.084% -0.137% 0.686 0.602
MIBTEL
-0.295% -0.019% -0.194% 0.180% -0.035% -0.060% 0.345 0.847
MADRID
-0.244% -0.058% -0.073% 0.214% -0.015% -0.035% 0.232 0.920
ASE
-0.319% -0.124% -0.100% 0.105% -0.053% -0.098% 1.480 0.207
The F-Stat refers to the F-Statistic of the Equality of means test.
If p-value < 0.050, then the hypothesis of equal means is rejected

12 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Table 3D: Standard deviation on common stocks indices by day of the week for the downtrend examined
period

Monday Tuesday Wednesday Thursday Friday Average L-Value p-value
FTSE 100
1.830% 1.507% 1.618% 1.812% 1.469% 1.660% 0.884 0.473
CAC 40
2.317% 2.137% 1.950% 2.376% 1.891% 2.147% 1.975 0.098
DAX

2.710% 2.340% 2.373% 2.479% 2.251% 2.432% 0.820 0.512
MIBTEL
1.510% 1.566% 1.389% 1.627% 1.384% 1.579% 0.334 0.854
MADRID
1.783% 1.612% 1.555% 1.745% 1.506% 1.642% 1.496 0.202
ASE
1.113% 1.104% 1.121% 1.100% 1.093% 1.109% 0.082 0.987
The L-Value refers to the Levene Value of the Equality of variance test. If p-value < 0.050, then the hypothesis of equal variances is rejected

Studying Tables 3E and 3F which present mean returns and standard deviations for each day of
the week for the uptrend period data series we conclude that mean returns are highest on Friday for
FTSE100, CAC and ASE, on Monday for GDAX, while for the rest stock markets, mean returns are
highest on Wednesday. Standard deviations are highest on Monday for all stocks. From equality tests
across all days of the week we accept the null hypothesis that means are equal in each subgroup, but we
reject it for the standard deviations as it does not hold for DAX, MibTel and General Madrid.

Table 3E: Mean returns on common stocks indices by day of the week for the uptrend period

Monday Tuesday Wednesday Thursday Friday Average F-stat p-value
FTSE
0.068% 0.021% 0.026% 0.039% 0.111% 0.053% 0.677 0.607
CAC 40
0.051% 0.029% 0.087% 0.084% 0.112% 0.073% 0.380 0.822
GDAX
0.177% 0.012% 0.105% 0.113% 0.093% 0.100% 0.675 0.609
MIBTEL
0.038% 0.010% 0.139% 0.069% 0.085% 0.068% 1.685 0.151
GMADRID
0.071% 0.033% 0.149% 0.113% 0.105% 0.094% 0.864 0.485
ASE

-0.034% 0.039% 0.159% 0.190% 0.219% 0.114% 1.922 0.104
The F-Stat refers to the F-Statistic of the Equality of means test.
If p-value < 0.050, then the hypothesis of equal means is rejected

Table 3F: Standard deviation on common stocks indices by day of the week for the uptrend period

Monday Tuesday Wednesday Thursday Friday Average L-Value p-value
FTSE 100
0.775% 0.077% 0.732% 0.750% 0.641% 0.716% 1.424 0.223
CAC 40
1.093% 0.798% 0.925% 0.944% 0.842% 0.925% 2.334 0.053
GDAX
1.282% 0.890% 1.100% 1.074% 0.995% 1.075% 2.503 0.040
MIBTEL
0.867% 0.599% 0.686% 0.704% 0.599% 0.698% 3.008 0.017
GMADRID
0.881% 0.639% 0.804% 0.708% 0.7225 0.755% 2.474 0.042
ASE
1.172% 1.000% 0.987% 1.036% 0.843% 1.016% 1.022 0.394
The L-Value refers to the Levene Value of the Equality of variance test. If p-value < 0.050, then the hypothesis of equal variances is rejected


V. The Results
The results of the regressions are presented in tables 4A, 4B and 4C. The estimation of the regressions
has been done by maximum likelihood method because heteroskedasticity was always present. The lag
lengths used for the conditional mean were determined on the basis of the Schwarz and Akaike
criterion. To avoid autocorrelation, the use of lagged return observations was necessary (in some of the
regressions).
As we examine the results of the whole period (Table 4A), not only we don’t find strong
evidence for the existence of the classical day of the week effect, but there is no any obvious pattern in

coefficient’s significances. For FTSE 100, DAX and CAC, the only statistical significant coefficients
are that of FTSE 100 on Friday and of DAX on Monday. Wednesday is positive for MibTel and for
General Madrid and ASE General there are positive, statistically significant coefficients on
13 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Wednesday-Thursday and Thursday-Friday respectively. All the significant coefficients are positive.
There is no classical version of the day of the week effect (Fridays with positive returns and Mondays
with negative returns and no substantial day effect for the developed stock markets.

