Corporate risk information in annual reports and stock price behavior in the United Arab Emirates.
Uddin, Md Hamid ; Hassan, Mostafa Kamal
INTRODUCTION
Corporations often utilize various devices, such as media and
newspapers, to communicate information about their activities, but
annual report is the official public document that includes information
about corporations' current activities and future plans. Although
different studies investigate level of voluntary disclosure in the
annual reports (e.g. Robb et. al., 2001; Cabedo and Tirado, 2004; and
Hassan, 2009), managers are generally skeptical about providing all
information to their shareholders without certain regulatory and
professional requirements.
Regulatory requirements are laid in two forms: accounting standards
and listing conditions. For example, Financial Accounting Standard Board
(FASB) and Accounting Standards Board (ASB) set rules of disclosure in
the US and UK respectively, while International Financial Reporting
System (IFRS) sets guidelines that aim at enhancing financial disclosure
globally. In this regard, Statement of Financial Accounting Standards
(SFAS) 119 & 133, Financial Reporting Standard (FRS) 13, and IFRS 7
require the inclusion of risk related information in annual reports.
Likewise, security exchange enforces binding rules for the listed
corporations to promptly release risk related information and market
sensitive information. One the other hand, professional institutions,
such as American Institute of Chartered Public Accountants (AICPA) and
Association of Investment Management and Research (AIMR), encourage
corporations to disclose information about their strategic plans,
business opportunities, risks, as well as process and operations
management. Therefore, we expect that if annual report provides adequate
risk related information to the shareholders then it should influence
the cross sectional behavior of stock prices in market trading.
Adequacy of information in annual report could influence the
behavior of stock prices because it may help reducing uncertainty of
valuation. This enables the shareholders to take informed decisions
about their portfolio. Most importantly, the corporate risk information
disclosure in annual reports would help reducing the agency problem
between the public shareholders and corporate managers. One the one
hand, the accounting literature includes plenty of studies related to
voluntary and/or mandatory disclosures and factors affecting such
disclosures (e.g., Hooks et. al., 2002; Leventis and Weetman (2004); and
Ferrell, 2007). On the other hand, the information effect on the stock
price behavior is well known in the market efficiency literature in
finance (e.g., Fama, 1970). However, there is a research gap in
investigating whether the inclusion of risk related information in
annual reports can influence the stock price behavior. The finding on
relationship between corporate risk disclosure and stock price behavior
would therefore enrich the current knowledge on the linkage between
accounting information and financial market.
This study extends on the prior work that develops a corporate risk
disclosure (CRD) index for the United Arab Emirate (UAE) listed
corporations (e.g. Hassan, 2009) to examine the CRD cross-sectional
relationship with the average volatility of stock price over the
different interval of periods after publication of annual reports. The
volatility of price movements reflects the behavioral pattern of stock
price and the potential risk of investment. This is measured by
calculating spread between the highest and lowest prices in each trading
week over different intervals until end of the financial year, when the
next annual reports become due.
There are several reasons to choose the United Arab Emirates (UAE)
for this study. First the UAE is an emerging market in oil-rich gulf
region, yet most listed corporations apply the IFRS. The UAE
corporations' annual reports demonstrate a pattern of risk
disclosure (Hassan, 2009). The limitation however is that the UAE stock
markets started in 2000; hence it has only eight years of trading
history. Nonetheless, the evidence should have academic value in
understanding the relationship between the disclosures of risk
information in annual report and behavior of stock price.
The empirical results, based on 36 companies whose financial year
ends in December 2005 and annual reports are made available by the next
month, show that CRD index has nonlinear relationship with the stock
price volatility after the publication of annual reports. The finding is
inconsistent with the hypothesis that more risk disclosure reduces
uncertainty and hence the volatility of stock price declines. It is also
found that CRD has also nonlinear relationship with the market risk
factor (beta coefficient) of the stocks.
The paper concludes that the findings may have three major
implications. First, more risk information reported in annual reports
intensifies the uncertainty about the future corporate
performance. Second, corporate risk disclosure helps investors to
diversify their investment portfolio that reduces the level of portfolio
risk, but excess and unnecessary disclosure of information surpasses the
benefits of relevant risk disclosures raising the level of market risk.
Third, the risk information revealed in annual reports may not
adequately address the investors' concern, hence a survey may be
conducted to ascertain the types of risk information that investors are
concerned about.
The rest of the paper is organized in five sections. Literature
review and hypotheses are presented in Section 2. Research design and
methodology are discussed in Section 3. Sample and data are described in
Section 4. Results are discussed in Section 5 before conclusions are
drawn in Section 6.
LITERATURE REVIEW AND HYPOTHESES
Information asymmetry among the different corporate stakeholders
(e.g., shareholders and management) is one of the sources of agency
problems (Jensen and Meckling, 1976). This problem can be mitigated
effectively by increasing the level of disclosure (Mahoney, 1995). This
is because more information enables shareholders to take informed
investment decisions, and thereby risks and agency costs are reduced
(Hutton, 2007). If corporation need to approach investors to fund their
projects they should voluntarily disclose all relevant information
[Armitage and Marston (2008) and Mazola et. al., 2006)]. This is because
high level of disclosure is more likely to attract investors, who become
confident that stocks are fairly priced in the market [Diamond and
Verrecchia (1991), Kim and Verrecchia (1991a, 1991b, 1994 and 2001)].
Despite the benefits of high level of disclosure, the level of
voluntary disclosure is found to be low and fall short of
investors' expectations (Hooks et. al., 2002). The problem is more
sever in the emerging markets where corporations are less likely to
disclose information related to their strategic plans as well as their
major weaknesses and risks [Leventis and Weetman (2004) for Greece and
Kuasirikun and Sherer (2004) for Thailand]. Compared to developed
capital markets, such as USA and UK, emerging markets lack binding rules
that enforce listed corporations to promptly release information
associated with their strategic plans, opportunities, risks, and key
business processes [Salter (1998) and Patel et. al., (2002)]. Disclosure
of such information not only enhances business competition but also
reduces agency problem (Ferrell, 2007).
Although corporations may disclose information by announcements,
annual reports remain the main document that contains both financial and
non-financial information (1). Patel and Dallas (2002) found a
significant relationship between the amount of information included in
annual reports and market risk as well as investors' valuation of
shares. Therefore, it is likely that annual reports' financial
and/or non-financial information are relevant for shareholders in
assessing the potential risk of investment. The effect of disclosing
accounting numbers on the stocks returns is well documented in the
literature [Ball and Brown (1968), Beaver et. al. (1979), Fama and
French (1992 & 1993) and Kraft et. al. (2007)], but the effect of
disclosing a certain category of information, such
as risk information, is yet to be adequately explored. In this
regard, Li (2007) found that the textual disclosure of information
(strategic plan, market competitions, key resources, growth prospects
etc), has relationship with future earnings and stock returns, because
investors can better ascertain the risk of future investment.
Although the above studies shed light on the importance of risk
information for the investors, this paper poses the question of
"whether all the disclosed risk information is relevant for risk
perspective". The corporate risk disclosure literature shows
different approaches to analyze and measure the level of risk
disclosure. Some scholars use the content analysis (e.g. Beretta and
Bozzolan, 2004; Lajili and Zeghal, 2005; Linsley and Shrives, 2006).
Others use disclosure index (e.g. Robb et. al., 2001; Cabedo and Tirado,
2004 and Naser et. al., 2006). Third rely on the readability of risk
sentences in annual reports (Linsley and Lawrence, 2007). Recently,
based on literature, Hassan (2009) identifies a total of 45 different
risk information items in annual reports.
The paper attempts to investigate whether more informative annual
reports, that include more risk related information, can help reducing
investors' uncertainty and consequently influence the market price
of shares. Accordingly, we investigate that more risk disclosure would
enable investors to take informed decisions on their investment (Marino
and Matsusaka, 2005) while reducing information asymmetry. Our argument
is that the higher the level of risk disclosure the more likely is the
reduction in the uncertainty associated with return from the investment,
and hence share prices will be less volatile in the periods following
the disclosure of annual reports. Therefore, our first null and
alternative hypotheses are constructed as follows:
[H.sub.0]: The volatility of stocks with higher levels of corporate
risk information disclosure will not be significantly lower than that of
the stocks with lower levels of risk disclosure.
[H.sub.A]: The volatility of stocks with higher levels of corporate
risk information disclosure will be significantly lower than that of the
stocks with lower levels of risk disclosure.
If the null hypothesis is not rejected then findings will shed new
light on the relevance of the different risk information supplied to the
shareholders. Accepting the null hypothesis not only raises the concerns
of corporations and/or regulators but also increases their willingness
to align the risk information in annual report with the information
needed by the shareholders. However, if the empirical findings reject
the null hypothesis then it can be concluded that more disclosure will
reduce investors' uncertainty of returns and corporations should
voluntarily publish risk information to their shareholders.
