Analysis of Companies Financial Information Quality

This paper aims to analyze financial information quality through financial and economic ratios, detecting whether they are affected by financial reporting standards; as well as, determining a group of factors that allow proving the capacity of ratios to measure accounting information quality, and thus, facilitate the analysis process to the groups of users. Design / methodology / approach - For a sample of 111 companies from Madrid Stock Exchange and 32 from Eurostoxx50, descriptive analysis and non-parametric variance analysis were carried out during the period 2005-2007. At the same time, reduction data techniques were performed to detect the underlying main factors, specifically the principal component analysis (PCA) for the year 2007. Findings - It has been confirmed that financial information quality is affected by financial reporting standards; additionally, a group of factors in which financial and economic ratios group has been found. Practical implications - This study provides evidences to measure financial information quality and the results can be useful to accounting users, as well as, contributing to literature related to this topic.

Originality and value - This study empirically shows, from the analyzed companies, that accounting information is affected by financial reporting standards. Additionally, some factors that group ratios are provided. Keywords Financial information quality, Ratios, Financial reporting standards, Accounting users, Kruskal-wallis Test, Principal component analysis. Type of document Research paper Introduction Financial information quality has been empirically approached since late 60s. The studies were focused on contrasting whether data provided by companies were really useful to accounting users. At first, some studies were concerned with investigating the information content (Ball and Brown, 1968).Then, at late 80s, a new orientation arose focusing on studying the relevance of information, using wider regression models to determine the relationship between financial information and market profitability (Ou and Penman, 1989a). It has also been observed that recent studies have focused on contrasting financial information quality in time. The models to show companies financial information have gone through important changes, specially, with the implementation of International Financial Reporting Standards (IFRS) to elaborate consolidated financial statements of companies listing in the stock exchange (Kenny and Larson, 2009). Among the harmonization changes in accounting, the measures taken on the transparency of information disclosure, a decisive factor, for the generation of quality accounting documents (Epstein, 2009) are distinguished. Therefore, financial information quality plays a fundamental role to know the companyA´s economic reality.

Furthermore, the used financial reporting standards can have an effect on the quality of disclosed data; and, therefore, on the formulation of appropriate Financial analysis of financial statements, in which there can be decisive factors in the evolution and tendency of some indicators (Choi et al., 1983). It is important to notice, that ratios can be a reflection of financial information quality, from the financial reporting standards point of view, used to help understanding financial and economic events that affect companies. Within this context, the purpose of this empiric study is to analyze financial information quality through financial and economic ratios, specifically focusing on detecting whether ratios are affected by the financial reporting standards applied by companies. At the same time, this study aims to determine a group of factors that allow demonstrating the capacity of ratios to measure accounting information quality, as well as, facilitating the financial information analysis process to the groups of users. Therefore, the study included the analysis of financial information from a sample of companies belonging to Madrid Stock Exchange and Eurostoxx50 during the period 2005-2007. This has lead to consider the financial reports presented with the Local Generally Accepted Accounting Principles (GAAP) and International Financial Reporting Standards (IFRS), as well as, individual and consolidated financial statements. The results obtained in this study reveal that some ratios show population averages differences, indicating that financial information quality is affected by the applied financial reporting standards. Additionally, a group of factors in which ratios group have been found, and they are, a way to measure the quality of companies financial and economic results by accounting users groups.

Background and empiric predictions

Financial information quality Financial information quality has been discussed in several studies, pointing out different aspects that lead to create new research lines. At late 60s, the accounting research took a new path, thanks to the utilitarian paradigm, leading to adopt a research methodology with a positivist approach that considers accounting as an information system. Consequently, the capital markets research line arose, this line researches accounting information performance for investors. The studies carried out by Ball and Brown (1968) and Beaver (1968) were pioneers. As a matter of fact, considering this research line, two work fields can be distinguished, on one hand, the research conducted to study information (information content and relevance), and on the other hand, the research that studies measuring. Regarding the studies on information, it can be mentioned that one of the examined aspects is related to the study of information content. That is to say, the market reaction is investigated in relation to the publication of earnings announcements. Some studies have indicated that market prices have a positive reaction in relation to earnings announcements (Weston, 1971).

