Master thesis | MSc BA Management Accounting & Control
THE QUALITY OF IFRS 7 RISK DISCLOSURES AND
FINANCIAL ANALYSTS’ EARNINGS FORECASTS
F.M. Jonker
S3751074
Supervisor dr. V.A. Porumb
Monday, 22
ndof June 2020
2
ABSTRACT
High-quality risk reporting has long been deemed useful because it assists the economic decisions of financial statements’ users. To cater to the informational needs of different firm stakeholders, standard setters and regulators, therefore, aim at increasing the quality of companies’ risk reporting. This thesis examines if IFRS 7 risk disclosures have an impact on financial analysts’ (1) earnings forecasts (2) firm following. I use hand-collected data from a sample of UK listed firms between the years 2011 and 2016. At odds with standard setters aims, I find that the IFRS 7 risk disclosure types (currency, price, liquidity, and remaining risk) are positively associated with financial analysts’ earnings forecasts characteristics (dispersion and error). Concurrently, I find that only credit risk has a significant effect on analysts following. Overall, my results suggest that, due to its complexity, IFRS 7 risk disclosures are not correctly incorporated in financial analysts’ earnings forecasts.
3
1. INTRODUCTION
During the past decade, corporate scandals and the 2008 global financial crisis have stemmed
a growing interest in the quality of risk reporting. Standard setters and regulators, therefore,
aimed to refine the quantity and quality of entities’ risk reporting (Elshandidy et al., 2018). The
International Accounting Standards Board (IASB) is one of the regulatory organisations that
has taken the initiative to enhance the disclosure requirements for firms. From January 1, 2007,
the IASB required firms to incorporate the revised “IFRS 7 Financial Instruments: Disclosure”
in their financial statements. According to IFRS 7, entities are obligated to disclose “the
significance of financial instruments for the entity’s financial position and performance” and “the nature and extent of risks arising from financial instruments … and how the entity manages those risks” (IFRS, 2020).
This regulatory development is important since risk disclosures generically reduce
information asymmetry between a firm and its stakeholders. When disclosures present
unexpected risks, this information is meaningful, and firm stakeholders incorporate it in their
assessment of a firms’ fundamental risk (Campbell et al., 2013). Specifically, the risk perception of investors increases when firms highlight in their disclosures unknow risk factors
and contingencies (Kravet & Muslu, 2013). Overall, this indicates that when the information
asymmetry between firms and stakeholders reduces, the risk perception of investors increases.
Additionally, prior research has shown that the quality of risk disclosures is determined by
either the requirements of standard setters (Bean & Irvine, 2015; Miihkinen, 2012), managers’
incentives to withhold bad news (Kothari et al., 2008), the institutional fit between the regulator
and regulate (Bischof et al., 2015), or risk comment letters from the SEC (Brown et al., 2018).
High-quality risk disclosures have, however, proven to increase the perceived riskiness among
different stakeholders (He et al., 2019; Karaibrahimoglu & Porumb, 2019; Kravet & Muslu,
4
sophisticated users of financial information. Analysts’ forecasts are used by investors and “can
serve as a reasonable proxy for the unobservable market expectation of earnings” (Givoly &
Lakonishok, 1984, p. 40). Reporting risk disclosures which are specific (Hope et al., 2016) and
which match the risk disclosure formats of hedged items and hedging instruments (He et al.,
2019) is essential for financial analyst since they enhance their understanding, and therefore
also the risk assessment. In this thesis, I explore whether IFRS 7 risk disclosures impact the financial analysts’ earnings forecasts.
Focusing on UK listed firms, I expect that the quantity of IFRS 7 risk disclosures
influences the characteristics of financial analysts’ earnings forecasts. The expectation is based
on the assumption that a reduction in information asymmetry between a firm and its
stakeholders is caused by providing more specific risk information (Hope et al., 2016), through
the overall quantity of IFRS 7 risk disclosures. In turn, this will improve financial analysts’
earnings forecasts due to financial analysts’ better understanding of the risks arising from firms’ financial instruments. To operationalise my research question, I use earnings forecast
dispersion, earnings forecast errors, and analysts following as characteristics of financial
analysts’ earnings forecasts. I use a sample of 896 observations from UK listed firms to run my estimations. I find that the quantity of IFRS 7 risk disclosures is not significantly associated
with the earnings forecasts of financial analysts. However, I find that the IFRS 7 risk disclosure
types currency, price, liquidity, credit, and remaining risk are positively and significantly
associated with financial analysts’ earnings forecasts characteristics. Specifically, results
indicate that firm disclosures regarding currency, price, and liquidity risk are positively
associated with the dispersion and currency, liquidity, and remaining risk are positively
associated with the error. Only credit risk has indicated to influence the analysts following. This
finding is consistent with the view that a more detailed description of firm risks, will be
5
correctly incorporated in financial analysts’ earnings forecasts. This finding provides support for the experimental study of He et al. (2019) and suggests that the qualitatively described IFRS
7 risk disclosures types affect financial analysts earnings forecasts. Being more specific (Hope
et al., 2016), by increasing the number of words used for IFRS 7 risk disclosure types, increases
the ambiguity in financial analysts earnings forecasts. Furthermore, this thesis contributes to
the literature by demonstrating that the quality of risk disclosures has an influence on users of
this information (Karaibrahimoglu & Porumb, 2019). Additionally, I add to prior literature by
providing evidence of the importance of risk disclosure types as a determinant of risk disclosure
quality.
The rest of the thesis is organised as follows: the literature review and hypotheses
development are provided in Section 2. Section 3 discusses the research methodology. Further,
the results are given in section 4 and finally, section 5, deals with the discussion and conclusion
of the thesis.
2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
2.1 Quality of risk disclosures and IFRS 7 Financial Instruments: Disclosure
According to IFRS 7, public companies need to disclose the significance of their financial
instruments and the risks that arise from the financial instruments (IFRS, 2020). The main
objective of the standard is to enhance the information on risks of a firms’ financial instruments.