Capital asset pricing theory suggests that market risk factor (beta
coefficient) determines the investors' expected return (i.e. cost
of capital). Accordingly, if the disclosure of more risk information
reduces investors' uncertainty then it should negatively affect the
stock's market risk factor (beta coefficient). This is because the
lower the uncertainty the lower should the risk of investment.
Theoretical literature shows that quality and quantity of financial
disclosure affect the cost of equity and, in turn, investors require
higher return for lower level of disclosure [Easley and O'Hara,
(2004), Hughes et. al. (2007), Lambert et. al. (2007)]. This suggests
that more disclosure of risk information should benefit investors and
stocks' market risk should decline. While direct evidence on this
issue is yet to be found, the present indirect evidence however is not
consistent with the conjecture. For example, although the fair
disclosure (FD) regulation in the US requires public corporations to
increase fair value disclosure to investors, that disclosure did not
reduce the cost of capital [Duarte et. al., (2007) and Gomes et. al.,
(2007)]. Likewise, Wang et. al., (2008) found that the increase in
voluntary disclosure in the Chinese capital market is associated with
the higher return on equity (Wang et. al., 2008). The implication of
these indirect evidences is that investors' market risk perhaps did
not decline after disclosure of more information. Given this background
our second null and alternative hypotheses are formulated as follows:
[H.sub.0]: The level of corporate risk information disclosure does
not have negative effect on the level of market risk of the stock.
[H.sub.A]: The level of corporate risk information disclosure has
negative effect on the level of market risk of the stock.
If the evidence cannot reject the null hypothesis then it may raise
further research question of "whether providing more risk
information in annual reports can help corporations to reduce their cost
of capital through reducing investors' market risk" (2). A
consequent query would be to investigate "why investors cannot
recognize all the risk information provided in annual reports".
However, if the null hypothesis is rejected then it will imply that
investors are benefited through reduction of information asymmetry.
Therefore, risk of investment decreases that in turn helps to reduce the
cost of equity capital.
RESEARCH DESIGN AND METHODOLOGY
Since the study empirically examines the impact of corporate risk
information on variability of stock returns and market risk factor, it
requires to measure three variables: (i) Corporate Risk Disclosure, (ii)
Volatility of Securities Return, and (iii) Market Risk Factor. The
following subsections discuss how each variable was measured in the
context of this study.
Corporate Risk Disclosure
Literature suggests different approaches to measure information
disclosure in annual reports. (3) For example, content analysis used by
Beretta and Bozzolan (2004), Lajili and Zeghal (2005) and
Linsley and Shrives (2006) among others; corporate information
disclosure level is used by Robb et. al. (2001), Cabedo and Tirado
(2004), Naser et. al. (2006) and Hassan (2009); and readability of risk
sentences in annual reports is assessed by Linsley and Lawrence (2007).
Each approach has weakness as well as strengths and some of them are
more subjective than others. The study needs a measure of risk
disclosure that comprehensively covers different business aspects such
as business operations, management, and financial matters. That
measurement, in turn, allows to properly assessing the risk of
investment. The study relies on Hassan's (2009) corporate risk
disclosure index (hereafter CRD) to rank the UAE corporations according
to level of risk disclosure. The index is based on an extensive review
of literature while, at the same time, underscoring the UAE statutory
requirements. Appendix I outlines CRD index items utilized in this
study. The CRD index items are checked against the study sample of 36
annual reports.
The study awards "1" for each risk information item found
in the annual report and awards "0" if the information is not
found. A firm receiving higher CRD rank means it discloses more risk
information and vise versa. It is noted that the CRD rank may be
computed by assigning specific weight to each information item, but we
consider it is less important in this study, as we are concern about the
extent of risk information disclosure. It is true that different
information item may have different importance to different users hence
weighting for information may be useful, but weighting process is
subjective and may be biased toward a particular group of users.
Therefore, an unweighted CRD rank is utilized since it serves the
purpose of this study.
Volatility of Stock
Stock volatility is primarily estimated by the standard deviation
or variance of returns, which is considered as a measure of risk in the
portfolio theory. There are sophisticated risk models, developed under
Autoregressive Conditional Heteroskedasticity (ARCH), Generalized
Autoregressive Conditional Heteroskedasticity (GARCH), and option
pricing frameworks, which are valid under specific assumptions. Since we
are uncertain whether any of the models provide right descriptions of
volatility that is easily understood by general investors, we do not
like to keep the findings subject to particular assumptions of any
model. Therefore, a model-free simple volatility measure is constructed.
This makes study findings easier to understand by general investors
while retaining the academic value.
We expect that there will be less uncertainty for the investors
following the disclosure of adequate risk information. Thus stock's
market price will become more stable and the spread between the highest
and lowest prices will become relatively narrower. We examine the
volatility of stocks by measuring the average spread of the highest and
the lowest prices over different intervals and we estimate it as follows
[AVO.sub.iT] = 1/N [N.summation over (t=1)] ([P.sup.H.sub.it] /
[P.sup.L.sub.it] -1) (1)
Where, [AVOL.sub.iT] is the average volatility of stock i for the
interval of period T, after publication of annual report.
[P.sup.H.sub.it] and [P.sup.L.sub.it] are
respectively the highest and lowest prices of the stock i in the
week t of the interval of period T, and N is the numbers of weeks
in each interval of period T.
The Average Volatility ([AVOL.sub.iT]) is computed over different
intervals (T), such as the 1st month, 1st quarter, 1st half-year and
full-year following the end of the financial year. We examined the
weekly volatility instead of intraday or monthly volatilities. This is
because price corrections generally not settled within the trading day;
rather it may roll on over the next few days. One-month period is also
too long to measure the behavior of general price movements. Therefore,
weekly volatility may be suitable in the study. The average weekly
volatility over different intervals, T, should depict general pattern of
price fluctuations. We check the corporate announcement files maintained
by stock exchanges to find out any major announcement (e.g., earning,
dividend, and major investments etc.,) made after publication of the
annual report. If any such announcement is found, the stock prices of
the announcement week and the week after announcement are excluded from
[AVOL.sub.iT] calculation. In order to reduce biasness in the
statistical calculations, we also exclude the extremely high and low
prices of the year that exists beyond [+ or -] 3 s levels.
The [AVOL.sub.iT] measures the total volatility of a stock, which
may be driven by the general market sentiments to certain extent. Hence,
the [AVOL.sub.iT]for individual stock i in the week t has been adjusted
for market volatility in same week t. We call the adjusted volatility of
individual stock as the excess volatility, which is estimated as
follows:
[AEVOL.sub.it] = 1/N [N.summation over (i=1)] ([P.sup.H.sub.it] /
[P.sup.L.sub.it] - [I.sup.H.sub.it] / [I.sup.L.sub.it] (2)
Where, [AEVOL.sub.iT] is the average excess volatility of stock i
for the interval of period T, after publication of annual report.
[P.sup.h.sub.it] and [P.sup.L.sub.it] are respectively
the highest and lowest prices of the stock i in the week t of the
interval of period T. [I.sup.H.sub.it] and [I.sup.L.sub.it] are
respectively the highest and lowest values of
market index in the week t of the interval of period T, and N is
the numbers of weeks in each interval of period T.
The Average Excess Volatility ([AEVOL.sub.iT]) is computed over the
different intervals (T), such as the 1st month, 1st quarter, 1st
half-year and full-year following the last financial year. Since the
volatility of market has been adjusted, the [AEVOL.sub.iT] may depict
the stock price volatility due to the reasons specific to the company.
Market Risk Factor
Investors' market risk factor for the sample stocks is
calculated over one-year period following the end of financial year. Our
objective is to examine the cross-sectional variation of market risks in
the period after disclosure of annual reports. The market risk for each
stock is calculated, as follows, by estimating the beta coefficient
([beta]) of time series market model using weekly returns.
[R.sub.it] = [[alpha].sub.i] + [beta] [R.sub.mt] +
[[epsilon].sub.l[tau]] (3)
Where, [R.sub.it] is the return of stock for the week t in the
period after current financial year. [R.sub.mt] is the market return for
the same week t calculated based on the all-share price index,
[[alpha].sub.i] is the constant and [[epsilon].sub.it] is the error
term.
In estimating the market model, selection of return interval and
estimation period is a crucial decision that may affect the biasness of
estimated beta coefficient, [beta], (henceforth we write as BETA). A
long return interval, e.g., one-month, is normally used to reduce the
problem of infrequent trading while a reasonably long period, e.g., five
years, is taken to ensure precision in estimation. The scope and purpose
of this paper limits our BETA estimation over one-year period following
the end of the financial year. If adequate risk information of company i
is disclosed in its annual report for year T, the stock's market
risk factor (BETA) in year T+1 may be affected. We use weekly returns so
that we have adequate observations for time-series regression. (4)
Test Design and Models
After constructing the test variables (CRD, AVOL, AEVOL, and BETA)
for each company, we test the hypotheses by cross-sectional analyses of
the stock volatility and investors' market risk factor across the
different subsamples based on the level of risk disclosure. Samples
companies are classified into four groups based on their CRD index.
