Disclosure, Enforcement and the Valuation of Equity

Empirical evidence shows that higher levels of disclosure and enforcement do not consistently translate into higher firm valuations. This observation implies a real-life setting in which a richer information environment can shift investors’ risk premiums upwards or downwards. However economic literature does not convincingly explain why adverse effects can occur. To fill this gap in research, we provide a model beyond the standard principal-agent framework that can explain valuation effects from augmented disclosure and enforcement regulation. We also show that country differences in regulatory effectiveness do not align with the legal system, but instead with structural strengths or difficulties of an economy. In a six country setting, investigating Canada, France, Germany, Japan, the UK and the U.S., we systematically capture regulatory changes and document varying valuation effects from mandatory disclosure regulation. Our analysis shows that valuation effects are driven by the share of “bad news” firms, which is higher in economies with structural difficulties. Effects from higher levels of disclosure are thus neither generalizable across economies nor dependent on the legal system as previously hypothesized.


Introduction
Almost every major country has witnessed increasing emphasis on the regulation of financial reporting. The arrival of transnational International Financial Reporting Standards (IFRS) or the advent of new national supervisory agencies represent only the most outstanding examples of ever increasing levels of disclosure and enforcement regulation (Christensen et al. 2013). Literature associates the introduction of more regulation typically with a higher information level and predicts a positive effect on liquidity, cost of equity and equity valuation (e.g.; (Lambert et al. 2007); (Strohmenger 2014); (Verrecchia 2001)).
Many empirical studies have assessed the relationship between changes in financial reporting regulation and equity valuation (see e.g. (Brüggemann et al. 2013); (Healy & Palepu 2001); (Leuz & Wysocki 2016) for an overview). There is substantial empirical evidence for positive capital market effects from event studies (e.g. (Eleswarapu et al. 2004); (Ernstberger et al. 2012); (Hope 2003)), especially for the introduction of the IFRS (e.g. (Brown et al. 2014); (Byard et al. 2011); (Christensen et al. 2013); (Daske et al. 2008); (Daske et al. 2013); (Frino et al. 2013); (Horton et al. 2013); long-term studies corroborate most findings (e.g. (Core et al. 2015); (Daouk et al. 2006); (Dhaliwal et al. 2014); (Fu et al. 2012); (Meser et al. 2015)). However, there is also scattered evidence for a negative association of new disclosure regulations and equity valuation. (Armstrong et al. 2010) find negative capital market effects for European code law countries adopting IFRS. At closer inspection, by using a dummy variable for the legal system their sample may have been dominated by individual countries (Ireland and the UK being the only two common law out of the 18 EU countries), and it is very likely that their documented effect is at the country-level, and not at the level of the legal system. In addition, (Hassan et al. 2009) conclude in their longterm study for the Egyptian market that there is a positive relationship between voluntary disclosure and firm value, but a negative one for mandatory disclosure regulation. They denote their results as "puzzling" and try to explain it ad hoc by regulatory costs. Both studies show that an explanation for the varying country-specific valuation effects in the course of mandatory disclosure regulation remains an open issue.
We provide a model that can explain positive or negative effects from a change in accounting regulation without invoking regulatory costs. In conformity with the majority of literature on capital market effects of accounting regulation we do not think that costs due to a higher level of mandatory disclosure regulation play such a major role that they outweigh the benefits. Instead, we posit the existence of systematic reasons for negative valuation effects when many firms are required to disclose of previously withheld bad news. Country-specific valuation effects depend on the heterogeneity of firms' disclosures within a country, controlled on the enforcement environment. Our model allows predicting capital market effects for specific countries based on the proportion of good news and bad news firms. A high proportion of bad news firms is likely to exist in countries with microeconomic or macroeconomic structural problems. Equals 1 if companies are required to follow international accounting standards for their consolidated financial statements; equals 0.5 if companies have the choice between local GAAP and international accounting standards; equals 0 otherwise.

D2: Management Commentary
Rating for the complexity of the Management Discussion and Analysis (MD&A, as it is named in the US; e.g. in the UK it is named Operating and Financial Review, in Germany Lagebericht). Equals 1, 0.75, 0.5, 0.25 or if no such report exists 0. D3: Material Information Equals 1 if companies are required to pursue ad hoc disclosure; equals 0 otherwise.

D4: Quarterly Reports
Equals 1 if companies are required to publish quarterly reports; equals 0.5 if reports have to be published semi-annually; equals 0 otherwise.

D5: Segment Information
Equals 1 if companies are required to publish detailed information on business segments and regional segments; equals 0.5 if companies are free to choose the extent of segment information disclosure; equals 0 otherwise.

D6: Compensation
Equals 1 if companies are required to disclosure total compensation of each top manager; equals 0.5 if regulation only requires the disclosure of the aggregate compensation of top management; equals 0 otherwise.

D7: Inside Ownership
Equals 1 if firms are required to disclose the number of shares owned by each member of the management; equals 0.5 if regulation only requires the aggregate number of the shares owned by the management to be disclosed; equals 0 otherwise .

D8: Prospectus
Equals 1 if companies are required to set up a prospectus for potential investors on any public offering; equals 0.5 if there exist material exemptions from publishing a prospectus, e.g. regarding tender size; equals 0 otherwise.

D9: Shareholders
Equals 1 if companies are required to disclose details (e.g. name, share property) of shareholders that directly or indirectly control at least 10 % of the company's stock; equals 0.5 if regulation only requires the disclosure of the direct or aggregate ownership of these shareholders; equals 0 otherwise.

D10: Transactions
Equals 1 if companies are required to disclosure all capital market transactions that involve related parties; equals 0.5 if only some transactions between the company and related parties have to be disclosed; equals 0 otherwise.