IFRS 7 should lead to more transparency and, therefore, improve assistance to users of financial
statements. In 2018, the IASB issued the revised conceptual framework for developing
accounting policies. Companies who use this set of concepts for financial reporting needed to
make this framework effective after 1 January 2020 (IFRS, 2018).1
1 Appendix A1 contains the official document regarding the formation and objective of IFRS 7 Financial
6
Users of financial statements have an important reason why they confide in firms’ disclosed
information. They use this information to make well-informed economic decisions (Mihaela,
2008). Subsequently, disclosures in financial statements provide additional value when there is
information asymmetry between stakeholders and entities. Information asymmetry “exists when
a party or parties possess greater informational awareness pertinent to effective participation in a given situation relative to other participating parties” (Clarkson et al., 2007, p. 828).
Research has shown that the unexpected portion of risk factor disclosures is important for
investors since they incorporate this information in their assessment of the firms’ fundamental
risk (Campbell et al., 2013). Kravet & Muslu (2013) found that the risk perception of investors
is increased when risk disclosures point out unknown risk factors and contingencies. These
results indicate that when the information asymmetry between firms and investors reduces, the
risk perception of investors increases. In addition, when the information in risk disclosures is
presented quantitatively and specific (He et al., 2019; Hope et al., 2016), the risk assessments
of disclosures becomes more enhanced. Enhancement lessens the information asymmetry
between stakeholders and entities and therefore increases the perceived risk.
Accordingly, high-quality information in risk reporting should help users to make informed
economic decisions. In contrast, Bean and Irvine (2015) analysed the usefulness of disclosures
for derivatives and found that disclosures are prepared in a generic and uninformative way.
Additionally, they note that companies are focused on their year-end positions, which stipulates
that these positions may be unrepresentative of the transactions made during the year (Bean &
Irvine, 2015). Usefulness for users can be improved, and therefore the quality of the disclosures,
when companies incorporate more material risks than is currently required under IFRS 7 (Bean
& Irvine, 2015). In addition to these findings, Miihkinen (2012) found that the coercive effect
of the risk disclosure standard improves the quality of risk disclosures, that is information is
7
studies, the quality of disclosures would, therefore, depend on the requirements made by
standard setters.
However, the quality of the information within risk disclosures depends also on other factors.
Firstly, Kothari, Shu & Wysocki (2008, p. 241) expected the “magnitude of a negative stock
price reaction to bad news disclosures to be greater than the magnitude of the positive stock price reaction to good news”. The analysis found that managers, on average, indeed delay
releasing bad news to investors (Kothari et al., 2008). IFRS 7 risk disclosures can be seen as
carriers of bad news, as they comprise the risks of a firms’ financial instruments. So, once the
quality of the information within IFRS 7 risk disclosures increases, and therefore the amount
of bad news, this should lead to negative stock price reactions. On the other hand, IFRS 7 risk
disclosures have contractual value for banks, namely the higher the quality of these disclosures
on financial instruments, the lower the loan spread for firms (Karaibrahimoglu & Porumb,
2019). So, managers that delay sharing bad news to minimise the negative stock price reaction
could miss the opportunity of lower loan spreads from banks. Nonetheless, banks do charge
higher prices for loans when firms disclose higher levels of risks in comparison with assurance,
mainly because of the increased perceived riskiness (Karaibrahimoglu & Porumb, 2019).
Secondly, the success of regulation and therefore the quality of risk disclosures depends also
on the institutional fit between the regulator and regulated entities, namely “having multiple
regulators may lead to inconsistent implementation and enforcement of the same rules”
(Bischof et al., 2015, p. 1). Thirdly, Brown, Tian, & Tucker (2018) found that increased
disclosure specificity, based on the market leader’s risk comments from the SEC, reduces the firm’s probability of getting a comment letter from the SEC on its new filing. Indicating that the quality of the 10-K filing regarding risk disclosures has been up to the quality standards of
8
Notwithstanding, in 2013, the SEC outed concerns in their Disclosure Effectiveness
Initiative, about firms’ risk factor disclosures becoming more prolonged, repetitive, and less
specific. Beatty, Cheng, and Zang (2019) supported these concerns by finding that the
informativeness of risk factor disclosures has significantly decreased after the financial crisis
and that there has been an underlying change in disclosure behaviour of firms. Overall, taking
into account the previously mentioned research results, the information necessary for investors
and financial analysts to make well-informed decisions has become less apparent in risk
disclosures during the past years.
2.2 Financial analysts’ earnings forecasts
As mentioned, risk disclosures accomplish something fundamental; they reduce information
asymmetry between a firm and its stakeholders. One group of stakeholders are the financial
analysts’, who are sophisticated users of financial information. Financial analysts’ expectations regarding earnings forecasts are based on firms’ growth and profitability. Following, the
earnings forecasts “can serve as a reasonable proxy for the unobservable market expectation
of earnings” (Givoly & Lakonishok, 1984, p. 40).
Forecast dispersion, the variation in forecasts around the average forecast, gives insight into
the uncertainty of a firms’ future economic performance (Barron & Stuerke, 1998). According to theory, both uncertainty and lack of consensus among market participants about future events
are incorporated in forecast dispersion (Barry & Jennings, 1992). Several studies found that the
quality of financial disclosures influences the forecast dispersion of financial analysts. Dechow
et al. (1996) found that so-called violations of GAAP increased the forecast dispersion and
Swaminathan (1991) suggests that the newly released corporate segment information imposed
by the SEC decreased the forecast dispersion. Additionally, Ittner and Michels (2017) found
9
disclosures, by finding that greater sophistication regarding risk-based forecasting and planning
processes are associated with lesser earnings forecast errors. Earnings forecast error, the
difference between the actual earnings per share and the average estimated earnings per share
by financial analysts, gives insight into the accuracy of the financial analysts’ earnings
forecasts. Generally, larger and well-known firms have a higher analyst following relative to
other firms. Derouiche, Muessing, and Weber (2020) also suggest that firms have bigger analyst
coverage when they have a greater risk disclosure. Some investors believe that a firm with
analyst coverage even benefits from more investment activity than a firm without analyst
coverage. Overall, these studies indicate that a firms’ risk disclosure is associated with the
earnings forecasts of financial analysts.