Table 4A: Regression results - whole period

Monday Tuesday Wednesday Thursday Friday c
1

0.067 0.009 0.018 0.062 0.101 -9.140
FTSE 100
1.651 0.225 0.406 1.364 2.451* -3.317**
0.141 0.020 0.112 0.114 0.053
DAX
2.279* 0.359 1.868 1.867 0.819
0.063 0.019 0.073 0.115 0.071
CAC
1.163 0.368 1.350 1.855 1.247
0.031 0.021 0.157 0.068 0.080
MIBTEL
0.622 0.512 3.724** 1.506 1.804
0.084 0.038 0.150 0.131 0.086
MADRID
1.649 0.935 3.120** 2.830** 1.651

-0.070 0.028 0.104 0.183 0.123 8.559
ASE
-1.220 0.471 1.801 3.170** 2.467* 3.182**
* denotes significance at the 5% level of significance
** denotes significance at the 1% level of significance

Of particular interest are the findings for the downtrend period (Table 4B). There are only two
statistically significant negative coefficients. Only the Tuesday returns for FTSE 100 and the Monday
returns for ASE General are statistically significant. Not only we can not find strong evidence for the
existence of specific day of the week effect for the downtrend period, but the markets, in a risky period
because of the economic environment that after the September’s 2001 attacks has been configured, look
they work efficiently.

Table 4B: Regression results – downtrend period

Monday Tuesday Wednesday Thursday Friday c
1

-0.086 -0.222 -0.134 0.194 0.009
FTSE
-0.690 -1.807* -0.951 1.376 0.075
-0.082 -0.200 -0.017 0.160 -0.122
DAX
-0.349 -1.104 -0.073 0.731 -0.591
-0.105 -0.215 -0.123 0.266 -0.087
CAC
-0.556 -1.064 -0.710 1.253 -0.580
-0.233 -0.013 -0.076 0.118 0.037
MIBTEL
-1.355 -0.074 -0.495 0.694 0.279

-0.189 -0.060 0.024 0.175 -0.027
MADRID
-1.046 -0.381 0.154 1.038 -0.182
-0.291 -0.172 -0.088 0.076 -0.075
ASE
-2.457* -1.355 -0.697 0.655 -0.646
* denotes significance at the 5% level of significance
** denotes significance at the 1% level of significance

Examining the uptrend period (results in Table 4C) we find a few statistically significant
positive coefficients. Just like the whole period, for FTSE 100, DAX and CAC, the only statistically
significant coefficients are that FTSE 100 on Friday and that of DAX on Monday. The results are
different for the less developed stock markets than these of the previous ones. For Madrid General, the
coefficients that are statistically significant are these of Monday, Wednesday, Thursday and Friday
which have a positive sign, as the returns Wednesday, Thursday and Friday for ASE General. Finally,
Wednesday is positive for MibTel. We conclude that, in the uptrend period, there is no classical day of
the week effect. The results from the most developed stock market’s indices are an evidence for market
14 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


efficiency as no trading pattern can be created. From the less developed stock market’s indices like
General Madrid and ASE General, the uptrend movement of the prices seems that have influence in the
same way almost the whole week. In general, the results of the uptrend period seem that have
influenced these of the whole period as it’s almost the same.

Table 4C: Regression results – uptrend period

Monday Tuesday Wednesday Thursday Friday c
1

0.085 0.028 0.037 0.052 0.107 -9.433
FTSE
1.889 0.663 0.797 1.085 2.447* -2.956**
0.163 0.037 0.121 0.113 0.070
DAX
2.540* 0.643 1.917 1.775 1.056
0.084 0.047 0.093 0.110 0.096 -7.522
CAC
1.415 0.919 1.630 1.820 1.587 -2.426*
0.064 0.025 0.166 0.072 0.087
MIBTEL
1.219 0.637 3.828** 1.543 1.877
0.136 0.053 0.138 0.142 0.108
MADRID
2.683** 1.270 2.766** 3.001** 2.066*
0.006 0.085 0.149 0.211 0.188 6.935
ASE
0.086 1.283 2.351* 3.125** 3.400** 2.162**
* denotes significance at the 5% level of significance
** denotes significance at the 1% level of significance

The statistical significance of mean returns indicates that past returns influence the current
returns. If the past returns influence the current returns, then there is violation of the Martingale
Hypothesis, according to which, information from past returns is useless in predicting future returns. If
this violation is true, it constitutes evidence of market inefficiency. The Martingale Hypothesis is
violated only for the case of FTSE and ASE General in the whole and in the uptrend periods (and CAC
in the uptrend too). In all the other indices the regression results show that past returns do not have any
influence on future returns.