Disclosure Score Value
Calculated as the average of items D1 to D10 and ranges between 0 and 1 Research Journal of Finance and Accounting www.iiste.org ISSN 2222-1697 (Paper) ISSN 2222-2847 (Online) DOI: 10.7176/RJFA Vol.10, No.8, 2019  Equals 1 if the Enforcement Institution has the legal power to demand documents from all persons that are related to the suspicious firm; equals 0.5 if the Enforcement Institution has the legal power to only demand documents from the directors of suspicious publicly-traded corporations; equals 0 otherwise.

E7: Review
Equals 1 if the Enforcement Institution requires a continuous filing of disclosure documents and submits them to a systematic, proactive and reactive review; equals 0.5 if a filing and only a reactive review is statutory, equals 0 otherwise.

E8: Sanctions Company
Equals 1 if the Enforcement Institution has far reaching competencies to impose financial and other legal penalties against companies; equals 0.5 if the Enforcement Institution can only impose financial penalties; equals 0 otherwise.

E9: Sanctions Management
Equals 1 if accountants/managers can be held criminally liable when they are unaware of fraud and misleading information; equals 0.5 if the accountants/managers can be held criminally liable when they are aware that the financial statements are misleading; equals 0 otherwise. E10: Sworn Statement Equals 1 if the directors have to sign for the material accuracy of financial statements; equals 0 otherwise. Enforcement Score Value Calculated as the average of items E1 to E10 and ranges between 0 and 1 With the exception of Management Commentary, which allows a differentiation between five levels, all items can take values of 0, 0.5 and 1, where 1 represents the highest regulatory activity. Score values change in time series whenever a new regulatory measure becomes effective. We consider all mandatory changes in disclosure and enforcement regulation. The overall score values are calculated as the average of the ten items and ranges between 0 and 1. We choose this equal weighting of score items to avoid subjective bias.
In the following, we analyze disclosure and enforcement regulation for the six countries (ordered by market size): U.S., Japan, UK, Germany, France and Canada between 1991 and 2010. We review all relevant rules, regulations and competences in the six countries to classify mandatory disclosure requirements and their enforcement, but we do not distinguish between a country's stock market segments as there are only small regulatory differences between the segments within a country. The appendix provides a detailed analysis of the country-specific score values and their changes over time for the six country cases.
The results of the leximetric analysis of the six countries can be compared in the easiest and most illustrative way by charting their yearly score values for disclosure and enforcement regulation, as presented in Figures 2 and 3. Three observations are apparent: First, mostly upward changes can be observed. Only the enforcement scores of Japan and the UK decrease, albeit temporarily. Our leximetric analysis adds further evidence that regulators around the world tighten the screw on disclosure and enforcement regulation. Second, especially the years after the enactment of the SOX are characterized by more frequent and steeper increases. Third, and not astonishingly, the U.S. are setting the pace for all other five countries. Even Germany and Japan, with mostly low score values,  Vol.10, No.8, 2019 31 dramatically catch up after 2003. Figure 2: Development of the disclosure scores in a cross country comparison This figure shows the development of the disclosure scores for the U.S., Japan, UK, Germany, France and Canada in comparison for the period between 1991 and 2010. The score ranges between 0 and 1, and is calculated as the average of ten variables capturing the independence and scope of an enforcer as described in Panel B of Table 1.

Figure 3: Development of the enforcement scores in a cross country comparison
This figure shows the development of the enforcement scores for the U.S., Japan, UK, Germany, France and Canada in comparison for the period between 1991 and 2010. The score ranges between 0 and 1, and is calculated as the average of ten variables capturing financial disclosures and governance disclosures as described in Panel A of Table 1.

Empirical Analysis 4.1. Model Specification
Following previous literature (e.g. (Daske et al. 2008), (Ernstberger et al. 2012); (Meser et al. 2015)), we perform panel ordinary least square (OLS) regressions with robust standard errors clustered by firm. We regress our three proxies for valuation effects (VALit), namely Tobin's Q (LN_TOBINit), the market-to-book ratio (LN_MTBit) and a valuation factor (VAL.FACit), on the yearly disclosure and enforcement score value and on various firm and country specific controls. We formulate the following equitation (1): VALit = β0 + β1DISCt + β2ENFOt + β3LN_SIZEit + β4LEVit + β5ROAit + β6AGit + β7GDPt + β7Dt + η (1) We compute Tobin's Q as the sum of total assets and the market value of equity minus the book value of equity, scaled by total assets. The market-to-book ratio is calculated as the yearly average market capitalization divided by the book value of common equity at the end of the fiscal year. In line with other studies, we apply principal component analysis and combine our variables into an additional valuation factor since a factor constitutes a more reliable metric to approximate security's valuation (see e.g. (Ernstberger et al. 2012) and (Christensen et al. 2013)).
Following our theoretical framework provided in section 2, enforcement regulation is expected to be positively associated with equity valuation. For mandatory disclosure regulation, we have identified various positive and negative individual effects in our framework. In an empirical setting, it is not possible to isolate and measure the amount of these individual effects of mandatory disclosure regulation on equity valuation. Only the total effect can be observed. With regard to our model, we expect varying country-specific total valuation effects depending on the proportion of good and bad news firm within the countries.
For our six country sample, we predict the predominance of firms withholding bad news for Japan and France. In Japan, informal communication mechanisms and low enforcement score values throughout our observation period may indicate more inherited burdens in the balance sheets of Japanese firms. Famous examples for inherited burdens are scandals around Toshiba and Olympus. Already (Leuz et al. 2003) provide empirical evidence that there is more earnings management in firms based in insider economies with low enforcement mechanisms. A tendency to earnings management can be interpreted as a signal that these firms also tend to withhold bad news. In the course of a higher level of mandatory disclosure regulation these bad news must be uncovered. As a consequence, we expect a negative coefficient for mandatory disclosure regulation in Japan.
Regarding France, the prediction of a proportion towards more bad news firms is reasoned by macroeconomic structural deficits. The need for structural reforms in France and the existence of powerful labor unions are problems on country level that are financially reflected in the balance sheets on firm level. For example, strikes in France led to the highest amount in Europe of non-work days per 1,000 employees throughout the observation period (Data from European Trade Union Institute; see e.g. (Carley 2010)). This is only one example for higher costs and decline in sales for French firms.
We select control variables in line with similar studies (see (Daske et al. 2008); (Meser et al. 2015)). In particular, to control for non-regulatory company effects, we include the following control variables: total assets (SIZEit), financial leverage (LEVit), return on assets (ROAit) and the asset growth rate (AGit). In addition, we incorporate the gross domestic change rate (in real terms; GDPt) and a dummy variable for the technical effects from the mandatory introduction of information accounting (Dt) as country specific controls. All other firm specific aspects are captured by firm fixed effects. A definition of all measurement and control variables is shown in Table  2.