2.3 Hypotheses development
The hypotheses formulated in this thesis are centred on whether IFRS 7 risk disclosures impact financial analysts’ earnings forecasts. The expectation is based on the assumption that a reduction in information asymmetry between a firm and its stakeholders is caused by providing
more specific risk information (Hope et al., 2016), through the overall quantity of IFRS 7 risk
disclosures. In turn, this will improve financial analysts’ earnings forecasts due to their better
understanding of the risks arising from firms’ financial instruments. To operationalise my
research question, I use earnings forecast dispersion, earnings forecast errors, and analysts
following as characteristics of financial analysts’ earnings forecasts.
Notwithstanding, financial analysts are sophisticated users of financial statements and can
retrieve information from numerous different reliable sources2. Moreover, the complexity of
hedging related actions is intricate and has been shown to be difficult to interpret correctly (He
2 While I acknowledge this explanation, I find it unlikely that financial analysts would not benefit from IFRS 7
10
et al., 2019). Therefore, the question remains if financial analysts genuinely benefit from IFRS
7 risk disclosures. Nonetheless, I expect that when the quantity of IFRS 7 risk disclosures
increases, financial analysts’ earnings forecast error and dispersion will decrease. Additionally,
I expect that an increase in the quantity of IFRS 7 risk disclosures will expand the number of
analysts following a firm. Accordingly, I hypothesise that:
Hypothesis 1a: The quantity of IFRS 7 risk disclosures is negatively associated with the
dispersion of analysts’ earnings forecasts.
Hypothesis 1b: The quantity of IFRS 7 risk disclosures is negatively associated with the
error of analysts’ earnings forecasts.
Hypothesis 1c: The quantity of IFRS 7 risk disclosures is positively associated with the
number of analysts following the firm.
Hypotheses 1a, 1b, and 1c would indicate that an increase in the quantity of risk reporting
decreases the variation in financial analysts’ earnings forecasts. However, a decrease in
information asymmetry, indicating an increase in quantity, doesn’t indicate the usefulness of
IFRS 7 risk disclosures to its users. Recent studies have shown that the disclosed items in
financial statements have had different implications for users of this information.
Karaibrahimoglu & Porumb (2019) found that the higher the quality of these disclosures on
financial instruments, the lower the loan spread for firms. Furthermore, the experimental study
of He et al. (2019) showed that a mismatch between the disclosures of hedged items and hedging
instruments, causes them to be less comparable, which therefore makes investors disregard the
relationship between the formats when assessing net risks. Additionally, analyses showed that “a consequence of making qualitative item disclosures is that participants assessed a hedged
11
presenting risks in specific ways, users assess risk disclosures differently. In turn, I expect that
the qualitatively described IFRS 7 risk disclosures types have an impact on the earnings
forecasts of financial analysts. Accordingly, this leads me to formulate an additional
expectation. Specifically, I expect that IFRS 7 risk disclosure types have an impact on the
earnings forecasts of financial analysts. Accordingly, I hypothesise that:
Hypothesis 2a: IFRS 7 risk disclosure type has an impact on the dispersion of analysts’
earnings forecasts.
Hypothesis 2b: IFRS 7 risk disclosure type has an impact on analysts’ earnings forecasts
errors.
Hypothesis 2c: IFRS 7 risk disclosure type has an impact on the number or financial analysts
following the firm.
See Figure 1 – Research framework for the conceptualisation of the hypotheses.
[Insert Figure 1 Here]
3. METHODOLOGY
3.1 Research sample
To test whether the quantity and type of IFRS 7 risk disclosures have an impact on financial
analysts’ earnings forecasts, I use data on UK listed firms. These firms are premium-listed on the LSE and are required to use IFRS in preparing their financial statements. The data that has
been used in testing is between the years 2011 and 2016. Specifically, I first extract archival
data for all UK firms on financial analysts’ earnings forecasts between 2011 and 2016 from the
IBES database. I match the hand-collected information on IFRS 7 risk disclosures and financial
12
information on IFRS 7 risk disclosure and financial analysts with firm-specific financial data
gathered from Compustat Global. Previously mentioned is also achieved based on the ISIN
codes and fiscal year ending dates.
The available IFRS 7 risk disclosures data provides us with the quantitative characteristics
of the disclosed information in the annual reports per firm. The quantitative data compresences
the total number of words used in the IFRS 7 disclosures, the total number of risk categories
addressed, and the total number of risk categories made quantitative in tables. Furthermore, the
data provides us with exposure to each type of risk at the firms’ reporting date. These types of
risk are the interest rate, currency, other price, liquidity, credit, and remaining risk.
The available data on financial analysts from the IBES database provides us information
about (1) the dispersion of earnings forecasts, (2) the error in earnings forecasts, and (3) the number of analysts’ following the firm.
After eliminating missing financial analysts’ data regarding earnings forecasts dispersion,
error, and number of analysts and missing financials for both IFRS 7 disclosures and financial
analysts’ data, I obtained a sample of 896 observations.
3.2 Research model 1
To test the predictions of Hypothesis 1a, 1b, and 1c, whether the quantity of IFRS 7 disclosures
affects financial analysts’ earnings forecasts, I developed the following equation:
Forecast_Characteristics = β0 + β1 IFRS_7_Disclosure_Quantity (1)
+ β3 Forecast_specific Controls + (Industry and Year dummies) + ε
I replace the dependent variable Forecast_Characteristics by the following items: (1)
13
Dispersion by dividing the standard deviation of earnings per share (EPS) forecasts by the share
price. I use this particular variable because it serves as a useful indicator of the uncertainty of
firms’ future earnings (Barron & Stuerke, 1998). Second, to test Hypothesis 1b, following Lusgarten and Mande (1998), I calculate the Error by taking the absolute value of the average
forecasted earnings minus the actual earnings divided by the share price. The error represents
the difference between the actual earnings and the estimated earnings by financial analysts. I
use this particular variable because it is likely to indicate the accuracy in the earnings forecasts.
Third, to test Hypothesis 1c, I use the variable Analysts_Following. This variable indicates the
total number of analysts following a firm and therefore specifies the analyst coverage in
earnings forecasts.