Therefore, Bae et al. (2008) working with a sample of Korean companies, found that the increments (decreases) of earnings announcements have a positive (negative) effect on the value of shares prices in the market. Meanwhile, other studies have pointed out the opposite (Su, 2003). It should be mentioned that some studies have considered the relationship between the negotiation volume and earnings announcements. Therefore, Choi and Choe (1998) stated that the volume of operations reacts when three-month earnings announcements are made. In this sense, some factors that will affect the negotiation volume, such as investors' agreement degree (Garfinkel and Sokobin, 2006) or information asymmetry (Mahipala et al, 2009) can be distinguished. Some researchers have considered the effects of earnings announcements in both, prices and volume of operations. Cheung and Sami (2000) analyzing a sample of companies from Hong Kong Stock exchange, and, the earnings announcements corresponding to the period 1992-1995, showed a significant reaction of prices and volume of operations when announcements were made.

On the other hand, Ahmed et al. (2003) analyzing the negotiations during the period 1992-1995 (without on-line trade) and the negotiations during the period 1996-1999 (with on-line trade), demonstrated that three-month earnings announcements produce changes in shares prices and volume of operations when the negotiations are on-line. Other research papers have studied the delayed reactions in the market when the announcements (post-earnings announcement drift. PEAD) are made, due to problems related to earnings predictions. According to Bernard and Thomas (1989) PEAD incorrect measures can be grouped in "false risk measures" and "other incorrect measures". In this sense, Zhang (2008) points out that the analysts' prediction sensibility is associated with the costs and benefits of predictions. It is necessary to mention the considerations made in some studies about the reactions that accounting changes produce in the market, and, therefore, end up affecting prices. It should be estimated that the results of some studies have indicated that accounting changes do not have any effect on the market (Vigeland, 1981). Meanwhile, other studies have evidenced, for example, that the adoption of International Financial Reporting Standards generates changes in the market (Floros, 2007). The results of some studies have emphasized the relationship between the investor's performance and changes in market prices. Caspar (2003) analyzing investors' performance when presenting information, carried out by Danish companies, showed the presence of significantly abnormal results few days after the presentation.

On the other hand, Chewning et al. (2004) evidenced that sophisticated and not sophisticated investors have improved their abilities in the use of information. Therefore, the subsequent effects of their performances have been evident on market prices. Regarding studies of relevance, research papers on the anticipation of prices have been found, emphasizing time extension for profitability calculation. In this sense, Kothari and Sloan (1992) determined the expectations of future earnings, using price and earnings regression, corresponding to the multi-year prior period. On the other hand, Schleicher et al. (2007) worked with a two-year period in their regression model to estimate earnings. In relation to the studies of relevance and variables of results, some research papers have emphasized the relationship of profitability in view of changes and different levels of results. In this sense, Shroff (2002) shows the relationship of profitability with the levels of results. Meanwhile, other studies have determined the relationship in both, changes and levels of results (Dumontier and Labelle, 1998). Regarding prediction studies, Ou and Penman (1989b) evidenced that the model based on accounting information was useful to predict earnings. Other researchers have also been concerned with this aspect (Skogsvik, 2008). It can be observed that profits predictions have an explanatory power for future earnings (Wu and Wang, 2000); so the accounting structure becomes a decisive factor for the carried out estimates (Fairfield et al., 1996). The ratios, factor analysis and principal component analysis The use of data multivariate analysis has allowed explaining accounting phenomena through the treatment of groups of variables.