Panel A: Measurement Variables
LN_TOBINit natural logarithm of total assets plus market value of equity as of seven months after the end of fiscal year t minus book value of common equity for firm i, scaled by total assets LN_MTBit natural logarithm of the average market capitalization in fiscal year t for firm i, divided by book value of common equity

Panel B: Control Variables
DISCt + / -country-specific disclosure score value in calendar year t as described in Table  2 -7; ENFOt + country-specific enforcement score value in calendar year t as described in Table 2  Dt + / -country-specific dummy variable that captures technical effects from the mandatory introduction of information accounting in calendar year t. For firm size we predict no sign. Small firms, for instance, will be more likely to have growth opportunities ((Lang et al. 2004), p. 612), and large firms are likely to be less risky and therefore have lower cost of capital ( (Botosan 1997), p. 341). Because of pecking order theory, we predict a positive sign for leverage ( (Frank & Goyal 2003)). For the return on assets and the asset growth rate we expect positive coefficients. To control for macroeconomic events and time effects, we include the gross domestic change rate, for which we expect a positive correlation ( (Levine & Zervos 1998)). We incorporate a dummy variable for the technical effects from the mandatory adoption of information accounting to our regression. This dummy catches the effect that information accounting does not only lead to a change in the information environment, but also affects the valuation of balance sheet items technically by recognizing hidden reserves in countries that previously emphasized prudence. Higher book values of equity and an increased balance sheet total affect our valuation metrics. We control for this effect in Germany and France. For the UK, we do not consider the information accounting effect in 2005 because of the similarity of UK GAAP to IFRS ( (Bae et al. 2008); (Brochet et al. 2013(Brochet et al. ), p. 1373.
We winsorize all variables at the extreme percentiles except for those with natural boundaries at both ends. According to previous empirical studies, we use the logarithm for our valuation metrics as well as for the firm size (e.g. (Daske et al. 2008); (Fu et al. 2012); (Meser et al. 2015)). As a result, by using logarithms we are able to capture non-linear relations better, to correct for heteroskedasticity and also to compress data to give less weight to outliers.

Sample Selection and Descriptive Statistics
As disclosure requirements do not equally affect all listed firms, we only consider firms that are listed on regulated domestic markets. Our initial data sample consists of all companies that have been listed on a regulated domestic market in one of our six sample countries. We treat the following stock exchanges or indices as regulated markets: NYSE, AMEX, NASDAQ and OTCBB in the U.S.; the TSE First and Second Sections in Japan; the FTSE All-Share in the UK; the CDAX in Germany; the CAC All-Share in France and the TSX Composite and TSX Venture Composite in Canada. As OTCBB has not been enforced by the SEC for the whole sample period (until 1999), we performed a robustness check without this market and arrived at similar results. We retrieve accounting and daily stock market data from Datastream/Worldscope for a maximum sample period from 1980 to 2010. Since data availability is extremely limited for all years before 1991, we constrain our sample for the main analysis to the 20year period of the fiscal years 1991 to 2010. We eliminate financial firms because they are subject to further special regulations. In addition, we delete observations with abbreviated financial years. We do not delete dead firms or firms that do not provide data for the entire observation period for survivorship bias reasons. Further, we only use observations that have data for each of our specifications. We do so to ensure comparability across our models. Especially for the equity markets of Canada and the UK, data availability notably constrains the sample. We end up with a cross-country sample of 92,211 firm-year observations.

Results
We proceed as follows. We first present a pooled cross-country regression. To examine the influence of the legal system on valuation effects with regard to (Armstrong et al. 2010), we will then divide our sample into two subsamples: a common law country and a code law country group. The U.S., UK and Canada are common law countries, while Japan, France and Germany are classified as code law countries (Nobes & Parker 2012). Third, we investigate all six countries separately to show that varying valuation effects are country-specific.
The three left-hand columns of Table 3 report results from regressing valuation metrics on disclosure and enforcement regulation and control variables for the pooled sample. Column 1 shows that our valuation model accounts for 59 % of the variation in Tobin's Q. Our second and third pooled valuation regressions, namely marketto-book ratio and the valuation factor, exhibit similar results, yielding an adjusted R 2 value of 60 % and 66 % respectively. We find a significantly negative coefficient for disclosure regulation in two of three regressions indicating that mandatory disclosure regulation is negatively associated with firm valuation. In terms of our theoretical model, the uncovering of bad news must, in the pooled sample, outweigh the positive effects. For enforcement regulation we find the predicted positive coefficient in all regressions. The signs of all control variables are as predicted. The respective magnitudes of the coefficients cannot be further interpreted or compared in the model setup. As an untabulated robustness check we also include country fixed effects into the pooled regression and arrive at similar results.