IFRS_7_Disclosure_Quantity
IFRS_7_Disclosure_Quantity is defined as the total amount of information disclosed in IFRS 7
risk disclosures, measuring the extent to which firms use words and tables. To assess the
quantity of the IFRS 7 risk disclosures, I use (1) Total_Words, (2) Risk_Catagories, and (3)
Hardscore. I use the variables since they represent the quantity of IFRS 7 risk disclosures. First,
following Miihkinen (2012), the Total_Words are the total number of words used in the IFRS
7 risk disclosures for each firm per fiscal year. Second, the Risk_Catagories are the total number
of risk categories mentioned in the IFRS 7 disclosures for each firm per fiscal year. Finally, by
building on He et al. (2019) that quantitative hedged item disclosures are assessed less risky, I
use Hardscore. These are the total number of risk categories made quantitative in tables for
each firm per fiscal year.
14
In all model estimations, I control for potential factors which may affect the association between
quantity of IFRS 7 risk disclosures and the financial analysts’ earnings forecast characteristics.
The standard control variables, firm size, leverage, profitability, and debt-paying ability are
used in testing. Additionally, I add the loss and negative equity dummy variables. The control
variables are further defined as follows: Size. Firm size is calculated by the natural logarithm
of the total assets. This control variable is included because larger firms are found to be
disclosing more information than smaller firms. Leverage. The firm’s debt-to-firm value ratio
equals total liabilities divided by total assets. For the calculation, the book value is used because
this does not reflect recent changes in the market’s valuation of the firm. Profitability. The
likelihood of high-performing firms defaulting in the future is far less than for low-performing
firms. Based on the operating income of firms, profitability is calculated by dividing the
earnings before tax and depreciation with the book value of total assets. Debt_Paying_Ability.
The firm’s ability to meet its short-term cash obligations is calculated by dividing the total
current assets by the total liabilities. When a firm has a good short-term financial strength, it
builds in believing it is a healthy business. Loss. A dummy variable which takes the value of 1
for firms that have a net income smaller than zero, and 0 otherwise. Negative_Equity. A dummy
variable which takes the value of 1 for firms when the total assets minus total liabilities is
smaller than zero, and 0 otherwise.
Additionally, I use industry and year indicator variables in all my estimations.
3.3 Research model 2
To test the predictions of Hypotheses 2a, 2b and 2c, whether the IFRS 7 risk disclosure type
has an impact on the financial analysts’ earnings forecasts, I developed the following equation:
15
+ β2 IFRS_7_Disclosure_specific Controls + β3 Forecast_specific Controls
+ (Industry and Year dummies) + ε
Where I replace – sequentially – the dependent variable Forecast_Characteristics by the
following items: (1) Dispersion, (2) Error, and (3) Analysts_Following.
IFRS_7_Disclosure_Type
IFRS_7_Disclosure_Type is defined as the amount of risk exposure a firm faces regarding
different types of risks, measuring the extent to which firms use words to describe these risks
in their IFRS 7 risk disclosures. To assess the IFRS 7 risk disclosures type, I use the following
items: (1) Interest_Rate_Risk, (2) Currency_Risk, (3) Other_Price_Risk, (4) Liquidity_Risk, (5)
Credit_Risk, and (6) Remaining_Risk. Interest_Rate_Risk represents the potential for
investment losses resulting from a change in interest rates. Second, Currency_Risk is the risk
that a currency changes in price in relation to another currency. Third, Other_Price_Risk are
the fluctuations caused by the changes in other market prices. Fourth, Liquidity_Risk is the risk
that a firm cannot meet its short-term financial demands. Fifth, Credit_Risk arises when there
is a possibility that the borrower defaults in repaying a loan or meet contractual obligations.
Finally, Remaining_Risk is the amount of risk exposure that remains, after factoring in all the
know risks. All these types of risks are most commonly referred to in IFRS 7 risk disclosures,
so, therefore, they are taken into consideration.
IFRS_7_Disclosure_specific Controls
To control for the IFRS 7 risk disclosures, I use (1) Total_Words, (2) Risk_Catagories, and (3)
16 3.4 Statistical analyses
The data at hand shows observations for different firms at different points in time. To analyse
this longitudinal data, I use the OLS linear regression model. The panel variable is set to be the
ISIN_codes and time variable is Year. I use the sandwich estimator to invalidate the need for
independent observations. By clustering based on ISIN codes, I require only that from cluster
to cluster the observations are independent.
4. RESULTS
4.1 Descriptive statistics
Table 1 presents the summary statistics for all variables used in the analyses. The summary
is based on all the observations available in the dataset. The average dispersion over the
complete dataset is 0.84, and the error is 1.44. The average number of analysts following a firm
is 15 (14.97).
Regarding the different types of IFRS 7 disclosure risks, the highest mean value is 487
(487.20) for Remaining_Risk, and the lowest mean value is 91 (90.99) for Other_Price_Risk.
Indicating that, in IFRS 7 risk disclosures, firms use most words for remaining risks and the
least for other price risk. Further, the average amount of words used in IFRS 7 risk disclosures
is 1757.
[Insert Figure 2 Here]
Finally, from the six possible risk categories mentioned in the IFRS 7 disclosures, the
average amount of categories mentioned by firms is 4 (4.15), and the average number of risk
categories made quantitative in tables is 3 (2.59).
17
[Insert Table 1 Here] 4.2 Pearson’s correlation matrix
The correlation coefficients between all variables are provided in Table 2. The results
demonstrate that the six risk disclosure types positively correlate with the quantity indicators
Total_Words, Risk_Catagories, and Hardscore. This finding is consistent with the view that
the quality of risk disclosures is at least partly a function of risk disclosure quantity.
Additionally, the Analysts_Following negatively and significantly correlates with earnings
forecast dispersion and error. This finding suggests that when the number of analysts following
a firm increases, the earnings forecasts of financial analysts’ become more enhanced.
However, the risk disclosure types and the earnings forecast dispersion and error have
non-significant Pearson correlation coefficients. Still, I expect these variables to have a non-significant
relation in the regressions, so I consider all indicators during testing.