VI. The Conclusions
In this study we examine the existence of the phenomenon «Day of the Week Effect» in six European
Indices after the September’s 2001 attacks in USA, dividing the whole period according to the stock
market‘s trend.
Balios and Xanthakis didn’t find specific day of the week effect for the period January 2004 to
August 2001 for the same indices, but examining the same period based on the stock market’s trend
they concluded that in four out of six indices there is no day effect in the sideway period, there are
many statistically significant positive coefficients in the uptrend period and a strong negative
Wednesday effect in all the examined indices in the downtrend period.
As we examine the same phenomenon for the period October 2001 – February 2007 we
conclude that not only there is no classical day of the week effect but the phenomenon seems to
disappear from the developed stock markets and not to have a specific pattern in general.
Nowadays, the stock markets are more liquid than ever and seem to be more efficient that the
previous decades because of the easiest capital transmission, the technological changes and the changes
in the stock market microstructure. So, it is logical for investors to react more mature, something that
induces less inefficient results. This is clear for the more developed stock markets and the indices of
DAX, FTSE and CAC as we consider the coefficient’s statistical significances and less clear for the
indices of MIBTel, General Madrid and ASE General.
Investor’s behavior looks more mature as we study the results of the downtrend and the uptrend
period and we compare them with the results of Balios and Xanthakis. Before the September’s 2001
15 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


attacks, in the downtrend period, there was a negative Wednesday effect. Balios and Xanthakis support
that this finding is a result of the waiting attitude of the investors for good news and especially the rate
of interest. After September 2001, in the downtrend period, we don’t find any effect during the week
(except for ASE on Monday). This period is characterized by hesitance and wariness because investors
are not interested only to clear economic events but to political – terrorism news too. So investor’s
behavior is characterized by behavioral trading, no specific trading strategy and a waiting attitude for
political – economical stability. As a result, there is no day of the week effect, something that changes

in the uptrend period.
When investors feel more secure about the political – economical environment, there optimism
boosts prices to higher levels and most of the trading sessions have a positive sign at the end of the day.
Our findings fit with Balios and Xanthakis results as the indices of the developed stock markets do not
have (or have at the most one) day effect in the uptrend period. In less developed stock markets the
phenomenon exists but not in a specific pattern.
Finally, we conclude that the phenomenon of the «Day of the Week Effect» seems to be weaker
than it was in previous decades as a result of investor’s behavior. Investors are more mature, well
educated, with more professional attitude, characteristics that help stock markets to become more
efficient.
16 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


References
1] Akaike, H. (1973), “Information theory and an extension of the Maximum Likelihood
Principle”, in B.N. Petrov and F. Csaki (eds), 2
nd
International Symposium on Information
Theory (Budapest: Academia Kiado)
2] Balios, D., E. Xanthakis (forthcoming). “Day of the week effect: New evidence from European
Stock Markets”. Essays in Honour of Professor Naoum, University of Piraeus.
3] Bollerslev, T. (1986). “Generalized Autoregressive Conditional Heteroskedasticity”. Journal of
Econometrics 31: 307 – 327.
4] Cabello, A., Ortiz E. (2004). “Day of the week and month of the year effects at the Latin
American Emerging markets”. International Finance Review 5: 273-304.
5] Chen G., Kwok C.C.Y., Rui O.M. (2001). “The day of the week regularity in the stock markets
of China”. Journal of Multinational Financial Management 11: 139-163.
6] Chukwuogor-Ndu, C. (2006). “Stock market annual returns patterns, day of the week effect, and
volatility: Evidence from three Latin American Financial Markets”. European Journal of
Economics, Finance and Administrative Sciences 5: 20-30