Table 3: Regression Results for Securities Market Regulation -Pooled Regression, Common Law and Code Law Group
This table reports coefficients and t values in parentheses of two-tailed tests (with significances at the 10 %, 5 % and 1 % levels indicated by *, ** and ***) of the following panel ordinary least square (OLS) regressions with firm fixed effects and robust standard errors clustered by firm: VALit = β0 + β1DISCt + β2ENFO + β3LN_SIZEit + β4LEVit + β5ROAit + β6AGit + β7GDPt + β7Dt + η Please refer to Table 2  With respect to the findings of (Armstrong et al. 2010) we next test the influence of the legal system. Breaking down the sample in common and code law groups and analyzing the six columns on the right-hand side of Table  10, we find again positive coefficients for enforcement in both groups and R 2 values ranging between 57 % and 67 %. The control variables again show the expected signs. The sample split now reveals a highly significant positive coefficient for mandatory disclosure in the common law group, while there are negative coefficients for the code law countries' sample. This is in line with the results of (Armstrong et al. 2010) and seems to indicate a dependence on the legal system. But the results have to be interpreted with caution. Within the groups, there is still substantial heterogeneity in terms of the distribution of observations. Findings could be dominated by the huge capital markets of the U.S. with 39,038 firm year observations and Japan with a number of 29,267 respectively. Furthermore, the group results contradict in parts the positive results on disclosure regulation in the code law country Germany (Meser et al. 2015). According to our model, we point out the proportion of good and bad news firms within a country as the deciding factor for varying country-specific effects, not the legal system. We therefore split the sample further down to the country level.  Table 4 shows regression results at the country level for the common law group. A very differentiated picture emerges. Our regressions for the U.S. in the columns 1 to 3 provide evidence for a positive relationship between both components of capital market regulation and equity valuation in line with most of the empirical studies. With one exception, the coefficients for disclosure and enforcement are significantly positive at a level of 1 %. Our model is able to explain between 56 % and 67 % of variation in the valuation measures. For the UK, our model is able to explain 61 % and 71 % of the variation in the valuation measurement variables. Except for the asset growth rate, all control variables show the expected signs and are in most cases highly significant. However, there is only one significant positive coefficient for disclosure and one significant negative coefficient for enforcement. All other coefficients for disclosure and enforcement are insignificant. We explain this result by the unusual variance in UK's regulatory environment: The UK is the country least affected by regulatory changes. For instance, the enforcement score increases only by 0.10 over the whole sample period of 20 years. This makes it difficult to find significant results for the UK, but also highlights the need for analyses at the country level. The last three columns report the results for Canada, which are again in the expected range. For Canada our model yields an adjusted R² value between 65 % and 72 %. The coefficients for enforcement are positive and significant at a level of 1 %. For disclosure we find one of three coefficients with a significantly positive sign at a level of 5 %. Except for ROAit and AGit all control variables have predicted signs and are mostly significant.

Table 4: Regression Results for Securities Market Regulation -Common Law Countries
This table reports coefficients and t values in parentheses of two-tailed tests (with significances at the 10 %, 5 % and 1 % levels indicated by *, ** and ***) of the following panel ordinary least square (OLS) regressions with firm fixed effects and robust standard errors clustered by firm:    Table 5 reports the results for the group of the code law countries. The three columns on the left show the results for Japan. The adjusted R² accounts for between 56 % and 65 % of the overall variation. All coefficients of control variables have the expected signs, and only one is insignificant. All three measurement variables show significant negative coefficients for disclosure and significant positive ones at a level of 1 % for enforcement. This indicates that a higher level of enforcement leads to positive capital market effects in Japan, while a higher level of mandatory disclosure regulation implies lower valuations. In the columns 4 to 6, we present results for Germany. The adjusted R² ranges between 52 % and 63 %. Except for ROAit, controls show expected signs and significance. The results for Germany document a situation similar to the U.S. and Canada in line with (Meser et al. 2015). The results show highly significant positive associations between the components of capital market regulation and the three valuation measurement variables. Higher levels of disclosure and enforcement leads to higher firm values. For France in the columns 7 to 9, we can present similar results to Japan. Our valuation regressions exhibit an adjusted R² between 48 % and 58 %. Controls are mostly significant and show the predicted signs. We find a significant negative relationship between disclosure regulation and the measurement variable Tobin's Q. The other variables are negative, but not significant to a level of 10 %. For enforcement regulation, we find a highly significant positive coefficient for Tobin's Q and a significant negative coefficient for the market-to-book ratio at the level of 10 %. Factor analysis reveals a significant positive effect.

Table 5: Regression Results for Securities Market Regulation -Code Law Countries
This table reports coefficients and t values in parentheses of two-tailed tests (with significances at the 10 %, 5 % and 1 % levels indicated by *, ** and ***) of the following panel ordinary least square (OLS) regressions with firm fixed effects and robust standard errors clustered by firm: Please refer to Table 2  Our results suggest that there are country-specific variations between mandatory disclosure regulation and equity valuation. We also document that it is not the legal system which is crucial in explaining differences. Country-specific characteristics have to be taken into account. For Canada, Germany and the U.S., the results indicate that a higher level of mandatory disclosure leads in total to a higher firm value. In France and Japan, the negative effects from the discovery of bad news outweigh the positive ones as predicted in section 4. The overall economic picture sees these two countries plagued by persistent structural problems. At the firm level, these problems were hidden in the balance sheets and were then gradually exposed when disclosure requirements increased. Examples such as Vivendi in France or Kanebo, Olympus and Toshiba in Japan are not only cases of individual problems but also indicate to some extent structural deficits in these countries. Consequently, investors use better disclosures to update their expected future cash flows and risk expectations.
While we provide evidence that mandatory disclosure can lead to negative capital market effects due to the reveal of bad news, the findings do not indicate that regulatory changes in Japan or France failed in strengthening the capital markets. For one, the positive effects from enforcement changes need to be taken into account. And expecting a consistently positive valuation effect from disclosure would be misplaced optimism. There are not always good news from additional mandatory disclosure regulation in the short run, but in the long run strengthening the information environment will always be good news for disclosure and for the investors that depend on them.