[Insert Table 2 Here] 4.3 Main findings
Table 3 presents the results of the equation regarding Hypotheses 1a, 1b, and 1c – the effect of
the quantity of IFRS 7 risk disclosures on the dispersion, error, and number of analysts. As
shown, the coefficients of the IFRS_7_Disclosure_Quantiy are not significant, indicating that the quantity of IFRS 7 risk disclosures does not influence the dispersion of analysts’ earnings forecasts, the difference between the actual earnings and the estimated earnings by financial
analysts, and the number of analysts following the firm. Nonetheless, the results do indicate
that firm size is negative and significantly associated with the error. Which stipulates that when
firm size increases the difference between the actual earnings and the estimated earnings by
financial analysts decreases. This finding could be due to the fact that bigger firms disclose
18
regarding earnings forecasts. Additionally, profitability is negative and significantly associated
with dispersion and error. Indicating that an increase in profitability decreases the dispersion in
financial analysts’ earnings forecasts and the difference between the actual earnings and the estimated earnings by financial analysts. Which is reasonable, since bigger firms generally
generate more profit.
[Insert Table 3 Here]
Table 4 presents the results of the equation regarding Hypotheses 2a, 2b, and 2c – the effect
of IFRS risk disclosure types on the dispersion, error, and number of analysts. Model 1 shows
that Currency_risk, Other_price_risk, and Liquidity_risk have coefficients that are positive and
significant – indicating that when a firm increases the amount of disclosed information about
these types of risks in IFRS 7 risk disclosures, the dispersion of financial analysts’ earnings
forecasts will therefore also increase. Moreover, there are also three types of risks,
Currency_risk, Liquidity_risk, and Remaining_risk, that have coefficients that are positive and
significant in Model 2; however, now in association with the error. This finding indicates that
when a firm increases the amount of disclosed information about these types of risks in IFRS 7
risk disclosures, the difference between the actual earnings and the estimated earnings by
financial analysts will also increase. The earnings forecast of financial analysts will, therefore,
be less accurate, when taking into account these types of risks. Lastly, Model 3 only shows a
positive and significant association for Credit_risk. This result indicates that when a firm
increases the amount of disclosed information about credit risk, the number of analysts
following a firm will also increase.
The only type of risk that does not show a significant association with (one of) the dependent
variables is Interest_rate_risk. This finding indicates that there is no association between this
type of risk and the dispersion, error, or number of analysts following a firm.
19
Finally, I examine whether the results are robust to including the Analysts_Following as a
control variable for Hypothesis 2a and 2b. In the tests reported in Table 4, I assume that the
Dispersion and Error are independent of the Analysts_Following; however, the dispersion and
error may be directly associated with the number of analysts following a firm. However, all of
my initial results remain unchanged after including Analysts_Following as a control variable.
4.4 Additional test – Factor analysis
As an alternative for the linear regression models regarding the type of IFRS risk disclosures
on the dispersion, error, and number of analysts (Hypotheses 2a, 2b, 2c), I constructed a factor
using explanatory principal component factor analysis. I rotated the factor using the
orthogonal varimax method. Based on the rotated factor results, I have taken into account the
items with a factor loading higher than 0.40. The full construction of the factor analysis is
included in Appendix A3.
The results of the factor analysis indicated 1 factor with an eigenvalue greater than 1,
explaining 33.23% of the variance. This factor, named Risk types, contains the following items:
(1) Interest rate risk, (2) Currency risk, (3) Other price risk, (4) Liquidity risk, (5) Credit risk,
and (6) Remaining risk.
Table 5 presents estimations for the equations regarding Hypotheses 2a, 2b, and 2c. Model
1 shows a positive and significant result for Factor 1, which indicates that disclosing more
information on the types of risks, will lead to increased dispersion in financial analysts’ earnings forecasts. These results are in line with Table 4 regarding the Currency_risk, Other_price_risk,
and Liquidity_risk. Moreover, Model 2 also shows a positive and significant outcome for Factor
1, which indicates that the difference between the actual earnings and the estimated earnings by
financial analysts will increase when a firm discloses more information on the types of risks.
20
Remaining_risk. Overall, these results imply that the unfavourable impact of the IFRS 7 risk
disclosures regarding types of risks on the uncertainty and accuracy of earnings forecasts is
strengthened.
[Insert Table 5 Here]
5. DISCUSSION & CONCLUSION
Throughout the last years, accounting standard setters have aimed to develop the reporting
requirements for firms further to assist firm stakeholders better. Simultaneously, firms have
invested significant resources in complying with these enhanced reporting requirements.
Given this significantly important phenomenon, it is of crucial relevance to determine to what
extend reporting standards are relevant to users’ decision making. More specifically, for sophisticated users of financial information like financial analysts. In this thesis, I, therefore,
examine whether IFRS 7 risk disclosures are associated with financial analysts’ earnings
forecasts. I test whether the quantity and type of IFRS 7 risk disclosures influence the
dispersion of earnings forecasts, the error in earnings forecasts, and the number of analysts’ following the firm.
I find that IFRS 7 risk disclosures quantity does not influence financial analysts’ earnings
forecasts. This may be because financial analysts do not tend to mind the overall amount of
words and tables a firm uses to indicate their risks. In contrast, findings do demonstrate that
IFRS 7 risk disclosure types (currency, price, liquidity, and remaining risk) are positively
associated with financial analysts’ earnings forecasts characteristics (dispersion and error).
Concurrently, I find that only credit risk has a significant effect on analysts following. These
21
perceived riskiness. Indicating that IFRS 7 risk disclosures are not correctly incorporated in financial analysts’ earnings forecasts.
This finding provides support for the experimental study of He et al. (2019) and suggests
that the qualitatively described IFRS 7 risk disclosures types affect financial analysts earnings
forecasts. Being more specific (Hope et al., 2016), by increasing the number of words used for
IFRS 7 risk disclosure types, increases the ambiguity in financial analysts earnings forecasts.
Furthermore, this thesis contributes to the literature by demonstrating that the quality of risk
disclosures has an influence on users of this information (Karaibrahimoglu & Porumb, 2019).