7] Cross, F. (1973). “The behaviour of stock prices on Fridays and Mondays”. Financial Analysts
Journal 29: 67-69.
8] Dickey, D.A. and W.A. Fuller (1979), “Distribution of the Estimators for Autoregressive Time
Series with a Unit Root”, Journal of the American Statistical Association, 74: 427–431
9] Dickey, D.A. and W.A. Fuller (1981), “Likelihood Ratio Statistics for Autoregressive Time
Series with a Unit Root”, Econometrica, 49: 1057-1072
10] Fama, E. (1976). “Efficient Capital Markets: Reply”. Journal of Finance 3: 143-145.
11] French, K. (1980). “Stock Returns and the Weekend Effect”. Journal of Financial Economics 8:
55- 69.
12] Gibbons, R.S., P. Hess, (1981), “Day of the Week Effects and Asset Return”, Journal of
Business, 54, 579-96.
13] Harris, L. (1986), “A transaction Data Study of Weekly and Intra daily Patterns in Stock
Returns”, Journal of Financial Economics, 16, 99-117.
14] Holden, K., J. Thompson and Y. Ruangrit (2005). ”The Asian crisis and calendar effects on
stock returns in Thailand”. European Journal of Operational Research, May 2005.
15] Hui, T.K. (2005). “Day of the week effects in US and Asia-Pacific stock markets during the
Asian financial crisis”. Omega 33: 277-282.
16] Joshi, N.K., F.K.C. Bahadur (2006). “The Nepalese Stock Market: Efficiency and Calendar
Anomalies”. Economic Review 17.
17] Keim, B.D., R. F. Stambaugh, (1984), “A further investigation of the weekend effect in stock
returns”, Journal of Finance, 39, 819-840.
18] Kiymaz, H. and H. Berument (2003). ”The Day of the week effect on Stock Market volatility
and volume: International evidence”. Review of Financial Economics, January 2003.
19] Lakonishok, J., M. Levi, (1982), “Week-end effects on stock returns: a note”, Journal of
Finance, 37, 883-89.
20] Lakonishok, J., S. Smidt, (1988), “Are seasonal anomalies real? A ninety-year perspective”,
Review of Financial Studies, 1, 403-25.
21] Levene, H. (1960), “Robust tests for equality of variances in contribution to probability and
Statistics”, (Ed) 1, Olkin: Stanford University Press, Palo Alto.
22] Lyroudi, K., D. Subeniotis, G. Komisopoulos (2002). “Market Anomalies in the A.S.E.: The

Day of the Week Effect”. European Financial Management Association, London Meetings.
23] Patev, P.G., K. Lyroudi and N.K. Kanaryan, “The Day of the Week Effect in the Central
European Transition Stock Markets”. Tsenov Academy of Economics Finance and Credit 03-06.
24] Rogalski, R. (1984), “New findings regarding day of the week returns over trading and
nontrading period”, Journal of Finance, 39, 1603-14.
17 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


25] Schwarz, G. (1978), “Estimating the Dimension of a Model”, Annals of Statistics, 6, pp 461-464
26] Smirlock, M., L. Starks, (1986), “Day of the week and intraday effects in stock returns”,
Journal of Financial Economics, 17, 197-210.

European Journal of Economics, Finance and Administrative Sciences
ISSN 1450-2887 Issue 8 (2007)
© EuroJournals, Inc. 2007



Readability of Financial Statement Footnotes of Kuwaiti
Corporations


Aly M. Hewaidy
Associate Professor of Accounting, Accounting Department
College of Business Administration Kuwait University, Kuwait
E-mail:


Abstract
This study examines the readability of the Kuwaiti corporate annual report footnotes and

whether their readability levels differ significantly among user groups according to their
educational background. The results revealed that footnotes to the financial statements of
Kuwaiti corporations are difficult to read and understand, and their readability levels vary
significantly among different groups of financial report users. The findings of the present
study are in line with the studies undertaken in western countries which reported that
annual reports are difficult for most users to understand. These findings encourage more
attention to improve the communicative ability of the corporate annual reports. They also
encourage further research in which additional methods of communicating corporate annual
report -such as graphs and visual methods- would be examined as to whether they improve
the communication function of corporate annual reports.


Introduction
Since the absence of a common meaning between producer and user may have damaging decision-
relevance consequences, the measurement of meaning is an important research area in accounting
(Smith and Taffler, 1992a, p.84). To alleviate this potential problem, it has been argued that corporate
reports should be communicated in such a way that makes them as readable as possible. The SEC
released an updated staff legal bulletin regarding plain English disclosures. The plain English rule is
designed to make prospectuses easier to read by requiring issuers to write in a clear and understandable
manner (1999, p.69).
Accounting research into the readability of annual reports was first published in 1952. Over the
past 50 years annual reports readability has been investigated in the USA (Kinnersley and Fleischman,
2001; Pashalian and Crissy, 1952; Schroeder and Gibson, 1990; Smith and Smith, 1971), the UK
(Jones, 1988; Smith and Taffler, 1992a, 1992b), Canada (Courtis, 1986, 1995), Australia (Lewis et al.,
1986; Parker, 1982; Pound, 1980, 1981) and New zeland (Healy, 1997). Findings have consistently
revealed annual report prose passages to be at a reading ease level of difficult to very difficult.
The purpose of the present research is to extent the literature on this topic by examining the
readability of the footnotes to financial statements in Kuwaiti corporate annual reports. That is, the
research is undertaken in an environment which differs from that in which previous readability studies
were undertaken. Thus, the findings of the present research are expected to contribute to the accounting

research generally and to the accounting practice in Kuwait and other Middle Eastern countries in
particular.
19 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Organization of the Research
The next section reviews alternative measures of readability, and the prior research in the area of
financial reporting. This is followed by an explanation of the research methodology. Next, the results
are presented and discussed. Finally the study concludes with a summary of the findings and their
implications, together with directions for further research.