Conclusion
Information and reporting risks give rise to risk premiums demanded by equity investors. Extended disclosure requirements and stronger enforcement mechanisms address these risks, and capital markets have witnessed ever stricter regulation by means of new disclosure requirements or enforcement institutions. Still, there is empirical evidence of unfavorable outcomes from additional mandatory disclosure regulation. Based on a comprehensive model for disclosure, enforcement and equity valuation, we can show that enforcement regulation will lead to a higher firm value, but stricter disclosure requirements do not. More disclosure regulation reduces information risk, but may give rise to new concerns about the firm's activities and its reporting risk due to the reveal of bad news. The overall effect of increased disclosures may lead to a downwards revision of equity valuation. We point out that it is not the legal system which is crucial. Depending on the proportion of firms that immediately disclose good and bad news on a voluntarily basis to firms that withhold bad news within a country, we expect varying country-specific capital market effects. Our model is thus also able to explain negative consequences of welldesigned regulatory actions.
Using a cross-country dataset of 92,211 firm-year-observations, we provide empirical evidence on the effects of regulatory changes. We analyze the regulatory framework of six countries (Canada, France, Germany, Japan, the United Kingdom and the U.S.) between 1991 and 2010 and test for the long-run association between disclosure, enforcement and equity valuation. We examine valuation effects by Tobin's Q, the market-to-book ratio and a valuation factor. Consistent with our model, we find positive associations between our enforcement score value and our valuation measurement variables for all settings: a pooled regression, a split between legal system groups of common and code law countries and for each country separately. The effects of disclosure regulation offer spurious results in the pooled regressions and in the legal system split. Instead, specific country characteristics seem to determine the effects of mandatory disclosure regulation. In Canada, Germany and the U.S., the countries typically examined for effects of disclosure regulations, a higher level of disclosures is associated with a higher firm value. However, we find a negative association for France and Japan. We posit the predominance of firms that have to discover previously withheld bad news in these economies, because they are plagued by long-running macroeconomic problems. For the UK, we are not able to find significant results. We reason this by the small overall change in the regulatory environment for the period under investigation. Our paper faces some limitations that can be addressed in future studies. The extension to additional countries is obvious. A broadening of the sample could indicate whether the results hold beyond the six largest equity markets. A possible shortcoming is our focus on the country level, which obscures firm level effects. Investigating additional voluntary disclosure on firm level is another possible extension. Matching at the country level or splitting samples by firm characteristics may lead to further insights about the valuation effects of disclosure and enforcement regulation. Especially investigating into the links of accounting choices at firm level and structural difficulties of an economy will demonstrate how these individual decisions take an effect on a market. This could be done by associating accounting quality measures with the performance of industries or countries. Moreover, we do not know a lot about costs. There is a need for future research regarding how to quantify regulatory costs and to which extent they may cause adverse capital market effects respectively. Such an estimation needs to comprise several perspectives as, for instance, administrative costs, compliance and litigation at firm level and, finally, technical obstacles created for accountants, auditors, investors and market makers.  1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 1991 1992 1993 1994 1995 1996 1997 1998 1999     The SEC is granted rule-making powers in by various provisions in the Acts of 1933 and 1934 (e.g., SEA s. 13(a), SEA s. 14(a)(1), SEA s. 15(d), SA-US s. 19(a)). Throughout the sample period, the number of applicable regulations rules increased and i.a. regulates accounting and disclosure (to name the most prominent ones: Regulations S-K, S-X). Courts or other authorities may not amend or repeal these rules. o Members of the SEC "can be removed from office by the President only for good cause" during the entire observation period ( (Peters 1988), p. 286). Because of this, we rate variable E5=0. It is in contrary to (La Porta et al. 2006)) and their element Appointment, which is conceptual similar.