Additionally, I add to prior literature by providing evidence of the importance of risk disclosure
types as a determinant of risk disclosure quality. Further, the results have managerial
implications for standard-setters and firms. Given that the IASB aims to achieve more
transparency and improve assistance to users of financial statements, this thesis provides insight
into the usefulness of IFRS 7 risk disclosures for financial analysts. The results indicate whether
the utility goal for IFRS 7, set by the IASB, is being met. Secondly, the results give firms a
better understanding of how their disclosed information, under IFRS 7, is perceived by its users.
Disclosing risks increasingly qualitative will enlarge the earnings forecasts characteristics. This
insight shows firms that the way of presenting risks does influence the overall risk assessments
of financial analysts.
Further, this thesis has limitations that can be considered a starting point for future research.
First, I used data recorded between 2011-2016 of UK listed firms. It is wise to repeat this
analysis for other countries or geographic areas during different periods. The cultural, political,
and social differences can play an essential role in the outcomes. By looking into a geographic
area as Europe, the results can become more generalisable. Secondly, I used different types of
UK listed firms. However, the outcomes can diverge when looking at specific branches and
22
industries. Finally, in this thesis, I used IFRS 7 risk disclosure quantity and types as quality
indicators. With other indicators, the outcomes may differ and give additional perspectives
23
REFERENCES
Barron, O. E. & Stuerke, P. S. (1998). Dispersion in Analysts’ Earnings Forecasts as a Measure of Uncertainty. Journal of Accounting, Auditing & Finance, 13(3), 245–270.
Barry, C. B. & Jennings, R. H. (1992). Information and Diversity of Analyst Opinion. Journal of Financial & Quantitative Analysis, 27(2), 169–183.
Bean, A. & Irvine, H. (2015). Derivatives disclosure in corporate annual reports: Bank analysts’ perceptions of usefulness. Accounting and Business Research, 45(5), 602–619.
Beatty, A., Cheng, L., & Zhang, H. (2019). Are Risk Factor Disclosures Still Relevant? Evidence from Market Reactions to Risk Factor Disclosures Before and After the Financial Crisis.
Contemporary Accounting Research, 36(2), 805–838.
Bischof, J., Daske, H., Elfers, F., & Hail, L. (2015). A Tale of Two Regulators: Risk Disclosures, Liquidity, and Enforcement in the Banking Sector. SSRN Electronic Journal, 1–50.
Brown, S. V., Tian, X. S., & Tucker, J. W. (2018). The Spillover Effect of SEC Comment Letters on Qualitative Corporate Disclosure: Evidence from the Risk Factor Disclosure. Contemporary Accounting Research, 35(2), 622–656.
Campbell, J. L., Chen, H., Dhaliwal, D. S., Lu, H., & Steele, L. B. (2013). The information content of mandatory risk factor disclosures in corporate filings. Review of Accounting Studies, 19, 396– 455.
Clarkson, G., Jacobsen, T. E., & Batcheller, A. L. (2007). Information asymmetry and information sharing. Government Information Quarterly, 24(4), 827–839.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1996). Causes and Consequences of Earnings Manipulation: An Analysis of Firms Subject to Enforcement Actions by the SEC. Contemporary Accounting Research, 13(1), 1–36.
Derouiche, I., Muessig, A., & Weber, V. (2020). The effect of risk disclosure on analyst following. The European Journal of Finance, 1–22.
24
Givoly, D. & Lakonishok, J. (1984). The Quality of Analysts’ Forecasts of Earnings. Financial Analysts Journal, 40(5), 40–47.
He, Y., Tan, H.-T., Yeo, F., & Zhang, J. (2019). When Do Qualitative Risk Disclosures Backfire? The Effects of a Mismatch in Hedge Disclosure Formats on Investors’ Judgments. Contemporary Accounting Research, 36(4), 2093–2112.
Hope, O.-K., Hu, D., & Lu, H. (2016). The Benefits of Specific Risk-Factor Disclosures. Review of Accounting Studies, 21, 1005–1045.
IFRS. (2018, March). IFRS Conceptual Framework Project Summary. IFRS Conceptual Framework Project Summary. https://www.ifrs.org/-/media/project/conceptual-framework/fact-sheet-project-summary-and-feedback-statement/conceptual-framework-project-summary.pdf IFRS. (2020). IFRS 7 Financial Instruments: Disclosures. IFRS 7 Financial Instruments: Disclosures.
https://www.ifrs.org/issued-standards/list-of-standards/ifrs-7-financial-instruments-disclosures/
IFRS Foundation. (2019). IFRS 7 Financial Instruments: Disclosures. eifrs.ifrs.org/eifrs/bnstandards/en/IFRS7.pdf
Ittner, C. D. & Michels, J. (2017). Risk-Based Forecasting and Planning and Management Earnings Forecasts. Review of Accounting Studies, 22, 1005–1047.
Karaibrahimoglu, Y. Z. & Porumb, V. A. (2019). The interplay between IFRS 7 disclosures and the expanded audit report: Evidence from syndicated loan contracting. 1–40.
Kothari, S. P., Shu, S., & Wysocki, P. D. (2008). Do Managers Withhold Bad News? Journal of Accounting Research, 47(1), 241–276.
Kravet, T. & Muslu, V. (2013). Textual risk disclosures and investors’ risk perceptions. Review of Accounting Studies, 18(4), 1088–1122.
Lustgarten, S. & Mande, V. (1998). The effect of insider trading on financial analysts’ forecast accuracy and dispersion. Journal of Accounting and Public Policy, 17, 311–327.
25
Miihkinen, A. (2012). What Drives Quality of Firm Risk Disclosure? The International Journal of Accounting, 47(4), 437–468.
26 FIGURES
Figure 1: Research Framework
Figure 2: IFRS 7 Risk Disclosure Types over the years
Figure 3: Quantity of IFRS 7 Risk Disclosure over the years Financial analysts'
earnings forecasts Quantity of overall
IFRS 7 risk disclosures (H1)
27
TABLE 1: Summary Statistics
Variable Obs. Mean Std. Dev. Min. Max.