Companies with CRD index below the 1st quartile are classified as
'Very Low Disclosure' group, companies with CRD index between
the 1st and 2nd quartiles as "Low Disclosure' group, companies
with CRD index between the 2nd and 3rd quartile as 'High
Disclosure' group, and those with CRD index above the 3rd quartile
as 'Very High Disclosure' group. The mean differences of AVOL,
AEVOL, and BETA across these four subsamples are tested for statistical
significance. The effect of CRD on the level of excess volatility and
investors' market risk is first tested in simple regressions and
later examined by estimating the following multiple regression models.
In these regressions, the actual CRD numbers are used as explanatory
variables.
[[AEVOL.sub.iT] = [[alpha].sub.i] [chi] [CRD.sub.i] + [N.summation
over (j=1)] [delta] [CH.sub.j] + [[epsilon].sub.i] (4)
[BETA.sub.i] = [[alpha].sub.i] + [chi] [CRD.sub.i], + [N.summation
over (j=1)] [delta] [CH.sub.j] + [[epsilon].sub.i]
Where, all the variables are described as above except [CH.sub.j]
(i=1 ... ... n) that are the factors determining characteristics of the
sample firms. We include a number factors based on the literature review
and the UAE market conditions. The following additional variables (as
control factors) are feasible for us to compute:
SIZE is the log of total asset at end of 2005.
DAR is the total debt divided by total asset of the company as at
the end of 2005.
TURN is the average weekly turnover volume calculated for each
company over the 1st month, 1st quarter, 1st half year and full year
period after last financial year.
FOWN is the percent of equity held by the foreign investors.
INDFIN identifies the company whether it is listed on the finance
and banking sector. (5) We include this as a dichotomous variable. INDIN
= 1 if the company is finance and banking company, else INDIN = 0.
ADEX identifies the company whether it is listed on Abu Dhabi Stock
Exchange We include this as a dichotomous variable. ADEX = 1 if listed
on Abu Dhabi exchange, else ADEX= 0.
MARGINis the net profit margin for 2005.
EPS is the earning per share for 2005.
ROE is the return on equity for 2005
We examine the degree of the effect of corporate risk disclosure
(CRD) on the level of stock volatility (AEVOL) and investors'
market risk (BETA) if other relevant factors are considered. Finally we
estimate a parsimonious model using the stepwise regression method, and
to observe whether CRD remains in the parsimonious model as a
significant variable.
SAMPLE DESCRIPTIONS
The primary sample consists of 49 UAE corporations listed in Dubai
Financial market (DFM) and Abu Dhabi Stock Market (ADSM). These
companies published their annual reports for the year 2005. Since Dubai
and Abu Dhabi stock markets started trading activities in 2000,
sufficient numbers of published annual reports were not available for
the period prior to 2005. Moreover, the UAE financial markets become
relatively matured over the five years period since the start of trading
activities. As of 2005, a total of 94 companies listed in both exchanges
(34 in DFM and 59 in ADSM). Of these, 41 stocks are new companies listed
in 2005. Therefore, 49 companies who published their annual report for
the year 2005 were listed prior to 2005.
To prevent undue disturbances in the analysis, caused by financial
year differences, five corporations with year-ended in the months other
than December are removed. Similarly, to maintain homogeneity of the
sample corporations, three non-UAE corporations listed in the DFM are
removed. Finally, the sample becomes 41 corporations spanning over banks
(12 samples), insurance (5 samples), finance/investment (7 samples),
hotels (2 samples), construction (5 samples), cement (2 samples),
telecommunication (2 samples), and others industries (6 samples). While
conducting regression tests, the sample size dropped to 36 because of
missing data for certain explanatory variables. However, the findings
based on a sample of 36 samples should have statistical significance
because it covers about 74 percent of all corporations publishing their
annual reports in 2005. The, sample also covers nearly 70 percent of the
whole market capital at the end of 2005.
RESULTS AND DISCUSSIONS
Average Volatility
Table 1 shows that average volatility (AVOL) varies with the level
of corporate risk disclosure (CRD). The volatility is higher for the
companies disclosing maximum risk related information. We divide the
samples into four subgroups: very low disclosure, low disclosure, high
disclosure and very high disclosure based on number of disclosures. It
reveals that the average volatility is consistently higher for the
'very high disclosure' group compared to those of the other
groups over the different intervals, e.g., one month, one quarter, half
year, and a full year after the end of the financial year. For example,
a comparison between the two extreme subgroups shows that one-month AVOL
appears to be 0.052 for the 'very high disclosure' group
compared to 0.028 for the 'very low disclosure' group; the
difference is statistically significant. Similarly, the one-quarter,
half-year and a full year AVOLs for the 'very high disclosure
group' are respectively 0.088, 0.094, and 0.070 that are generally
higher than the respective AVOLs of same intervals for the 'very
low disclosure group'. The differences however are not significant.
All of the AVOLs for different subsamples presented in column 1
through 4 are statistically significant at one percent level with t
values varying from 3.043 to 14.574. The average volatility (AVOL) of
stocks is calculated on weekly basis over different intervals of period,
e.g., one month, one quarter, half year, and full year after official
publication of annual report where the corporate risk related
information are disclosed. CRD is an index calculated based on the
number of risk related information disclosed in the annual report.
Samples companies are classified into four groups based on their CRD
index. Companies with CRD index below the below the 1st quartile are
classified as 'Very Low Disclosure' group, companies with CRD
index between the 1st and 2nd quartiles as "Low Disclosure'
group, companies with CRD index between the 2nd and 3rd quartile as
'High Disclosure' group, and those with CRD index above the
3rd quartile as 'Very High Disclosure' group. Asterisks ** and
* measure the level of significance at five percent and ten percent
levels respectively.
However, the AVOLs of two intermediary groups: 'high
disclosure' and 'low disclosure' depict a different
pattern. The AVOLs of 'high disclosure' group is generally
lower than those of the 'low disclosure' group over the
different intervals. The differences though are not significant. The
overall finding tends to indicate a non-linear relationship between the
volatility of stock prices and the level of corporate risk related
information disclosure.
Average Excess Volatility
There is a probability that the volatility of stock is driven, to a
certain extent, by the general market sentiments. Therefore, we subtract
the market-wide volatility from the individual stock volatility in order
to measure the excess volatility that may be more related to the company
risk characteristics/reasons. Table 2 shows that average excess
volatility (AEVOL) for the subsamples across different intervals are
lower than the AVOLs (in Table 1), and many of them are insignificant
because the effect of market sentiment is adjusted. The full-year AEVOL
for 'low disclosure' and 'very high disclosure'
groups are however remain statistically significant. This odd result
could be due to information leakage on current operating results, which
normally appears in the last quarter. Uncertainty arises if the
corporate risk information disclosed in last annual report does not
adequately support the current operating results. It is likely that such
uncertainty was higher for the companies belong to 'low
disclosure' and 'very high disclosure' groups.
Nonetheless, the results may imply that a part of the whole-year stock
volatility is due to uncertainty arises from the company risk
characteristics.
None of the AEVOLs for different subsamples presented in column 1
through 4 are statistically significant, except the full-year AEVOLs for
the 'low-disclosure' and 'very high disclosure'
subsamples that are significant with t values of 2.517 and 1.844
respectively. The average excess volatility (AEVOL) of stocks is
calculated on weekly basis over different intervals of period, e.g., one
month, one quarter, half year, and full year after official publication
of annual report where the corporate risk related information are
disclosed. CRD is an index calculated based on the number of risk
related information disclosed in the annual report. Samples companies
are classified into four groups based on their CRD index. Companies with
CRD index below the below the 1st quartile are classified as 'Very
Low Disclosure' group, companies with CRD index between the 1st and
2nd quartiles as "Low Disclosure' group, companies with CRD
index between the 2nd and 3rd quartile as 'High Disclosure'
group, and those with CRD index above the 3rd quartile as 'Very
High Disclosure' group. Asterisks ** and * measure the level of
significance at five percent and ten percent levels respectively
The important finding is that AEVOLs are consistently higher for
the 'very high disclosure' group of subsamples compared to
those of the other subsamples. For example, a comparison between the two
extreme subgroups shows that one-month AEVOL is 0.013 for the 'very
high disclosure' group compared to -0.011 for the 'very low
disclosure' group; the difference is statistically significant.
Similarly, the one-quarter, half-year and a full year AEVOLs for the
'very high disclosure group' are respectively 0.030, 0.035,
and 0.028 that are generally higher than the respective AEVOLs of same
intervals for the 'very low disclosure group'. The differences
however are not significant.