The availability of companies' great quantity of data, due to financial and economic ratios disclosure has facilitated the use of this statistical tool. Some studies, using multivariate techniques have tried to determine a group of factors that synthesize economic and financial indicators, in conglomerate groups that facilitate, among other aspects, the measuring of companies' information quality. Consequently, the most used multivariate techniques are the factor analysis (FA) and principal component analysis (PCA). Therefore, different studies have chosen to apply the factor analysis technique. Pinches et al. (1973) working with a sample of 180 industrial companies whose data were obtained from Compustat database, and using the varimax rotation, determined 7 factors (return on investment, capital intensiveness, inventory intensiveness, financial leverage, receivables intensiveness, short-term liquidity and cash position). In another study, Pinches et al. (1975) using the same sample of companies and ratios from the previous study, and applying the oblique rotation method, found the same factors. Short (1980) carried out a study selecting 259 industrial and commercial companies, considering data adjusted to inflation rate. Therefore, using varimax rotation, 7 factors of historical ratios were found (return, capital intensiveness, asset turnover, financing policy, inventory turnover, working capital and current position), and, the same number of factors of price-level ratios (return, capital intensiveness, asset turnover, financing policy, inventory turnover, working capital and cash position). The previous ideas are consistent with those stated by Yli-olli and Virtanen (1989) and Liao (2008).

Other studies have preferred to apply the principal component analysis. Stevens (1973), selecting a sample of 80 industrial companies, reduced 20 ratios to 6 factors (leverage, profitability, liquidity, activity, dividend policy and price and earnings). Pinches and Mingo (1973) examining 180 industrial companies with the varimax rotation method, determined 7 factors in which the indicators (size, financial leverage, long-term capital intensiveness, return on investment, short-term capital intensiveness, earnings stability and debt coverage stability) are grouped. Laurent (1979) selected a sample of 63 companies that were analyzed with the varimax rotation method, 10 components (return on investment, leverage, working capital, fixed assets, long-term solvency, liquidity, cash position, stability of debts coverage, self-financing policies and credit policies) were proposed.

These ideas are consistent with Chen and Shimerda (1981), Sorensen (2000) studies. Hypothesis formulation The measuring of ratios began to be developed in the United States to carry out financial analysis, acquiring relevance in the financial sector, specifically in the analysis of credit risk and management. In the course of time, studies on ratios-type use and ratios groups modeling arose. Therefore, considering the idea that a ratio is meaningless by itself, and, so that its interpretation is valid, it is compared with an internal standard (based on true data conveniently settled), or with an external standard (corresponding to other organizations in a similar situation). From the 20s on, ratios are used in inter-companies comparative studies, to overcome the difficulties of carrying out comparisons in the technical-industrial field. As a consequence, some patterns arose allowing characterizing the differences in corporations' organizational structure. Therefore, the ratio-type employment has made possible the analysis of groups of companies in similar situations with certain characteristics (Mulla, 2003). FitzPatrick (1932) carried out one of the first studies that lead to develop the descriptive stage of financial analysis through ratios. The study consisted on three documents, in which the use of ratios to determine companies' bankruptcy was proposed.

Then, the predictive stage was started, focusing on the analysis of ratios prediction capacity. Beaver (1968) was one of the pioneers. He emphasized the analysis of ratios through statistical advanced methods to prove their predictive capacity. He also pointed out that ratios are accounting data that can be evaluated due to their usefulness that is defined in terms of their predictive capacity. Additionally, it is necessary to mention that the evolution of the international harmonization accounting process has been estimated during some time. This has been highly related to economic and institutional events that affect financial information (Kenny and Larson, 2009). In this sense, reforms in the juridical international regulations due to the adoption of IFRS has supposed, on one hand, benefits for companies (Epstein, 2009); and on the other hand, changes in the presentation of information. However, the existence of one accounting system has brought some problems related to accounting, juridical and economic aspects (Slot and Gerrits, 2009); and for that reason, in some cases, it has implied IFRS adaptations to each countryA´s social reality to prepare and communicate financial information (Tsakumis et al., 2009). Considering the previous ideas, some studies have discussed the implicit changes in the formulation of financial statements with IFRS application and their implications in the quality of accounting information. Therefore, Choi et al. (1983) developed one of the first studies approaching this topic, comparing financial reports elaborated with the American GAAP and the reports elaborated with the Japanese and Korean GAAP.