Dispersion 1,389 0.84 7.42 0.00 192.80 Error 1,389 1.44 11.49 0.00 247.70 Number_Analysts 1,389 14.97 8.98 1.00 44.00 Interest_Rate_Risk 1,935 262.00 204.20 0.00 1626.00 Currency_Risk 1,935 348.50 266.30 0.00 2257.00 Other_Price_Risk 1,935 90.99 239.60 0.00 3543.00 Liquidity_Risk 1,935 289.10 214.60 1.00 740.00 Credit_Risk 1,935 265.20 192.30 0.00 2075.00 Remaining_Risk 1,924 487.20 484.70 0.00 4882.00 Total_Words 1,935 1757.00 949.90 92.00 6919.00 Risk_Catagories 1,935 4.15 0.73 1.00 5.00 Hardscore 1,935 2.59 1.57 0.00 6.00 Size 2,516 6.67 1.81 1.82 11.42 Leverage 2,516 0.59 0.27 0.03 2.81 Profitability 2,515 0.10 0.20 -4.89 5.37 Debt_Paying_Ability 2,484 0.94 1.04 0.01 20.28 Loss 2,515 0.81 0.39 0.00 1.00 Negative_Equity 2,516 0.24 0.43 0.00 1.00
28
TABLE 2: Pearson Correlation Matrix
29
TABLE 3: Quantity of IFRS 7 on Earnings Forecasts
Model 1 Model 2 Model 3
Dependent variables: Dispersion Error Analyst_Following
IFRS_7_Disclosure_Quantity Total_Words -0.034 0.014 -0.054 (0.038) (0.038) (0.042) Risk_Catagories -0.050 -0.020 0.037 (0.046) (0.048) (0.042) Hardscore -0.032 -0.050 0.018 (0.051) (0.061) (0.033) Forecast_specific Controls Size -0.131 -0.291*** 0.866*** (0.093) (0.088) (0.054) Leverage 0.185*** 0.260** 0.045 (0.064) (0.113) (0.043) Profitability -0.208*** -0.330*** 0.033 (0.055) (0.087) (0.037) Debt_Paying_Ability 0.045 0.098 0.086* (0.039) (0.059) (0.052) Loss -0.036 -0.143* -0.133** (0.125) (0.083) (0.064) Negative_Equity 0.001 0.017 -0.070 (0.079) (0.087) (0.063) Constant -0.647*** -0.469*** -1.797*** (0.182) (0.131) (0.123) Observations 896 896 896 Number of ISIN 184 184 184 R-squared 0.121 0.110 0.646
30
TABLE 4: IFRS 7 Disclosure Type on Earnings Forecasts
Model 1 Model 2 Model 3
Dependent variables: Dispersion Error Analyst_Following
IFRS_7_Disclosure_Type Interest_Rate_Risk 0.123 0.127 0.001 (0.087) (0.091) (0.043) Currency_Risk 0.161** 0.106* -0.019 (0.074) (0.057) (0.049) Other_Price_Risk 0.078* 0.026 0.070 (0.046) (0.048) (0.058) Liquidity_Risk 0.080** 0.077** 0.034 (0.039) (0.037) (0.033) Credit_Risk 0.073 0.022 0.073* (0.056) (0.046) (0.040) Remaining_Risk 0.119 0.133* 0.017 (0.072) (0.074) (0.058) Disclosure_specific Controls Total_Words -0.363** -0.273 -0.112 (0.179) (0.170) (0.107) Risk_Catagories -0.092* -0.033 0.013 (0.055) (0.051) (0.043) Hardscore -0.048 -0.060 0.022 (0.046) (0.054) (0.035) Forecast_specific Controls Number_Analysts Size -0.132 -0.274*** 0.862*** (0.092) (0.088) (0.055) Leverage 0.187*** 0.260** 0.049 (0.065) (0.113) (0.043) Profitability -0.208*** -0.328*** 0.036 (0.055) (0.087) (0.037) Debt_Paying_Ability 0.045 0.106* 0.088* (0.039) (0.058) (0.051) Loss -0.023 -0.124 -0.139** (0.123) (0.081) (0.064) Negative_equity -0.006 0.008 -0.072 (0.078) (0.086) (0.063) Constant -0.658*** -0.417*** -1.915*** (0.201) (0.153) (0.132) Observations 892 892 892 Number of ISIN 184 184 184 R-squared 0.131 0.114 0.653
31
TABLE 5: Factor analysis - IFRS 7 Disclosure Type
Model 1 Model 2 Model 3
Dependent variables: Dispersion Error Analyst_Following
Factor 1 - Risk types 0.302** 0.185* 0.086
(0.120) (0.112) (0.091) Disclosure_specific Controls Total_Words -0.279** -0.137 -0.124 (0.109) (0.103) (0.079) Risk_Catagories -0.089 -0.044 0.027 (0.055) (0.053) (0.044) Hardscore -0.048 -0.059 0.014 (0.049) (0.058) (0.034) Forecast_specific Controls Size -0.141 -0.297*** 0.866*** (0.092) (0.090) (0.055) Leverage 0.189*** 0.263** 0.046 (0.064) (0.112) (0.043) Profitability -0.208*** -0.330*** 0.033 (0.055) (0.087) (0.037) Debt_Paying_Ability 0.043 0.097 0.088* (0.040) (0.060) (0.051) Loss -0.029 -0.138* -0.128** (0.121) (0.083) (0.064) Negative_Equity 0.003 0.017 -0.077 (0.079) (0.087) (0.063) Constant -0.733*** -0.522*** -1.818*** (0.172) (0.138) (0.125) Observations 892 892 892 Number of ISIN 184 184 184 R-squared 0.128 0.112 0.650
*** p<0.01, ** p<0.05, * p<0.1 Robust standard errors in parentheses. In the regressions, all continuous variables in the are winsorised and standardised. All variables are described in Table A2.
32 APPENDIX
Appendix A1: IFRS 7 Financial Instruments: Disclosures IFRS 7
Financial Instruments: Disclosures
In April 2001 the International Accounting Standards Board (Board) adopted IAS 30 Disclosures in the Financial Statements of Banks and Similar Financial Institutions, which had originally been issued by the International Accounting Standards Committee in August 1990.