Readability and its Measurement: Literature Review
Readability Measurement
Readability is generally considered to be one of the most important characteristics of effective
reporting. Alternative formulas have been adopted to measure readability. Most of them attempt to
assess the reading age required to understand a particular written passage, or a degree of difficulty
associated with a particular level of education (Lewis et al., 1986, p. 200). In assessing the reading
difficulty, readability formulas concentrate on characteristics of writing style. Generally, they are based
on two features: Word length - related to speed of recognition; and Sentence length - related to a recall
of words in the immediate memory (Smith & Taffler, 1992a, p. 86).
A review of the literature reveals that, the most common readability formulas are Dale-Chall,
Flesch, Fog, and Lix. These formulas are described below with the aim of determining their relevance
to the measurement of readability of Kuwaiti corporate reports (Lewis et al., 1986; Smith & Taffler,
1992; and Stevens et al, 1992).

Dale-Chall
Dale-Chall Index = 15.79 (U/W) + 0.0496 (W/S)
Where: U = number of unfamiliar words, W = number of words and S = number of sentences. The
formula is based on an assumption that readability difficulty is a function of two factors: unfamiliar

words and sentence length. Therefore, the higher the score the more difficult the passage.

Flesch
Flesch Index = 206.385 - (0.846 X L/100W) - (1. 015 X W/S)
Where: L/100W = word length (number of syllables per 100 words) and (W/S) = sentence length
(average number of words per sentence).The Flesch formula is simple to calculate. The calculation
represents a deduction from the base constant for both word and sentence complexity, so that the higher
the score the easier the readability.

Fog
Fog Index = 40 (P/W) + 0.4 (W/S)
Where: P/W = percentage of “hard” word in the passage and W/S = sentence length (average number
of words per sentence). The Fog formula is similar to the Flesch approach except that it is based on the
percentage of polysyllabic words (i.e. words of three or more syllables) in a passage. The measure is
therefore a function of sentence length and percentage of hard words. The calculation is incremental
with increased complexity so that the higher the score the more difficult the readability.

Lix
Lix Index = 100 (B / W) + 1.0 (W / S)
Where: B/W = percentage of words of seven or more letters, and W/S = sentence length (average
number of words per sentence). The aggregate score is compared with a scale of difficulty ranging from
20 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


20 represents “very easy”, through stages to 60 representing “very difficult”. (Lewis et al., 1986, p.
207).

The Cloze Test
The above formulas highlight the importance of word and sentence factors to the measurement of
readability. The emphasis on words and sentences length, without reference to context, poses certain

problems and limitations. Several authors have argued that measures based upon word and sentence
length may provide misleading results. Lewis et. al (1986, p. 202) stated that, measures based upon
word and sentence difficulty may be an oversimplification. One or both of these variables may be
unreliable as an indicator of readability. That is, reducing sentence length will not necessarily improve
readability. Lesikar and Lyons (1986, p. 151) highlight the use of the passive voice as a factor known to
affect readability but argue that they are reflected in traditional formulae measures only to the extent of
the longer sentence invariably generated.
Moreover, these measures do not take account of the reader’s interest level. Reading studies
have clearly demonstrated that a high level of motivation towards the subject matter of the text has a
markedly positive effect on the apparent reading competence of a poor reader. Also, it is unlikely that
these measures account for the impact of reading speed which is a function of time required to assess
issues in the text, breadth of the reader’s previous knowledge and organization of material in the text
being read (Lewis et. al, 1986, p. 202).
These problems and limitations suggest that words and sentences length are not the only
variables affecting readability, and simplification alone does not automatically guarantee improved
understandability. Consequently, another measure is required. Reading studies highlight the reliability
and usefulness of the Cloze Test in measuring readability. It goes beyond word and sentence length to
reflect the interaction of several elements influencing readability. Such elements as word meaning,
language patterns, reader’s language ability and interests.
The Cloze Test involves the mutilation of a narrative passage by the deletion of every nth word
and its replacement by a blank space. Respondents are then required to predict the correct insertion for
the blank based on the surrounding context (Smith & Taffler, 1992a, p. 87). A readability measure is
then derived by calculating the percentage of words correctly predicted by each respondent.
The Cloze Test has been used to assess readability of accounting texts. Steven et al. (1983) used
it to measure the readability of Financial Accounting Standards Board Statement 33: “Financial
Reporting and Change Prices”. They emphasized the reliability of their procedure in comparison to
those based on sentence and word difficulty. Adelberg and Razek (1984) applied the Cloze procedure
to discriminate between accounting textbooks and recommended its further validation with accounting
communications. Several other studies, notably Smith & Taffler (1992a), emphasize the validity and
reliability of the Cloze procedure.