Japan
Although Japan has one of the largest equity and debt markets in the world, and substantial parts of its Securities Exchange Law (SEL) were transposed from the U.S. after the Second World War, its securities market regulation intensity falls behind the other countries in almost every year during our observation period. Three reasons come into play: The financing via large blockholders and banks (Markham 2002); an administration which for a long time tried to protect insiders in order to the conserve structures and foster growth of domestic companies (Pardieck 2001); and lastly a regulatory system based on consensual solutions instead of litigation and enforcement (Small 2003). This approach was also reflected in the institutional setting: initially, a department within the Ministry of Finance (MOF) was competent for market oversight, but with limited powers (Lu 1993). The Japanese banking crisis in the early 1990s changed the perspective on financial markets: bank recapitalization, supervision and securities regulation were transferred -in the so-called "Japanese Big Bang" -from the MOF to a unique and separate institution, the Financial Services Agency (FSA-JP) in 1998 ( (Hoshi & Ito 2004); (Markham 2002)). In the beginning (1998), the Supervisor was named Financial Supervisory Agency, and was not responsible for bank recapitalization. The shift to today's name and competences happened in 2000. Under its umbrella, the Securities Exchange and Surveillance Commission (SESC) is specially designated for enforcement. Despite these institutional rearrangements, the FSA-JP and SESC have remained not truly independent and are lacking rulemaking power. Apart from early changes (1989)(1990)(1991)(1992) in the major and insider ownership disclosure as well as in establishing a reactive review of accounting rules, the only increase in relevant regulation happened after the millennium: the requirement of quarterly reporting and independent auditors came into force (both in 2004), and the so called "J-SOX" was enacted in 2008 (Wisenbaker 2010). It transferred the most fundamental of the U.S. provisions to Japanese law, like increased management commentary disclosures and sworn statements. Even considering these recent changes, Japan can be characterized as having a "low regulation" framework. Table A2 shows and explains changes for Japan.  1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 93 1991 1992 1993 1994 1995 1996 1997 1998 1999       Until 1998, the MOF was competent in financial market supervision -until 1992, its Securities Bureau, later the (at that time dependent) Securities and Exchange Surveillance Commission (SESC) ( (Markham 2002), p. 63). In 1998, the FSA-JP (at first Financial Supervisory Agency, since 2000 Financial Services Agency) was founded as the Enforcement Institution, and under its umbrella the SESC remained responsible for enforcement. Even though the SESC is formally independent from the FSA-JP, it is not competent to process substantial cases (e.g., issuing the order to pay an administrative penalty) which has to be delegated to the FSA-JP ( (Small 2003), p. 339). For that reason, we assumed the FSA-JP as the Enforcement Institution. Its Commissioner is appointed by the Prime Minister, but the FSA-JP does not dispose of a board ( (Bebenroth et al. 2009), p. 188). For these reasons, we set E1 = 0 throughout the complete time period. l Weak provisions were in force regarding auditor independence before 2004. In that year, it was introduced that auditors may not perform non-audit services -which increases E2 to 0.5. Auditors must not be dependent, but a certificate of independence is not necessary (

United Kingdom
From an institutional perspective, regulation history falls into two phases. The first lasted until 1996 when the Department of Trade and Industry (DTI) and various Self-Regulating Organizations (SROs) like the London Stock Exchange (LSE) shared the capital market oversight. In the literature, this period is usually characterized as an intransparent regime with few statutory rules (Pimlott 1985). While mostly true for brokers/dealers over the entire period, changes happened for listed firms already in 1984. The enactment of the Companies Act (1985), implementing European directives on company law, and amendments to the LSE Listing Rules brought about a steep incline in disclosure and also in enforcement (Lorenz 1987). Even though the 1986 "Big Bang" in the UK meant a new Financial Services Act and a framework of stronger SRO regulation, it did not directly affect accountancy and equity issuers ( (Hablutzel 1992); (High & Shah 1993)). Only in 1989/1990, major steps were taken with the set up of a reactive enforcement of accounting rules by the Financial Reporting Review Panel (FRRP), and disclosure rules were strengthened (Fearnley & Hines 2003). The second phase started in 1997 with gradually delegating powers to the Financial Services Authority (FSA-UK). The competences to oversee the capital market was gradually transferred to the Securities and Investment Board (SIB) starting in 1997 which was renamed FSA-UK later. Full statutory powers were only granted with the enactment of the Financial Services and Markets Act (FSMA) which came into force in 2001. It is a separated enforcement agency and assumed competences e.g. from the LSE with broad rule-making (Listing Rules, Disclosure and Transparency Rules), investigative and sanctioning powers (against firms and managers). The FRRP is still competent for most accounting rule enforcement, and also gained pro-active powers in 2005. In general, the UK can be termed as a "high regulation" regime. The changes in disclosure and enforcement values are reported in Table A3.  1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 1991 1992 1993 1994 1995 1996 1997 1998 1999    Initially, a list tabulating the directors' shareholdings had to be made accessible at the firm's office and at its general meeting (CA, 1948, s. 195 (7)). After that, insider ownership had to be reported to the firm and to the LSE to release this information (CA, 1985, ss. 324 & 329  Investors owning more than 5 % of the shares have had to issue a report to the company and a disclosure via the CAO (Goergen & Renneboog 1998), p. 7;CA, 1985, s. 201, 202) since 1985. Later, the LSE LR (2002, and the DTR (2010) ss. 5.1.2 R & 5.9.1 R require this or a stronger disclosure. j Directors' transactions have to be reported to the firm and to the LSE to release this information (CA, 1985, ss. 324 & 329). Listing Rules also require the disclosure of related parties (e.g., LSE LR (2002) s. 11.1). Since 2005, this requirement is part of the DTR (2005)  Until 1996, the Department of Trade and Industry (DTI) fulfilled the functions of securities market surveillance without operating a separate Enforcement Institution. In parallel, the Securities and Investment Board (SIB) was founded in 1987 with a board appointed by the DTI ( (Hablutzel 1992)). It only had competences to supervise Self-Regulating Organizations (SROs) but not listed firms, which is why we do not label it as the Enforcement Institution. Only in 1997, these powers were transferred to the SIB that was renamed FSA of the UK. Its chairman and governing body are appointed by the Treasury (FSMA, 2000, Sched. 1 s. 2 (3)). Consequently, E1 always takes the value of 0. The DTI was empowered to require the production of any document ((Sikka & Willmott 1995); CA, 1985, ss. 432, 434 & 447), and equally is the FSA-UK (FSMA, 2000, ss. 165, 172). q A continuous filing requirement exists with the Companies House, a DTI agency, which is not empowered to review the documents ( (Fearnley & Hines 2003)). This power was given to the Financial Reporting Review Panel (FRRP) which was set up in 1990 as a reactive enforcement body. Its competences were increased to a pro-active review by means of the Companies (Audit, Investigations and Community Enterprise) Act, 2004, ss. 10-15. r Companies could be brought to court or fined by the DTI (CA, 1985, ss. 61 (2) & 438). The FSA-UK may e.g. resort on adverse publicity, impose financial penalties, suspend the listing, or bring a firm to court ( (Fearnley & Hines 2003);FSMA, 2000, ss. 77, 91, 93 & 206;CA, 2006, s. 456). s Managers could be fined or held criminally liable under the DTI regime (CA, 1948, ss. 147 (4) Even though the directors had to sign financial reports in the years of this observation period, the Companies Act did not explicitly impose a liability (CA, 1948, s. 169). This was introduced in 1985 and exists since this date (CA, 1985, ss. 238 & 245;CA, 2006, s. 414).