These findings suggest that companies, that disclose maximum risk
information, have higher excess stock volatility than those of companies
that disclose minimum risk information. However, like AVOL results, the
AEVOLs of the two intermediary groups reveal a different pattern. The
AEVOLs of 'high disclosure' groups are lower than those of the
'low disclosure' group across different intervals of period
after the financial year. Therefore, the sub-samples' findings
indicate that the level of corporate risk disclosure may have non-linear
effect on the level excess volatility of stocks.
Investors Market Risk
It was argued that if investors' uncertainty reduces with the
disclosure of more risk information in the annual report then it should
negatively affect the stock's market risk factor (beta coefficient)
that determines investors' expected return (cost of capital) under
the capital asset pricing model. The results, presented in Table 3, show
that the average beta coefficient for the 'very low
disclosure' group is higher than those of the other groups. For
example, the average beta of the 'very low disclosure' group
of sample companies is 1.3350 while that for the 'very high
disclosure' group is 0.6919, though the difference is not
statistically significant. The average betas for the two intermediary
groups of samples depict different characteristics: beta for the
'low disclosure' group is 0.5391 while that of the 'high
disclosure' group is 0.8098. The average betas of the four
different groups of sample companies based on the level of risk
information disclosure tend to indicate a non linear relationship
between the invertors' market risk and level of risk disclosure.
The systematic risk (beta co-efficient) is calculated using weekly
returns over one year period after publication of annual report. Samples
companies are classified into four groups based on their CRD index.
Companies with CRD index below the below the 1st quartile are classified
as 'Very Low Disclosure' group, companies with CRD index
between the 1st and 2nd quartiles as "Low Disclosure' group,
companies with CRD index between the 2nd and 3rd quartile as 'High
Disclosure' group, and those with CRD index above the 3rd quartile
as 'Very High Disclosure' group
Regression Findings
Table 4 shows the effect of corporate risk disclosure (CRD) on the
level of stock volatility (AEVOL) and investors' market risk (BETA)
in a simple regression model. The subsample analyses above indicate
nonlinear relationship, which has been tested with linear and nonlinear
regressions that used actual CRD index as the explanatory variable. The
linear and nonlinear findings are presented in Panel A and Panel B of
Table 4 respectively. The linear regression findings in Panel A show
that corporate risk disclosure positively affects the stock volatility
but do not affect the market risk. The CRD variable depicts positive
effect on the level of AEVOL. However, the effect is not very strong
because the CRD regression coefficient for the first-month AEVOL is
insignificant while those for the other AEVOLs are significant only at
10 percent level. Therefore, the simple linear regression results are
inconsistent with the null hypotheses with respect to the stock
volatility and investors' market risk.
However, the nonlinear results--presented in Panel B of table
4--show that risk disclosure has a negative relationship with all the
dependent variables (as expected). The inclusion of [CRD.sup.2]
(non-linear measure of CRD), as an explanatory variable, has led to
having linear CRD coefficients across different intervals to be
insignificant for all the dependent variables. The coefficients of
[CRD.sup.2] are significant for the first-quarter, half-year, and
full-year AEVOLs. The respective adjusted R2 for these nonlinear models
increased significantly compared to those of the linear models. This
suggests that more information disclosure in the annual report may
indeed confuse the investors and uncertainty increases at higher rate
than the degree of disclosure. This could be because of possible
difference between the investors' expectations and the information
contents.
The results of multiple regressions, using 11 explanatory
variables, presented in Table 5 reveal that the nonlinear effect of
corporate risk disclosure (CRD) on the level of stock volatility (AEVOL)
and investors' market risk (BETA) has been sustained in multiple
regressions. It is found that [CRD.sup.2] coefficients of different
multiple regressions appear to be more significant compared to the
univriate [CRD.sup.2] coefficients presented in Table 4. In addition,
the linear CRD coefficients of multiple regressions remain
insignificant, except the one for the model that uses BETA as the
dependant variable. These findings cannot reject our first null
hypothesis that more corporate risk disclosure has no negative effect on
the stock volatility. However, the second null hypothesis is partially
rejected because CRD coefficient in the model with BETA as dependent
variable depicts a negative sign and significant at 5 percent level
while [CRD.sup.2] depicts a positive sign that is also significant.
Therefore, results tend to suggest that more disclosure of
corporate risk information may increase investment uncertainty hence
stock volatility increases, though additional relevant information may
help investors to diversity their portfolio and to minimize the market
risk as beta tends to decline with the additional risk information
disclosure. However, excess and unnecessary information is not useful
for reduction of stock volatility and market risk. This finding is
consistent with the recent evidence that voluntary information in the
annual report may not contain value relevant information about future
earnings or investors are not capable of incorporating information in
the firm value estimates (Banghoj and Plenborg, 2008).
This table shows the univariate power of CRD in linear and
quadratic models to determine the stock volatility and investors'
market risk. CRD is an index of risk information calculated based on the
number of risk related information disclosed in the annual report. AEVOL
is the average excess volatility of stocks over the different intervals
after last financial year for which the annual report has been
published. BETA measures the investors' market risk and calculated
using weekly returns over one year period after last financial year. The
value in parenthesis is the t-statistic of regression co-efficient.
Asterisks ** and * measure the level of significance at five percent and
ten percent levels respectively.
This table shows the effect of CRD variable in determining the
excess stock volatility and investors' market risk in different
setting of multiple regressions. We considered a number of other
explanatory variables based on literature and UAE market conditions.
These include total asset size (SIZE), debt asset ratio (DAR), stock
turnover in market trading (TURN), percentage of foreign ownership
(FOWN), listing as finance and banking company (INDFIN), listing on Abu
Dhabi Exchange (ADEX), profit margin (MARGIN), earning per share (EPS),
and return on equity (ROE). AEVOL is the average excess volatility of
stocks over the different intervals after last financial year for which
the annual report has been published. BETA measures the investors'
market risk and calculated using weekly returns over one year period
after last financial year. We have calculated the cross-correlation
among the explanatory variables to examine the severity of
multicolinearity problem. It is found that all correlation coefficients
are statistically insignificant, except the coefficient of correlation
(0.2777) between EPS and ROE that is significant at 10 percent level. We
apply stepwise regression method to select only the significant
variables to estimate the parsimonious models that are presented in the
table. Asterisks ***, **, and * measure the level of significance at one
percent, five percent and ten percent levels respectively.
We applied stepwise regression method for estimating parsimonious
model to identify the most relevant variables explaining AEVOL and BETA.
Results show that [CRD.sup.2], TURN, and ADEX are the most relevant
explanatory variables in the models using AEVOL as the dependent
variable. These three variables together can explain about 51.5 percent
of the stock volatility in the UAE market. Therefore, the results
suggest that excess and unnecessary disclosure of risk information
though intensify the stock volatility due to higher investment
uncertainty, the stock volatility also depends on the volume of stock
turnover.
In additional, the level of volatility is higher in Abu Dhabi
market compare to that in Dubai market. It is found that CRD
[CRD.sup.2], MARGIN and ADEX are the most relevant explanatory variables
in the model using BETA as the dependent variable. These variables
together can explain about 14 percent of investors' market risk.
These findings depict that more information disclosure allows the
investors to diversify their portfolio and to reduce market risk, but
excess and unnecessary disclosure of information surpasses the benefits
of relevant risk disclosures raising the level of market risk. The
profit margin (MARGIN) has significant positive effect on the level of
market risk, which is not consistent with the general idea that profit
making companies are to be less risky. Results also show that average
market risk of Abu Dhabi stocks is lower than that of the Dubai stocks.
Robustness Check
The AEVOL results, presented above, are checked for robustness by
estimating the stock volatility using Generalized Auto Regressive
Conditional Heterscadasticty (GARCH) model. We estimate GARCH (1,1)
model for each stock using 258 daily time-series returns following
completion of a financial year for which annual report is published. The
GARCH (1,1) model is as follows
[[sigma].sup.2.sub.t] = [omega] + [alpha] [[epsilon].sup.2.sub.t=1]
+ [beta] [[sigma].sup.2.sub.2t-1] (6)
Where, [[epsilon].sup.2.sub.[[tau]-1] is one-period lag squared
error generated by Auto
Regressive Moving Average (ARMA) model for stock return [r =
LN([Price.sub.t] / [Price.sub.t-1])]. ARMA(1) is estimated as [r.sub.t]
= [b.sub.0] + [b.sub.1] [r.sub.t-1] + [[epsilon].sub.t].
[[sigma].sup.2.sub.[tau]-1] is the variance of the last period.
After estimating the parameters of model 6, the GARCH (1,1)
volatilities are calculated for each trading day. The Average GARCH
(1,1) Volatility ([AVOL_GARC[H.sub.iT]) for stock i is computed over
different intervals (T), such as the 1st month, 1st quarter, 1st
half-year and full-year following the end of the financial year.