In August 2005 the Board issued IFRS 7 Financial Instruments, which replaced IAS 30 and carried forward the disclosure requirements in IAS 32 Financial Instruments: Disclosure and Presentation. IAS 32 was subsequently renamed as IAS 32 Financial Instruments: Presentation. IAS 1 Presentation of Financial Statements (as revised in 2007) amended the terminology used throughout IFRS, including IFRS 7. In March 2009 the IASB enhanced the disclosures about fair value and liquidity risks in IFRS 7. The Board also amended IFRS 7 to reflect that a new financial instruments Standard was issued—IFRS 9 Financial Instruments, which related to the classification of financial assets and financial liabilities. IFRS 7 was also amended in October 2010 to require entities to supplement disclosures for all transferred financial assets that are not derecognised where there has been some continuing involvement in a
transferred asset. The Board amended IFRS 7 in December 2011 to improve disclosures in netting arrangements associated with financial assets and financial liabilities.
In May 2017 when IFRS 17 Insurance Contracts was issued, it added disclosure requirements for when an entity applies an exemption for specified treasury shares or for an entity’s own repurchased financial liabilities in specific circumstances.
In September 2019 the Board amended IFRS 9 and IAS 39 by issuing Interest Rate Benchmark Reform to provide specific exceptions to hedge accounting requirements in IFRS 9 and IAS 39 for (a) highly probable requirement; (b) prospective assessments; (c) retrospective assessment (IAS 39 only); and (d) separately identifiable risk components. Interest Rate Benchmark Reform also amended IFRS 7 to add specific disclosure requirements for hedging relationships to which an entity applies the exceptions in IFRS 9 or IAS 39.
Other Standards have made minor amendments to IFRS 7. They include Limited Exemption from
Comparative IFRS 7 Disclosures for First-time Adopters (Amendments to IFRS 1) (issued January 2010), Improvements to IFRSs (issued May 2010), IFRS 10 Consolidated Financial Statements (issued May 2011), IFRS 11 Joint Arrangements (issued May 2011), IFRS 13 Fair Value Measurement (issued May 2011), Presentation of Items of Other Comprehensive Income (Amendments to IAS 1) (issued June 2011), Mandatory Effective Date and Transition Disclosures (Amendments to IFRS 9 (2009), IFRS 9 (2010) and IFRS 7) (issued December 2011), Investment Entities (Amendments to IFRS 10, IFRS 12 and IAS 27) (issued October 2012), IFRS 9 Financial Instruments (Hedge Accounting and amendments to IFRS 9, IFRS 7 and IAS 39) (issued November 2013), Annual Improvements to IFRSs 2012–2014 Cycle (issued September 2014), Disclosure Initiative (Amendments to IAS 1) (issued December 2014), IFRS 16 Leases (issued January 2016) and Annual Improvements to IFRS Standards 2014–2016 Cycle (issued December 2016).
33 Objective
1 The objective of this IFRS is to require entities to provide disclosures in their financial statements that enable users to evaluate:
(a) the significance of financial instruments for the entity’s financial position and performance; and
(b) the nature and extent of risks arising from financial instruments to which the entity is exposed during the period and at the end of the reporting period, and how the entity manages those risks.
2 The principles in this IFRS complement the principles for recognising, measuring and presenting financial assets and financial liabilities in IAS 32 Financial Instruments: Presentation and IFRS 9 Financial Instruments.
34
Appendix A2: List of Variables
Variable Description
Dispersion The standard deviation of earnings per share (EPS) forecasts divided by the absolute value of the mean EPS forecast;
Error The absolute value of the average forecasted earnings minus the actual earnings divided by the share price;
Analysts_Following The total number of analysts following for each firm per fiscal year; Interest_Rate_Risk The potential for investment losses resulting from a change in interest rates; Currency_Risk The risk that a currency changes in price in relation to another currency; Other_Price_Risk The fluctuations caused by the changes in other market prices;
Liquidity_Risk The risk that a firm cannot meet its short term financial demands;
Credit_Risk Arises when there is a possibility that the borrower defaults in repaying a loan or meet contractual obligations;
Remaining_Risk The amount of risk exposure that remains, after factoring in all the know risks; Total_Words The total number of words used in the IFRS 7 risk disclosures for each firm per
fiscal year;
Risk_Catagories The total number of risk categories mentioned in the IFRS 7 disclosures for each firm per fiscal year;
Hardscore The total number of risk categories made quantitative in tables for each firm per fiscal year;
Size The natural logarithm of total assets (from Compustat Global);
Leverage The firm’s debt-to-firm value ratio equals total liabilities divided by total assets (from Compustat Global);
Profitability Based on the operating income of firms, profitability is calculated by dividing the earnings before tax and depreciation with the book value of total assets (from Compustat Global);
Debt_Paying_Ability The total current assets divided by the total liabilities (from Compustat Global); Loss A dummy variable which takes the value of 1 for firms that have a net income
smaller than zero, and 0 otherwise;
35
Appendix A3: Construction of the Factor Analysis
Observations 2681
Factor analysis/correlation
Method Principal-component factors
Rotation Unrotated
Principal component factors; 1 factor retained
Factor Eigenvalue Difference Proportion Cumulative
Factor1 1.994 1.002 0.332 0.332 Factor2 0.992 0.0932 0.165 0.498 Factor3 0.899 0.0307 0.15 0.647 Factor4 0.868 0.171 0.145 0.792 Factor5 0.697 0.147 0.116 0.908 Factor6 0.55 . 0.0917 1 Factor loadings
Variable Factor1 Uniqueness
Interest_Rate_Risk 0.599 0.641 Currency_Risk 0.719 0.483 Other_Price_Risk 0.424 0.82 Liquidity_Risk 0.469 0.78 Credit_Risk 0.703 0.506 Remaining_Risk 0.472 0.777 Factor analysis/correlation
Method: principal-component factors Rotation: orthogonal varimax (Kaiser off)
Factor Eigenvalue Difference Proportion Cumulative
Factor1 1.994 . 0.332 0.332
Rotated factor loadings
Variable Factor1 Uniqueness
Interest_Rate_Risk 0.599 0.641 Currency_Risk 0.719 0.483 Other_Price_Risk 0.424 0.82 Liquidity_Risk 0.469 0.78 Credit_Risk 0.703 0.506 Remaining_Risk 0.472 0.777
Factor rotation matrix
Factor1