Germany
Capital market regulation in Germany used to be weak since German financing patterns traditionally relied on bank loans and large blockholder financing (Ball et al. 2000). Installing a powerful enforcement regime was not in the focus of regulatory initiatives. Only the effort of the European Union to foster high quality financial reporting and converging international capital markets caused a shift in securities market regulation towards investor protection. Since the mid 1990s, Germany made great efforts in enhancing disclosure requirements and enforcement. The adoption of IFRS for consolidated accounts and a most recent initiative with the Accounting Modernization Act (Bilanzrechtsmodernisierungsgesetz, BilMoG) of 2009 constitute major milestones in disclosure rules for German listed companies. Those advances in company disclosure were accompanied by changes in the enforcement structure, finding its peak in the foundation of the first single capital market regulator in Germany, the Federal Financial Supervisory Authority (FFSA; Bundesanstalt für Finanzdienstleistungsaufsicht, BaFin). In corporation with the Financial Reporting Enforcement Panel (FREP; Deutsche Prüfstelle für Rechnungslegung, DPR) which was established in 2004, Germany operates a two-tier control system for capital market oversight. Given the remarkable development within the last 15 years, securities market regulation still remains on a low level suggesting that disclosure regulation is not fully geared towards investor orientation compared to highly regulated regimes such as the U.S. Given these circumstances, Germany can be seen by as a "low regulation" model. All changes of disclosure and enforcement changes are shown in Table A4. Table A4: Score Values by Year, Germany This table shows the yearly disclosure and enforcement regulation score values and provide explanations for the changes in Germany for the period between 1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 85 1991 1992 1993 1994 1995 1996 1997 1998 1999    According to §1 of the Selling Prospectus Act (Verkaufsprospektgesetz) companies that offered shares to the public had to deliver prospectus to potential investors, starting from January 1991. In 2005, this provision was replaced by the Securities Prospectus Act (SPA -Wertpapierprospektgesetz) which pronounces the prospectus requirement in §3. Controls on the formal correctness of prospectus were only established in 1998 (SPA §8a). These prospectus disclosure requirements have always included various exemptions (SPA §4; SPA §4). i STA §21 (1) and STA §26 (1) oblige publicly listed firms to disclose shareholdings if they exceed thresholds of 3,5,10,15,20,25,30,50 or 75 %, starting from 1995. The rules do not clearly indicate whether the thresholds include indirect ownerships. We therefore allocate a score value of 0.5 to the item. Later, we augment the value to 1 since CC-DE §315(4)3 coming into force in July 2006, states that direct and indirect ownerships have to be disclosed in the management commentary if they exceed 10 %. j Information on share transactions of board members (directors' dealings) or related parties has to be published starting from July 2002. This requirement is settled in STA §15a. The initiation of the Transparency Realization Act (Transparenzrichtlinie-Umsetzungsgesetz), which came into force in January 2007, extended these requirements (i.e., information has to be provided to the official business register). k In 1995, the Federal Securities Supervisory Office (FSSO -Bundesaufsichtsamt für den Wertpapierhandel) commenced its operations as a financial market supervisory authority. The government used to propose a candidate for the presidency of the FSSO to the Federal President of Germany. The latter then appointed the FSSO president (STA §3 (2) STA §37p). Hence, the FREP can only require documents from companies that agree to the examination. If a company refuses, the case will be handed to the FFSA that disposes -since then -of the legal power to order withdrawal of documents from any person of a firm under investigation (STA §4 (3) & §37o (4)-(5)). q Since 1995 the FSSO has the competences to reactively review financial reports (FSSO, 1995, s. 29). In 2005, a systematic review of filing documents was established. In this year, the FREP was set up and the FSSA gained the competences to proactively review financial reports (CC-DE §342b(2); STA §37o(1)). The separation into a voluntary and compulsory stage as outlined in footnote p also extends to the review. r Along with its foundation in 1995, the FSSO was given the power to impose financial sanctions. Its successor, the FFSA, assumed these competencies. If the FFSA suspects a criminal act, it passes the case to the federal prosecutor (STA §4 (5)). In the area of banking supervision, the FFSA can impose further sanctions, like the cancellation of the banking license ( §33 of the Bank Credit Act (Kreditwesengesetz)). However, banking supervision is not in the focus of our study. We therefore keep a constant value of 0.5 throughout the time series.
s The disclosure of misleading information has been punished by fine or imprisonment throughout the entire observation period. In 1986 moved to §264a of the Criminal Code (see also  With the enactment of the Transparency Realization Act in January 2007, board members and directors have been required to affirm the material accuracy of financial statements. The requirement was settled in CC-DE §264 (2).