Finally, the empirical test Model 4 has been redefined as follows:
[AVOL_G4RCH.sub.iT] = [[alpha].sub.i] + [chi] [CRD.sub.i] +
[N.summation over (j=1)] [delta] [CH.sub.j] + [[epsilon].sub.i] (7)
Where, all variables except AVOL_GARCH are as defined earlier. The
effect of corporate risk disclosure (CRD) on the level of GARCH
volatility is examined by estimating the above model with 11 explanatory
variables using stepwise regression method. The findings are presented
in Table 6. A review of results in Table 5 and Table 6 shows that CRD
has similar effect on both excess volatility (AEVOL) and GARCH
volatility (AVOL_GARCH).
Table 6 shows that all [CRD.sup.2] coefficients are positive and
statistically significant. This suggests that CRD has non-linear
positive effect on the level of GARCH volatility over different
intervals of periods from one-month to one-year following the end of a
financial year. However, it is noted that over the shorter periods,
i.e., one-month and one-quarter, the corporate disclosure has
significant negative linear effect on the GARCH volatility. As a whole,
GARCH volatility results cannot unequivocally reject the null hypothesis
that 'more corporate risk disclosure has no negative effect on the
stock volatility'. Therefore, GARCH volatility results reinforce
our earlier suggestion (based on AEVOL) that excess and unnecessary
disclosure of risk information intensifies the stock volatility due to
higher investment uncertainty.
CONCLUSIONS
Accounting regulators and international accounting standards
enforce different binding and non-binding rules for the corporate firms
to publish adequate risk information in their annual reports. These
reports include audited financial statements and other information about
current activities and future plans which may be important for the
shareholders. Adequate risk information in annual reports may reduce
uncertainty. Therefore, investors can take informed investment decisions
that reduce agency problems between the shareholders and managers. Based
on this, we hypothesized that more disclosure of risk information may
have negative relationship with the level of stock price volatility and
investors' market risk. An empirical test based on 36 UAE companies
listed in Dubai and Abu Dhabi stock exchanges showed that disclosures of
corporate risk information have no linear negative effect on the level
of stock volatility and investors' market risk. Instead, the
results depict a non-linear quadratic effect of risk disclosure on the
level of stock volatility and market risk. The findings tend to suggest
that more disclosure of corporate risk information may indeed increase
uncertainty of investment in UAE market, but more information allows the
investors to diversity their portfolio and minimize the market risk.
Finally, readers should take note of an inevitable limitation that
sample size is relatively small though it covers about 74 percent of the
population.
APPENDIX I
Items of risk information reported in the annual report that are
utilized in constructing the corporate risk disclosure index
(CRD Index)
Sources (Hassan, 2009)
Alfredson Beretta & Lajili &
et al., Bozzolan, Zeghal,
2006 2004 2005
General Risks Information
1 Competition in product X X
market
2 Brand name erosion / X X
change / addition
3 New alliances and X X
joint ventures
4 Relationship to X X
Government developments
plans
5 Customer acquisition X X
processes
6 Recruiting of qualified X X
and skilled professional
7 Regulations/Sharia's X X
law/Overseas tax law
8 Events beyond balance X X
sheet
9 Political environment X X
10 Natural disasters X X
Accounting Policies
11 Use of estimates / X
judgments
12 Collateral assets X
against loans
13 Objectives of X
provisions / legal
constructive
14 Financial assets X
impairment
15 Other assets X
impairment
16 De-recognition of X
financial assets
17 Risk management X X
18 Detailed risk X X
management
19 Objective of holding X
derivatives / instruments
20 Contingent liabilities X
21 Contingent assets X
22 Inventory Lower of X
Cost or Market
23 Key sources of X X
estimation uncertainty
Financial Instruments
24 Classifying X
instruments by risks
25 Principal, stated X
value, face value
26 Reclassification X
of instruments
27 Cumulative/ change X
in Fair value
Derivatives hedging
28 Hedging description X
29 Change in Fair Value X
of assets or liability
30 Cash flow hedge X
Reserves
31 Statutory
32 Legal
33 Contingency /
special/ general
Segment information
34 Business major X X
segments
35 Geographical X X
concentration
36 Customer / X X
(asset/liabilities)
concentration
Financial and Other Risks
37 Operational risk / X
Insurance risk
38 Market risk X X
39 Interest rate = X X
pricing risk sharia'a
40 Exchange rate X X
41 Liquidity X X X
42 Credit X X X
43 Pricing risk sharia'a
44 Tabular presentation
45 Sensitivity analysis
Linsley & Abraham and Lopes &
Shrives, Cox, 2007 Rodrigues,
2006 2007
General Risks Information
1 Competition in product X X
market
2 Brand name erosion / X X
change / addition
3 New alliances and X
joint ventures
4 Relationship to X
Government developments
plans
5 Customer acquisition X X
processes
6 Recruiting of qualified X
and skilled professional
7 Regulations/Sharia's X
law/Overseas tax law
8 Events beyond balance
sheet
9 Political environment X X
10 Natural disasters X X
Accounting Policies
11 Use of estimates /
judgments
12 Collateral assets X
against loans
13 Objectives of
provisions / legal
constructive
14 Financial assets X
impairment
15 Other assets X
impairment
16 De-recognition of X
financial assets
17 Risk management X
18 Detailed risk
management
19 Objective of holding X
derivatives / instruments
20 Contingent liabilities X
21 Contingent assets X
22 Inventory Lower of
Cost or Market
23 Key sources of
estimation uncertainty
Financial Instruments
24 Classifying X
instruments by risks
25 Principal, stated X
value, face value
26 Reclassification X
of instruments
27 Cumulative/ change X
in Fair value
Derivatives hedging
28 Hedging description X
29 Change in Fair Value X
of assets or liability
30 Cash flow hedge X
Reserves
31 Statutory
32 Legal
33 Contingency /
special/ general
Segment information
34 Business major
segments
35 Geographical
concentration
36 Customer /
(asset/liabilities)
concentration
Financial and Other Risks
37 Operational risk / X
Insurance risk
38 Market risk X
39 Interest rate = X X
pricing risk sharia'a
40 Exchange rate X X
41 Liquidity X X
42 Credit X X X
43 Pricing risk sharia'a
44 Tabular presentation
45 Sensitivity analysis
Robb et Cabedo & Barako et
al., 2001 Tirado, al., 2006
2004
General Risks Information
1 Competition in product X X X
market
2 Brand name erosion / X X X
change / addition
3 New alliances and X X X
joint ventures
4 Relationship to X X X
Government developments
plans
5 Customer acquisition X X X
processes
6 Recruiting of qualified X X X
and skilled professional
7 Regulations/Sharia's X X X
law/Overseas tax law
8 Events beyond balance X X X
sheet
9 Political environment X X X
10 Natural disasters X X X
Accounting Policies
11 Use of estimates /
judgments
12 Collateral assets
against loans
13 Objectives of
provisions / legal
constructive
14 Financial assets
impairment
15 Other assets
impairment
16 De-recognition of
financial assets
17 Risk management
18 Detailed risk
management
19 Objective of holding
derivatives / instruments
20 Contingent liabilities
21 Contingent assets
22 Inventory Lower of
Cost or Market
23 Key sources of
estimation uncertainty
Financial Instruments
24 Classifying
instruments by risks
25 Principal, stated
value, face value
26 Reclassification
of instruments
27 Cumulative/ change
in Fair value
Derivatives hedging
28 Hedging description
29 Change in Fair Value
of assets or liability
30 Cash flow hedge
Reserves
31 Statutory
32 Legal
33 Contingency /
special/ general
Segment information
34 Business major X X
segments
35 Geographical X X
concentration
36 Customer / X X
(asset/liabilities)
concentration
Financial and Other Risks
37 Operational risk / X X
Insurance risk
38 Market risk X X
39 Interest rate = X X
pricing risk sharia'a
40 Exchange rate X X
41 Liquidity X X
42 Credit X X
43 Pricing risk sharia'a
44 Tabular presentation
45 Sensitivity analysis
Ahmed et Meier, 1995 ICAEW 1997,
al., 2004 2000
General Risks Information
1 Competition in product X
market
2 Brand name erosion / X
change / addition
3 New alliances and X
joint ventures
4 Relationship to X
Government developments
plans
5 Customer acquisition X
processes
6 Recruiting of qualified X
and skilled professional
7 Regulations/Sharia's X
law/Overseas tax law
8 Events beyond balance X
sheet
9 Political environment X
10 Natural disasters X
Accounting Policies
11 Use of estimates / X
judgments
12 Collateral assets X
against loans
13 Objectives of
provisions / legal
constructive
14 Financial assets X
impairment
15 Other assets X
impairment
16 De-recognition of
financial assets
17 Risk management X
18 Detailed risk X
management
19 Objective of holding
derivatives / instruments
20 Contingent liabilities X
21 Contingent assets X
22 Inventory Lower of X
Cost or Market
23 Key sources of X
estimation uncertainty
Financial Instruments
24 Classifying
instruments by risks
25 Principal, stated
value, face value
26 Reclassification X
of instruments
27 Cumulative/ change
in Fair value
Derivatives hedging
28 Hedging description
29 Change in Fair Value
of assets or liability
30 Cash flow hedge
Reserves
31 Statutory
32 Legal
33 Contingency /
special/ general
Segment information
34 Business major X
segments
35 Geographical X
concentration
36 Customer / X
(asset/liabilities)
concentration
Financial and Other Risks
37 Operational risk / X
Insurance risk
38 Market risk X X
39 Interest rate = X X
pricing risk sharia'a
40 Exchange rate X
41 Liquidity X
42 Credit X
43 Pricing risk sharia'a X
44 Tabular presentation X
45 Sensitivity analysis X
AICPA 1987, UAE laws Dhanani,
1994 2003
General Risks Information
1 Competition in product X
market
2 Brand name erosion / X X
change / addition
3 New alliances and X
joint ventures
4 Relationship to X
Government developments
plans
5 Customer acquisition X
processes
6 Recruiting of qualified X
and skilled professional
7 Regulations/Sharia's X
law/Overseas tax law
8 Events beyond balance X X
sheet
9 Political environment X
10 Natural disasters
Accounting Policies
11 Use of estimates / X
judgments
12 Collateral assets X
against loans
13 Objectives of
provisions / legal
constructive
14 Financial assets
impairment
15 Other assets
impairment
16 De-recognition of
financial assets
17 Risk management X X
18 Detailed risk X X
management
19 Objective of holding
derivatives / instruments
20 Contingent liabilities
21 Contingent assets
22 Inventory Lower of
Cost or Market
23 Key sources of
estimation uncertainty
Financial Instruments
24 Classifying
instruments by risks
25 Principal, stated
value, face value
26 Reclassification
of instruments
27 Cumulative/ change
in Fair value
Derivatives hedging
28 Hedging description
29 Change in Fair Value
of assets or liability
30 Cash flow hedge
Reserves
31 Statutory X
32 Legal X
33 Contingency / X
special/ general
Segment information
34 Business major X
segments
35 Geographical X
concentration
36 Customer / X
(asset/liabilities)
concentration
Financial and Other Risks
37 Operational risk / X
Insurance risk
38 Market risk
39 Interest rate = X
pricing risk sharia'a
40 Exchange rate X X
41 Liquidity X
42 Credit X
43 Pricing risk sharia'a X
44 Tabular presentation
45 Sensitivity analysis
REFERENCES
Abraham, S. and Cox, P. (2007), "Analyzing the determinants of
narrative risk information in UK FTSE 100 annual reports", The
British Accounting Review, Vol. 39, pp. 227-248.