Litrature Review
Using one or more of the preceding measures, there have been several readability studies published in
the accounting literature. They covered most of the narrative disclosure areas. They also examined
trends in readability over time. Generally, these studies concluded that corporate reporting was couched
in an academic, technical style that the unsophisticated reader would find difficult, or very difficult, to
read and understand.
Studies of footnotes to financial statements found that footnotes are difficult to read. Smith and
Smith (1971) found that less than 20 percent of the US adult population had achieved an educational
level sufficient to comprehend the message appearing in 86 per cent of the notes to financial
statements. Healy (1977) examined the reading ease of footnotes contained within 50 New Zealand
public corporations. He found that the majority of footnotes to financial statements are of low
21 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


readability, and this prevents many investors from gaining the information that may be necessary for
rational economic decision making.
Readability of other parts of annual reports was noted to be no better than readability of the
footnotes. Schroeder and Gibson (1990) in the USA found the level of readability of footnotes and
management’s discussion and analysis to be similar, both significantly more difficult than that of the
President’s Letter. Still (1972) found that 77 per cent of a sample of 50 UK Chairman’s Statements
required a level of reading ability which was beyond that of 80.7 per cent of US adults. Pound (1981)
also examined readability of the audit report, and concluded that audit reports require a level of
education of at least university undergraduate in order to comprehend the messages contained in those
reports.
Studies of readability of annual reports over two or more periods found that corporate reports
had become more difficult to read over time. Lewis et al. (1986) using a sample of Australian reports to
employees, concluded that “reports to employees, while supposedly simplified for ease of lay reader
understanding, were still pitched at a “difficult to read” level, and the readability of the reports did not

appear to improve noticeably over time. Jones (1988) examined the readability of the chairman’s
narrative of a UK company from 1952 to 1985, and found that readability had declined significantly
over this period.
It is apparent that previous research has been undertaken in the English speaking countries. The
different language, legal, economic and social situations in these countries as compared to the Arab
environments may impair the applicability of their findings to an Arab country. This highlights the
importance of and the need for the present research.


Research Methodology
Choice of Readability Measurement
As mentioned above, alternative methods have been adopted to measure readability. For the purpose of
the present study, the Cloze Test was used. The choice of this measure is conditioned by the limitations
associated with other measures and its relevance and applicability to measure the readability of an
Arabic text.
The method of applying the Cloze test is defined as “a method of intercepting a message from a
transmitter (writer or speaker), mutilating its language pattern by deleting parts, and so administering it
to receivers (readers or listeners) that their attempts to make patterns whole again potentially yield a
considerable number of Cloze units. Thus, a close unit is any single occurrence of a successful attempt
to reproduce accurately a part deleted from a message by deciding, from the text that remains, what the
missing part should be (Adelberg & Razek, 1984, p.111).
The Cloze test has several advantages over alternative techniques. Among these advantages are:
(1) Cloze tests are economical, easy to prepare, and easy to score. They provide valid, reliable and
easily interpretable results. (2) Because Cloze tests are prepared entirely from a set of standard
mechanical operations, they are not subject to conscious or unconscious biases of the investigator. (3)
Cloze tests measure the reader’s ability to predict the precise word used by the writer. This ability is
indicative of the reader’s understandability of the writer’s total meaning (Adelberg and Razek, 1984, p.
115).
In measuring the readability of an Arabic text, the Cloze Test seems to be more applicable than
those techniques based on word and sentence difficulty. As an example, an accounting term like

“accounts receivable” can be written down in Arabic in several ways. It can be written using a word of
5 letters or a word of 8 letters. Although the two words have the same meaning, the latter would be
considered as a hard word according to Lix formula. Moreover, using number of syllables as a measure
of words length would be difficult in Arabic language due to the difficulty associated with the division
of words into syllables. Moreover, calculation of number of syllables of an English texts can be easily
22 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


undertaken due to the availability of microcomputer software packages, which is not the case with an
Arabic texts.
To support the results obtained by the Cloze Test, another test called reading Comprehension
test was used to measure the interaction between specific readers and specific messages. This test is
assumed to determine the validity of the Close test as an instrument of readability measurement.

Sampling Methodology
The Corporations Tested
Out of the annual reports of the corporations listed in the Kuwaiti Stock Exchange, two were selected
given the relative importance of the Kuwaiti economic sectors. As financial and real estate sectors
dominate the economy, a bank and a real estate corporation were chosen: the National Bank of Kuwait
(NBK) and Kuwait Real Estate Corporation (KREC). In terms of total assets, both NBK and KREC
represent the largest corporation in their sector.
It may be argued that the sample is small, and this would bias the results. Choice of the small
sample was due to the time factor associated with the application of the Cloze procedure. Processing a
passage of 300 words needs about 30 to 35 minutes. Given that, any attempt to test more than two
reports would be difficult and cause respondent fatigue. Therefore, two reports were found to be
relevant.