France
Along with the creation of the French Securities and Exchange Commission (COB -Commission des Opérations de Bourse) in 1967, France was one of the first countries in Europe to establish an influential enforcement body following the example of the SEC. Acting as a stock market regulator, the COB was given authority to set accounting guidance statements and to impose sanctions in case of non-compliance (Brown & Tarca 2005). In contrast to a powerful enforcement structure, disclosure regulation first played a minor role in capital market regulation; company disclosure in France was rather shaped by tax-related issues than by motives to provide outside investors with useful information ( (Giner & Rees 2001); (La Porta et al. 1997);). Major changes in disclosure have only been promoted during the years between 2001 and 2010, coming along with the overall aim to improve financial reporting quality in the European Union. Those changes are primarily related to the mandatory adoption of IFRS and associated increased disclosure requirements (e.g. in-depth segment reporting and management commentary). Disclosure regulation only caught up to the U.S. after the millennium. The traditionally high enforcement level was further increased in 2003 when the COB and two further enforcement bodies were merged into the Financial Markets Authority (AMF -Autorité des Marchés Financiers). It acts as the single capital market regulator and is endowed with additional power for market interventions. Considering all aspects of disclosure and enforcement, France is a "high regulation" system. Table A5 reports all changes in disclosure and enforcement score values.  This table shows the yearly disclosure and enforcement regulation score values and provide explanations for the  changes in France for the period between 1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 85 1991 1992 1993 1994 1995 1996 1997 1998 1999 (Cervellati & Fioriti 2007)). o No explicit rules exist in the legal bases establishing the COB or the AMF regarding the tenure of its commission members. For that reason, we set the variable to 0 throughout the whole observation period. p The COB was empowered since 1989, and the AMF has ever been empowered for investigation purposes to demand information from any person that has the ability to deliver information (Act No. 89-531 Art. 5ter; MFC Art. L621-12). q The obligation to deliver periodic information existed prior to 1991 (Decree No. 67-236 Art. 296). Both the COB had and the AMF has the power to proactively review financial documents and demand the correction of accounts ( (Lascoumes 1985), p. 8); (Hong Phu Dao 2005), p. 112). r The COB was and AMF is able to impose sanctions on misconducting companies. These sanctions include warnings, trade suspensions and financial penalties (Act No. 96-597 Art. 71; MFC Art. L621-15). In case of criminal infringement and market manipulation, the AMF directs the reports to the public prosecutor. The predecessor of the AMF, the Conseil de Discipline de la Gestion Financière (CDF) was able to use the same sanctions ((Hong Phu Dao 2005), p. 112). Therefore we set the variable to 1 over the entire observation period. s In line with Ordinance No. 67-833 Art. 10-1 the COB could and in line with MFC Art. L621-9 & L621-15, the AMF can likewise sanction individuals that act on behalf of entities (see footnote r). There is no indication that those individuals can be sanctioned even when they are unaware of fraudulent actions. t Management has to certify the correctness of financial reports in line with COB Regulation No. 2002-05 Art. 2 ter-1 (as endorsed by the Order of 18 June 2002). This requirement has been in place also under the AMF regime, currently set out in its General Rules Art. 212-14.

Canada
Capital market regulation in Canada is a special case since it is the only country within our sample with few institutions and rules at the national (i.e., federal) level. Each province or territory may issue own securities laws and empower own enforcement agencies, whereas banking supervision as well as accounting rules and practice have developed nationally . Most listed firms are traded at the Toronto Stock Exchange (TSX); hence for the past 30 years the Ontario Securities Act (SA-ON) is relevant where no federal regulation exists, and the Ontario Securities Commission (OSC) enforces this regulation. Generally, all aspects are more or less closely linked to U.S. provisions as a considerable share of TSX-listed firms is cross-listed on one of the New York exchanges, which ties them to U.S. rules . With respect to disclosure, both the Canadian and the U.S. regimes differed only in few items during our observation period. The relevant enforcement only caught up in 1994, and more prominently, in 2003/2004 when several regulations were introduced or amended that mirror the SOX (Carnaghan & Gunz 2007). To this end, the Ontario administration strengthened the OSC (giving the power to impose administrative financial penalties, conducting a pro-active review of filing documents) and requires auditors to certify their independence. Other provincial regulations were superseded (e.g., remuneration disclosure, management commentary requirements) or accompanied by federal ones (directors' certification of financial reports) (Kuras 2004). In stylized terms, the Ontario/Canada case represents a "high regulation" regime. Table A6 describes disclosure and enforcement score values for Canada.  This table shows the yearly disclosure and enforcement regulation score values and provide explanations for the  changes in Canada for the period between 1991 and 2010. 1991 1992 1993 1994 1995 1996 1997 1998 1999 1991 1992 1993 1994 1995 1996 1997 1998 1999 (Taylor 1978)) finds that the tenure of commission members is unclear. Hence, we code E5 = 0. p Ontario: As regulated by the Ontario Securities Act, the OSC can seize any relevant document (SA- ON, 1980, s. 11 (4); current unchanged rule: SA- ON, 1994, s. 13 (1)). q Until 2003, the OSC filings were used to reactively enforce financial reports. From this year onwards, the continuous disclosure review was introduced which includes a pro-active enforcement of accounting rules (SA- ON, 2003, s. 20.1). This provision mirrors the SOX s. 408 in U.S. law. r Ontario: In 2003, the OSC was empowered to impose administrative fines. Since then both, financial and legal penalties can be issued against companies (SA- ON, 2003, s. 127 (1)). s Ontario: Management can be sanctioned with fines or imprisonment if the person was aware of committing an offense (SA- ON, 1980, s. 118;current unchanged rule: SA-ON, 1994, s. 122). Until 1994, the Minister had to approve any sanction (SA-ON, 1994, s. 119 was repealed in this year); consequently, up to this year, item E9 is set to 0.