Ahmed, A. S., Beatty, A. and Bettinghus, B. (2004), "Evidence
on the efficiency of interest rate risk disclosures by commercial
banks", The International Journal of Accounting, Vol. 39, pp.
223-251.
Alfredson, K., Leo, K., Picker, R., Pacter, P., Radford, J. and
Wise, V. (2007), Applying International Financial Reporting Standards,
enhanced edition, John Wiley and Sons Australia, Ltd.
American Institute of Certified Public Accountants, (1994), Special
committee on financial reporting: Managing the information needs of
investors and creditors AICPA, Accounting Standards Division, New York.
Armitage S. and Marston, C., (2008), "Corporate disclosure,
cost of capital and reputations: Evidence from finance directors",
The British Accounting Review, Vol. 40, No, 4, pp. 314-336.
Banghoj, J. and Plenborg, T., (2008) "Value Relevance of
Voluntary Disclosure in the Annual Report", Accounting and Finance,
Vol. 48, No. 2, pp. 159-180.
Ball, R. and Brown, P. (1968), "An Empirical Evaluation of
Accounting Income Numbers", Journal of Accounting Research, Vol. 6,
pp. 159-178.
Barako, D., Hancock, P. and Izan, H. (2006), "Factors
influencing voluntary corporate disclosure by Kenyan companies",
Corporate Governance, Vol. 14, No. 2, pp. 107-125.
Beaver, W., Clarke, R., and Wright, W. (1979), "The
Association between Unsystematic Security Returns and Magnitude of
Earning Forecast Errors", Journal of Accounting Research, Vol. 17,
pp. 316-340.
Beretta, S. and Bozzolan, S. (2004), "A Framework for the
Analysis of Firm Risk Communication", The International Journal of
Accounting, Vol. 39, No. 3, pp. 265-288.
Cabedo, J. D. and Tirado, J. M. (2004), "The Disclosure of
Risk in Financial Statements", Accounting Forum, Vol. 28, No. 2,
pp. 181-200.
Chordia, T., Roll, R., Subrahamyam A. (2000), "Commonality in
Liquidity," Journal of Financial Economics, Journal of Financial
Economics, Vol. 56, pp. 3-28.
Chordia, T., Roll R., Subrahmanyam A., Anshuman, V. (2001)
"Trading Activities and Expected Stock Returns," Journal
ofFinancial Economics, Vol. 59, pp. 2-32.
Dhanani, A. (2003), "Foreign currency exchange risk
management: a case of the mining industry", The British Accounting
Review, Vol. 35, pp.35-63.
Diamond, Douglas W. and Verrecchia, Robert E. (1991),
"Disclosure, Liquidity, and The Cost of Capital", Journal of
Finance Vol: 46, pp. 1325-1360.
Duarte, J., Han, X., Harford, J., and Young L. (2007),
"Information asymmetry, information dissemination and the effect of
regulation FD on the cost of capital", Journal ofFinancial
Economics, Vol: 87, No. 1, pp. 24-44.
Easley, D. and O'hara, M. (2004), "Information and the
Cost of Capital," Journal of Finance, Vol. 59, No. 1, pp. 15531583.
Fama, E. F. (1970), "Efficient Capital Markets: A review of
theory and empirical work", Journal of Finance, Vol. 25, No. 2, pp.
383-417.
Fama, E. F. and French, K. R. (1992),"The Cross Section of
Expected Stock Returns", The Journal of Finance, Vol. 47, pp.
427-466.
Fama, E. F. and French, K. R. (1993), "Common risk factor in
the returns on stocks and bonds", Journal of Financial Economics,
Vol. 33, pp. 3-56.
Ferrell, A. (2007), "The Case Mandatory Disclosure in
Securities Regulation around the World", Brooklyn Journal of
Corporate Financial and Commercial Law, Vol. 2, pp. 81-132.
Gomes, A., Gorton, G., and Madureia, L. (2007), "SEC
Regulation Fair Disclosure, information, and the cost of capital",
Journal of Corporate Finance, Vol: 13, No. 2-3, pp. 300-334.
Hassan, M. K. (2009), "UAE corporations-specific
characteristics and level of risk disclosure", Managerial Auditing
Journal, Vol. 24, Issue 7, pp. 668-687.
Hooks, J. Coy, D. and Davey, H. (2002), "The Information Gap
in Annual Reports", Accounting Auditing and Accountability Journal,
Vol. 15, pp. 501-522.
Hughes, J. S., Liu, J. and Liu, J. (2007), "Information
Asymmetry, Diversification, and Cost of Capital" The Accounting
Review Vol. 82, pp. 705-730.
Hutton, A. P., (2007), "A discussion of corporate disclosure
by family firms", Journal of Accounting and Economics, Vol. 44, No.
1-2, pp. 287-297.
Institute of Chartered Accountants in England and Wales, (1997),
Financial reporting of risks: proposal for a statement of business risk,
issued by financial reporting committee at the ICAEW, London.
Institute of Chartered Accountants in England and Wales, (2000), No
surprise: the case for better risk reporting, issued by financial
reporting committee at the ICAEW, London.
Jensen, M. and Meckling, W. (1976), "Theory of Firm:
Managerial Behavior, Agency Cost, and Ownership Structure", Journal
ofFinancial Economics, Vol. 3, pp. 306-360.
Kim, O, and Verrecchia, R. E. (1991a), "Trading volume and
price reactions to public announcements", Journal of Accounting
Research, Vol. 29, pp. 302-321.
Kim, O, and. Verrecchia, R. E. (1991b), "Market reaction to
anticipated announcements", Journal of Financial Economics Vol. 30,
pp. 273-310.
Kim, O, and Verrecchia, R. E. (1994), "Market liquidity and
volume around earnings announcements", Journal of Accounting and
Economics, Vol. 17, pp. 41-67.
Kim, O. and Verrecchia, R. E. (2001), "The Relation Among
Disclosure, Returns, And Trading Volume Information", Accounting
Review Vol. 76, pp. 633-654.
Kraft, A., Andrew J., and Wasley, C. E. (2007),
"Regression-Based Tests of Market Pricing of Accounting Numbers:
The Mishkin Test and Ordinary Least Squares" Journal of Account
Research, Vol. 45, No. 5, pp. 1081-1113
Kuasirikun, N. and Sherer, M. (2004), "Corporate Disclosure in
Thailand", Accounting Auditing and Accountability Journal, Vol. 17,
pp: 629-659.