The Massage Tested
Corporate annual reports consist of several parts (e.g. financial statements, footnotes to the financial
statements, auditor’s report, management’s discussion and analysis). Footnotes represent a large part of

a corporate report and crucial to understanding the figures in the balance sheet and the income
statement. Given that, footnotes to the financial statements of both NBK and KREC were selected for
the purpose of readability measurement. To facilitate the application of both the Cloze and
Comprehension test, footnotes were produced in the form of paragraphs. The resulting paragraphs were
retyped and mutilated so as to every fifth word was deleted and replaced by a dotted line (Adelberg and
Razek, 1984, p. 113). The deletion commenced at the fifth word in the passage and continued at equal
intervals throughout except for the incidence of proper name, dates and numbers, which are
disregarded.
Smith and Taffler (1992a) found that deletions in excess of 100 took too long to process and
caused respondent fatigue. Taylor (1956) has suggested that 50 deletions provide a sufficient sample
for a stable and reliable score (see Adelberg and Razek, 1984, p. 113). Miller and Coleman (1967) used
30 deletions in each of their validation studies (see Smith & Taffler, 1992a, p. 88). In the present study
50 deletions were used in each passage. To facilitate tabulating the results, each deleted word carried
two marks (2 for correct and 0 for wrong). Thus, the maximum score that each subject could attain
ranged between zero and 100 marks.
Regarding the reading comprehension test 50 true-false questions were assigned on each
passage. Each question carried two marks (2 for correct and 0 for wrong). Thus, the maximum score
that each subject could attain ranged between zero and 100 marks.

The Subject Tested
Four hundred students from Kuwait University participated in the field work. They were equally
distributed over business and non-business colleges (i.e. each group consists of 200 students). Students
in the two groups are in the third-year undergraduate study. Students participating are assumed to
represent different groups of users of corporate annual reports with different fields of education. As
compared to business students, non-business students are assumed to represent the ordinary users of
annual reports.

23 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)

Process of Data Collection
The Cloze and Comprehension tests were administered to business and non-business students, a test
period being 60-75 minutes. The administration process started with providing an idea of the
importance of the readability of corporate reports. This was followed by distribution of the passage
with the Cloze test. The test instructions were read to the participants as the participants read them
silently. The participants then recorded their responses directly in the spaces.
Finally, the passage was distributed without any deletions. The participants were given the time
to go through the passage and hand it back. This was followed by distributing the reading
comprehension test. The participants then recorded their responses based on what they have understood
during the process of reading the passage.
This process was followed for each report (i.e. the KREC’s report and the NBK’s report) and
with each category of students. During the administration process, efforts have been made to ensure
that each participant worked individually and that external distractions were minimized.

Research Hypotheses
The Cloze and Comprehension tests can be adapted to the study of any component of the accounting
communication process (i.e. financial report preparers, financial report message, or financial report
user).
The present research used the Cloze test and Comprehension test to measure the readability of
Kuwaiti corporate’ financial report messages, and whether the message readability level differed
among report users depending on their educational background. Given that, the following hypotheses
were tested:
Hypothesis One (H
1
): The Kuwaiti corporate annual report footnotes communicate at a level which is
not difficult to read and understand.
Hypothesis Two (H
2
): The level of readability of Kuwaiti corporate annual report footnotes does not
vary significantly among different groups of financial report users.

Data Analysis and the Research Results
This section deals with analysis of the data and the results of testing the research hypotheses. First, data
and results related to annual report readability is presented (H
1
). This is followed by those data and
results related to the potential difference between business and non-business students in understanding
the Kuwaiti corporate annual report (H
2
).

Readability of Annual Reports (H
1
)
It has been agreed that a cloze score of 44 percent is comparable to a conventional multiple-choice
comprehension test score of 75 percent (the instructional reading level- hereinafter referred to as the
cut-off point). Thus, a cloze score of 44 percent can be considered to be a criterion level for
instructional materials. That is, if at least 44 percent of the deleted items are replaced correctly, the
materials can be considered to be written at a level of understandability at or over the instructional level
(Adelberg and Razek, 1984, p.114).
Following this criterion, student who scored the cut-off point and above presented an indicator
of reader’s understandability and readability of the given passage. Given that, testing H
1
was based on
the following standards:
1. The average scores attained by students in each category (i.e. business and non-business) as
compared to the cut-off point indicated the difference between readability level of passage
tested and the instructional reading level.
24 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)

2. In evaluating the acceptance or rejection of H1, 2-tailed t-Tests were performed to test the
significance of the difference between the observed mean readability scores of messages tested
and the instructional reading level referred to above.