Lajili, K. and Zeghal, D. (2005), "Labor Cost Voluntary
Disclosures and Firm Equity Values: Is human capital information is
value relevant?", Journal of International Accounting, Auditing and
Taxation, Vol. 14, No. 2, pp.121-138.
Lambert, R., Leuz, C. and Verrecchia, R. E. (2007) "Accounting
Information, Disclosure, and the Cost of Capital", Journal of
Accounting Research Vol. 45, pp. 385-420
Leventis, S. and Weetman, P. (2004), "Voluntary Disclosures in
Emerging Capital Market: Evidence from Athens Stock Exchange",
Advances in International Accounting, Vol. 17, pp. 227-250.
Li, Feng, (2007), "Do Stock Market Investors Understand the
Risk Sentiment of Corporate Annual Reports?" Working paper,
University of Michigan.
Linsley, P. M. and Lawrence, M. (2007), "Risk Reporting by the
Largest UK Companies: Readability and Lack of Obfuscation--Research
Note", Accounting, Auditing and Accountability Journal, Vol. 20,
No. 4, pp. 620-627.
Linsley, P. M., and Shrives, P. J., (2006), "Risk Reporting: A
Study of Risk Disclosure in the Annual Reports of UK Companies",
The British Accounting Review, Vol. 38, No. 4, pp. 387-404.
Lopes, P. T. and Rodrigues, L. L. (2007), "Accounting for
financial instruments: an analysis of the determinants of disclosure in
the Portuguese stock exchange", The International Journal of
Accounting, Vol. 42, pp. 25-56.
Mahoney, P. G., (1995), "Mandatory Disclosure as a Solution to
Agency Problems", The University of Chicago Law Review, Vol. 62,
No. 3, pp. 1047-1112.
Marino, Y. M. and Matsusaka, N G. (2005), "Decision Processes,
Agency Problems, and Information: An Economic Analysis of Capital
Budgeting Procedures", The Review of Financial Studies, Vol. 18,
No. 1, pp. 301-325,
Meier, H. H.; Tomaszeweski, S. G.; and Tobing, R. (1995),
"Political risk assessment and disclosure in annual financial
reports: the case of Persian Gulf War", Journal of International
Accounting, Auditing and Taxation, Vol. 4, No.1, pp.49-68.
Mazzola, P.; Ravasi, D.; and Gabbioneta, C. (2006), "How to
Build Reputation in Financial Markets", Long Range Planning, Vol.
39, No. 4, pp. 385-407.
Naser, K., Al-Hussaini, A., Al-Kwari, D., and Nuseibeh, R. (2006),
"Determinants of Corporate Social Disclosure in Developing
Countries: The Case of Qatar", Advances in International
Accounting, Vol. 19, pp. 1-23.
Pastor, L., and Stambaugh, R. F. (2003), "Liquidity Risk and
Expected Stock Returns," Journal of Political Economy, Vol. 111,
No. 3, pp. 642-685.
Patel, S. A. and Dallas, G. (2002), "Transparency and
Disclosure: Overview of Methodology and Study Results" Standard
& Poor's T&D study report.
Patel, S. A.; Balic, A.; and Bwakira, L. (2002), "Measuring
transparency and disclosure at firm-level in emerging markets",
Emerging market Review, Vol. 3, No. 4, pp. 325-337.
Robb, S. W. G., Single, L. E. and Zarzeski, M. T. (2001), "Non
Financial Disclosures Across Ango-American Countries", Journal of
International Accounting, Auditing, and Taxation, Vol. 10, No. 1, pp.
71-83.
Salter, S. B., (1998), "Corporate Financial Disclosure in
Emerging Markets: Does economic development matter?", The
International Journal of Accounting, Vol. 33, No. 2, pp. 211-234.
Wang, K., Sewon, O., and Claiborne, M. C. (2008),
"Determinants and consequences of voluntary disclosure in an
emerging market: Evidence from China", Journal of International
Accounting, Auditing and Taxation, Vol. 17, No. 1, pp. 14-30
Md Hamid Uddin, University of Sharjah
Mostafa Kamal Hassan, University of Sharjah
ENDNOTES
(1) For example in the US, publicly traded firms must disclose in
their annual report detailed information on the firm's financial
results, its assets and financial condition, legal proceedings against
the firm, information on the firm's officers and directors. See
Item 303, Regulation S-K, Securities Exchange Act of 134, 17 C.F.R.
section 249.308a (2002).
(2) According to the Capital Asset Pricing Model (CAPM),
'market risk' determines the level of investors' risk
premium. However, firm size, book-to-market ratio, and stock liquidity
also play role in determining the level of risk premium [(Fama and
French, 1993), Chordia et. al., (2000 and 2001) and Pastor and Stambaugh
(2003)].
(3) Information also discloses through corporate announcements,
periodic and occasional reports submitted to the market authorities on
interim results, market sensitive information, and analysts'
reports. These types of information usually revealed before annual
report, and stock price adjusts around the event time. But the annual
report is a final document that incorporates all the information and
events of the last financial year as well as the information on future
business plans and strategies. Annual report thus reveals the
fundamental characteristics of a corporation which is useful for
cross-sectional analyses
(4) Phillip et. al (2000) found that BETA estimated using shorter
return interval over a period of less than three years results in lower
standard error.
(5) The risk information of Finance and Banking companies may be
different from those of the non-finance companies. This is because
Finance and Banking companies are governed by UAE Central Bank under the
Federal Financial Regulations that may require additional information
disclosure, e.g., percentage of non-performing loans (bad loans) and
sector-wise loans distribution.
Table 1: Average volatility (AVOL) of stocks across different
subsamples and periods after official publication of annual report
containing risk related information
Period after Classification of samples based on
official corporate risk disclosure (CRD) index
publication
of annual report 1 2 3 4
Very Low Low High Very High
Disclosure Disclosure Disclosure Disclosure
One Month 0.028 0.042 0.034 0.052
One Quarter 0.050 0.067 0.052 0.088
Half year 0.058 0.069 0.048 0.094
Full Year 0.048 0.056 0.040 0.070
Period after Difference t
official (4-1) statistics
publication
of annual report
One Month 0.024 2.151 **
One Quarter 0.038 1.346
Half year 0.036 1.380
Full Year 0.022 1.731
All of the AVOLs for different subsamples presented in column 1
through 4 are statistically significant at one percent level with t
values varying from 3.043 to 14.574. The average volatility (AVOL)
of stocks is calculated on weekly basis over different intervals of
period, e.g., one month, one quarter, half year, and full year
after official publication of annual report where the corporate
risk related information are disclosed. CRD is an index calculated
based on the number of risk related information disclosed in the
annual report. Samples companies are classified into four groups
based on their CRD index. Companies with CRD index below the below
the 1st quartile are classified as 'Very Low Disclosure' group,
companies with CRD index between the 1st and 2nd quartiles as "Low
Disclosure' group, companies with CRD index between the 2nd and 3rd
quartile as 'High Disclosure' group, and those with CRD index above
the 3rd quartile as 'Very High Disclosure' group. Asterisks ** and
* measure the level of significance at five percent and ten percent
levels respectively.
Table 2: Average excess volatility (AEVOL) of stocks across
different subsamples and periods after official publication of
annual report containing risk related information
Period after Classification of samples based on corporate
official risk related information disclosure (CRD) index
publication of
annual report 1 2 3 4
Very Low Low High Very High
Disclosure Disclosure Disclosure Disclosure
One Month -0.011 0.005 -0.007 0.013
One Quarter -0.009 0.012 -0.010 0.030
Half year -0.001 0.013 -0.013 0.035
Full Year 0.006 0.016 -0.004 0.028
Period after Difference t
official (4-1) statistics
publication of
annual report
One Month 0.024 1.816 *
One Quarter 0.039 1.225
Half year 0.036 1.296
Full Year 0.022 1.482
None of the AEVOLs for different subsamples presented in column 1
through 4 are statistically significant, except the full-year
AEVOLs for the 'low-disclosure' and 'very high disclosure'
subsamples that are significant with t values of 2.517 and 1.844
respectively. The average excess volatility (AEVOL) of stocks is
calculated on weekly basis over different intervals of period,
e.g., one month, one quarter, half year, and full year after
official publication of annual report where the corporate risk
related information are disclosed. CRD is an index calculated based
on the number of risk related information disclosed in the annual
report. Samples companies are classified into four groups based on
their CRD index. Companies with CRD index below the below the 1st
quartile are classified as 'Very Low Disclosure' group, companies
with CRD index between the 1st and 2nd quartiles as "Low
Disclosure' group, companies with CRD index between the 2nd and 3rd
quartile as 'High Disclosure' group, and those with CRD index above
the 3rd quartile as 'Very High Disclosure' group. Asterisks ** and
* measure the level of significance at five percent and ten percent
levels respectively
Table 3: Average systematic risk (BETA co-efficient) of stocks
across different subsamples over the one year period after
publication of annual report containing risk related information
Classification of samples based on risk
related information disclosure (CRD) index
1 2 3 4
Very Low Low High Very High
Disclosure Disclosure Disclosure Disclosure
1.3350 0.5391 0.8098 0.6919
Difference t
(4-1) statistics
-0.6431 -0.8923
The systematic risk (beta co-efficient) is calculated using weekly
returns over one year period after publication of annual report.
Samples companies are classified into four groups based on their CRD
index. Companies with CRD index below the below the 1st quartile are
classified as 'Very Low Disclosure' group, companies with CRD index
between the 1st and 2nd quartiles as "Low Disclosure' group,
companies with CRD index between the 2nd and 3rd quartile as 'High
Disclosure' group, and those with CRD index above the 3rd quartile
as 'Very High Disclosure' group
Table 4: Regression findings on the effect of corporate risk
disclosures (CRD) on the level of stock volatility and investors'
market risk
Panel A: Linear regression using CRD as the only explanatory
variable.
Explanatory Dependent Variables
Variable
AEVOL AEVOL
First Month First Quarter
Constant -0.021 (-1.309) -0.041 (-1.464)
CRD 0.001 (1.394) 0.002 (1.783) *
Adjusted [R.sup.2] 0.026 0.059
F Value 1.943 3.173 *
Explanatory Dependent Variables
Variable
AEVOL AEVOL
Half Year Full Year
Constant -0.036 (-1.417) -0.015 (-1.046)
CRD 0.002 (1.876) * 0.001 (1.950) *
Adjusted [R.sup.2] 0.067 0.074
F Value 3.519 * 3.802 *
Explanatory Dependent
Variable Variables
BETA
Constant 1.573 (2.233) **
CRD -0.036 (-1.097)
Adjusted [R.sup.2] 0.006
F Value 1.204
Panel B: Nonlinear (quadratic) regression using CRD as the only
explanatory variable
Explanatory Dependent Variables
Variable
AEVOL AEVOL
First Month First Quarter
Constant 0.006 (0.150) 0.066 (0.985)
CRD -0.002 (-0.457) -0.009 (-1.355)
CRD (2) 0.0001 (0.734) 0.0001 (1.737) *
Adjusted [R.sup.2] 0.013 0.111
F Value 1.228 3.191 *
Explanatory Dependent Variables
Variable
AEVOL AEVOL
Half Year Full Year
Constant 0.074 (1.225) 0.046 (1.321)
CRD -0.009 (-1.577) -0.005 (-1.498)
CRD (2) 0.0001 (1.986) ** 0.0001 (1.919) *
Adjusted [R.sup.2] 0.141 0.142
F Value 3.884 ** 3.893 **
Explanatory Dependent
Variable Variables
BETA
Constant 2.871 (1.627)
CRD -0.171 (-0.996)
CRD (2) 0.003 (0.803)
Adjusted [R.sup.2] -0.005
F Value 0.918
This table shows the univariate power of CRD in linear and quadratic
models to determine the stock volatility and investors' market risk.
CRD is an index of risk information calculated based on the number
of risk related information disclosed in the annual report. AEVOL is
the average excess volatility of stocks over the different intervals
after last financial year for which the annual report has been
published. BETA measures the investors' market risk and calculated
using weekly returns over one year period after last financial year.
The value in parenthesis is the t-statistic of regression co-
efficient. Asterisks ** and * measure the level of significance at
five percent and ten percent levels respectively.
Table 5: Multiple regression findings of the effect of corporate
risk disclosures (CRD) on the level of stock volatility and
investors' market risk
Explanatory Dependent Variables
Variable
AEVOL AEVOL
First Month First Quarter
Constant -0.045 -0.064
(-5.05) *** (-4.11) ***
CRD
CRD (2) 0.0001 0.0001
(2.33) ** (2.43) **
SIZE
DAR
TURN 0.277 0.065
(3.17) *** (2.37) **
FOWN -0.001
(-2,32) **
INDFIN
ADEX 0.026 0.055
(5,11) *** (4,13) ***
MARGIN
EPS
ROE 0.058 *
(1,93) *
Adjusted [R.sup.2] 0.567 0.304
F Value 10.176 *** 6.087 ***
Explanatory Dependent Variables
Variable
AEVOL AEVOL
Half Year Full Year
Constant -0.056 -0.032
(-3.15) *** (-6.03) ***
CRD
CRD (2) 0.0001 0.0001
(2.03) ** (4.50) ***
SIZE
DAR
TURN 0.022 0.008
(2,22) ** (2.20) **
FOWN
INDFIN
ADEX 0.053 *** 0.041
(5,35) *** (3.37) ***
MARGIN
EPS
ROE
Adjusted [R.sup.2] 0.345 0.515
F Value 7.141 *** 13.386 ***
Explanatory Dependent
Variable Variables
BETA
Constant 5.637
(4.22) ***
CRD -0.415
(-2.13) **
CRD (2) 0.008
(2.40) **
SIZE
DAR
TURN
FOWN
INDFIN
ADEX -1.090
(-2,09) **
MARGIN 1.195
(1,81) *
EPS
ROE
Adjusted [R.sup.2] 0.140
F Value 2.421 *
This table shows the effect of CRD variable in determining the
excess stock volatility and investors' market risk in different
setting of multiple regressions. We considered a number of other
explanatory variables based on literature and UAE market conditions.
These include total asset size (SIZE), debt asset ratio (DAR), stock
turnover in market trading (TURN), percentage of foreign ownership
(FOWN), listing as finance and banking company (INDFIN), listing on
Abu Dhabi Exchange (ADEX), profit margin (MARGIN), earning per share
(EPS), and return on equity (ROE). AEVOL is the average excess
volatility of stocks over the different intervals after last
financial year for which the annual report has been published. BETA
measures the investors' market risk and calculated using weekly
returns over one year period after last financial year. We have
calculated the cross-correlation among the explanatory variables to
examine the severity of multicolinearity problem. It is found that
all correlation coefficients are statistically insignificant, except
the coefficient of correlation (0.2777) between EPS and ROE that is
significant at 10 percent level. We apply stepwise regression method
to select only the significant variables to estimate the
parsimonious models that are presented in the table. Asterisks ***,
**, and * measure the level of significance at one percent, five
percent and ten percent levels respectively.
Table 6: Multiple regression findings of the effect of corporate
risk disclosures (CRD) on the level of GARCH (1,1) volatility
Explanatory Dependent Variables
Variable
AVOLGARCH AVOLGARCH
(First Month) (First Quarter)
Constant -0.0157 -0.0111
(-3.57) **** (-6.12) ***
CRD -0.0014 -0.0032
(-1.88) * (-1.69) *
CRD (2) 0.0002 0.0001
(4.12) *** (5.11) ***
SIZE
DAR
TURN 0.0651 0.0323
(1.77) * (2.35) **
FOWN
INDFIN
ADEX 0.0351 0.0441
(3.62) *** (2.21) **
MARGIN
EPS
ROE
Adjusted [R.sup.2] 0.377 0.401
F Value 9.11 *** 8.21 ***
Explanatory Dependent Variables
Variable
AVOLGARCH AVOLGARCH
(Half Year) (Full Year)
Constant -0.0112 -0.0211
(-7.23) *** (-8.25) ***
CRD
CRD (2) 0.0002 0.0002
(3.65) *** (3.89) ***
SIZE -0.0251 -0.0151
(-1.74) * (-1.69) *
DAR
TURN 0.0222 0.0124
(1.91) * (2.21) **
FOWN
INDFIN
ADEX 0.0121 0.0222
(4.39) *** (4.21) ***
MARGIN
EPS
ROE
Adjusted [R.sup.2] 0.461 0.555
F Value 15.21 *** 18.21 ***
This table shows the effect of CRD variable in determining the
GARCH (1,1) volatility in different setting of multiple
regressions. We considered a number of other explanatory variables
based on literature and UAE market conditions. These include total
asset size (SIZE), debt asset ratio (DAR), stock turnover in market
trading (TURN), percentage of foreign ownership (FOWN), listing as
finance and banking company (INDFIN), listing on Abu Dhabi Exchange
(ADEX), profit margin (MARGIN), earning per share (EPS), and return
on equity (ROE). AEVOL is the average excess volatility of stocks
over the different intervals after last financial year for which
the annual report has been published. We have calculated the
cross-correlation among the explanatory variables to examine the
severity of multicolinearity problem. It is found that all
correlation coefficients are statistically insignificant, except
the coefficient of correlation (0.2777) between EPS and ROE that is
significant at 10 percent level. We apply stepwise regression
method to select only the significant variables to estimate the
parsimonious models that are presented in the table. Asterisks ***,
**, and * measure the level of significance at one percent, five
percent and ten percent levels respectively.