Results of Applying the Cloze Test
Table 1 shows the observed mean scores and minimum and maximum scores attained by students
through Cloze Test which was applied to the KREC's annual report. It also displays coefficient of
variation (CV) and t-Test results at 0.05 level of significance. Table 2 presents the same information
regarding the NBK's annual report.
Figures in table 1 reflect substantial difference in the mean scores attained by business and non-
business students. As compared to the cut-off point of 44 percent, it is apparent that the observed mean
scores attained by each group is below the cut-off point. This is supported by the results of t-Test in
which the observed mean scores and the established readability score (i.e. 44%) differ significantly for
both business and non-business students, as well as for the entire population.
Results of applying the Cloze Test to the NBK’s annual report (Table 2) seem to be similar to
those of the KREC. The observed mean scores are less than the established readability score. They
differ significantly either for business or non-business students as well as for the entire population.
Apparently, some of the students in each group attained scores which are higher than the cut-off point
of 44 percent. However, they represent a very low proportion of the sample. As the Sign-Test results
indicate (Table 5), 49 students out of 400 attained a score which was higher than the 44 percent.

Table 1: Results of Cloze Test (Kuwait Real Estate Corporation)

T-Test
Groups Mean Scores CV Min - Max
T-Value P
Business students
38.740 .287 12-58 - 5.63 .0000

Non-business students
25.820 .324 8-50 - 22.90 .0000
Entire population
32.280 .364 8-58 - 15.30 .0000

Table 2: Results of Cloze Test (National Bank of Kuwait)

T-Test
Groups Mean Scores CV Min - Max
T-Value P
Business students
35.930 .200 18-52 - 12.62 .0000
Non-business students
28.840 .282 8-42 - 19.89 .0000
Entire population
32.380 .260 8-52 - 21.15 .0000

Results of Applying Reading Comprehension Test
Tables 3 and 4 show the observed mean scores and minimum and maximum scores attained by students
through a reading comprehension tests to the annual report of both KREC and NBK. Tables also
display coefficient of variation (CV) and t-Test results at .05 level of significance.
The results presented in Tables 3 and 4 seem to be similar to those reported in Tables 1 and 2.
They indicate that the observed mean scores attained by students and the cut-off point of 75 percent
differ significantly at a 0.05 level of significance. This was the case in all situations, i.e., whatever the
group tested (business or non-business) or the report tested (KREC’s report or NBK’s report).
25 European Journal of Economics, Finance And Administrative Sciences - Issue 8 (2007)


Table 3: Results of Reading Comprehension Test (Kuwait Real Estate Corporation)

T-Test
Groups Mean Scores CV Min - Max
T-Value P
Business students
67.700 .129 44-88 - 8.39 .0000
Non-business students
60.800 .171 32-80 - 13.67 .0000
Entire population
64.250 .158 32-88 - 14.95 .0000

Table 4: Results of Reading Comprehension Test (National Bank of Kuwait)

T-Test
Groups Mean Scores CV Min - Max
T-Value P
Business students
64.200 .124 48-88 - 13.10 .0000
Non-business students
59.720 .162 36-80 - 15.83 .0000
Entire population
61.720 .150 36-88 - 19.98 .0000

Based on the results of applying both the Cloze and Comprehension tests, it seems reasonable to
reject H
1
and conclude that footnotes to the financial statements of Kuwaiti corporations are written and
communicated at a level and in a style which is difficult to read and understand.
Two other aspects can be observed from the results reported in Tables 1, 2, 3 and 4:
a. The readability level of the annual report of both KREC and NBK is relatively similar. That is,
there is no substantial difference between the observed mean scores for the two reports. This

can be easily noticed when results reported in table 1 are compared to those reported in Table 2,
and the results reported in Table 3 are compared to those reported in Table 4.
b. The mean scores reported as a result of applying the reading comprehension test (Tables 3 and
4) are higher than those resulting from the application of the Cloze test (Tables 1 and 2). This
may be due to the methodology of applying the two tests. That is, application of a
comprehension test requires students to read the text before answering the test questions. With
the Cloze test, however, students are required to predict the correct deleted words without
having been given the chance to become familiar with the report.

Table 5: Results of sign-test

Groups No. of cases LT* Criterion Level GT** Criterion Level
Cloze Test

Business Students

Kuwait Real Estate Corporation 100 65 35
National Bank of Kuwait 100 87 13
Non-Business Students

Kuwait Real Estate Corporation 100 99 1
National Bank of Kuwait 100 100
Total
400 351 49 (12%)
Comprehension Test
Business Students

Kuwait Real Estate Corporation 100 74 26
National Bank of Kuwait 100 85 15
Non- Business Students

Kuwait Real Estate Corporation 100 90 10
National Bank of Kuwait 100 92 8
Total
400 344 56(14%)
* LT = number of cases with score less than the readability criterion level.
**GT = number of cases with score greater than the readability criterion level.