The results prove that ratios show significant differences, due to different factors related to regulatory aspects in each country. Jun and Wang (2001) showed some significant discrepancies in the disclosed data and in the main financial ratios examining the financial information of three Chinese companies, elaborated with China GAAP, Hong Kong GAAP or using IFRS, during the period 1995-1998. Meanwhile, Duangploy and Gray (2005) when comparing Japanese and American companies financial reports showed that the financial reporting standards, applied in each country, affect ratios. Agca and Aktas (2007), when comparing the obtained ratios with financial reports elaborated with IFRS and Turkish GAAP, found significant differences. Considering ratios empiric antecedents and studies that support information quality, considered from the financial reporting standards application changes, the hypotheses are formulated as follows: Hypothesis 1 (H1): the values of profits ratios and financial information quality are affected by the financial reporting standards used by companies. Hypothesis 2 (H2): the values of operational ratios and financial information quality are affected by the financial reporting standards used by companies. Hypothesis 3 (H3): the values of structure ratios and financial information quality are affected by the financial reporting standards used by companies. Hypothesis 4 (H4): the values of per employee ratios and financial information quality are affected by the financial reporting standards used by companies.

Research design and Sample selection

Study design The financial information of the selected companies from Madrid Stock Exchange and Eurostoxx50 was examined at an initial stage. Data were obtained from Amadeus Database. At first, the companies financial statements from the period (2005-2007) were reviewed; and, considering the characteristics of the obtained data, the financial information was classified into four groups: a) individual financial statements elaborated with the local Generally Accepted Accounting Principles (IFS-GAAP), b) consolidated financial statements elaborated with the local Generally Accepted Accounting Principles (CFS-GAAP), c) individual financial statements elaborated with International Financial Reporting Standards (IFS-IFRS) and d) consolidated financial statements elaborated with International Financial Reporting Standards (CFS-IFRS). Furthermore, data were grouped according to the economic sector of the companies' activity area, aspect that will be useful when making statistical contrasts. In the case of Madrid Stock Exchange companies, the separation issued on January, 1, 2005 for all companies admitted to list in the Spanish Stock Exchange was selected.

In the case of Eurostoxx50, the companies were treated as an additional sector to have a general idea of their performance. It is necessary to point out that there were companies belonging to both Markets (Repsol YPF, Endesa, Iberdrola, Arcelor, Bayer and Telefónica) when the study was carried out. At a second stage, ratios that will help to examine financial information quality were selected. It is evident that the indicators obtained from accounting data are numerous. That is why, it is necessary that the selection must be as relevant as possible. Therefore, a revision and a comparison with ratios selected in other studies (Stevens, 1973 and Pinche and Mingo, 1973) have been made. It was chosen to use the ratios presented in Amadeus database, considering that most of them are similar with those used in other mentioned studies; additionally, due to their revealed data reliability and consistency, avoiding them to be significantly different from one period to another. In table 1, the relationship of the 26 selected ratios and the calculation form are shown.

Total assets per employee (monetary value) Total assets/number of employees Methods In relation to the kruskal-wallis test, it can be noticed that it is a non-parametric test equivalent to ANOVA, used to compare population medians in case data are not normally distributed. The hypotheses are defined by: H0: The k medians of variables are all the same. H1: At least one median from one of the variables is different. Therefore, the observations of k random samples are placed in growing order of magnitude, assigning ranks to each one of the variables to determine the ranks for each sample. Regarding the statistic to test, this follows a distribution with k-1 free grades. In relation to the principal component analysis (PCA), it is a multivariate statistical technique to reduce the number of variables to a smaller number, losing as less information as possible (Foguet, 1989). At the same time, it is able to identify a group of fictitious variables formed from the combination of those previously observed. Therefore, it is possible to synthesize data and to relate them without making any previous hypothesis about the meaning of each initial factor.

The new principal components will form a linear combination of origin variables and each one of them is independent. It should be added that the kruskal-wallis test and the principal component analysis were run using the statistical program SPSS (Statistical Package for Social Sciences), version 15.0. Sample The population subject of study is formed by companies listing in Madrid Stock Exchange and Eurostoxx50. The sample submitted to observations is constituted by all productive sectors, excluding the financial and real estate sector, because it is regulated by specific standards. Therefore, companies in the sample have increased to 111 belonging to Madrid Stock Exchange and 32 to Eurostoxx50. In table 2, it is proved that most companies presented consolidated financial statements with the International Financial Reporting Standards (CFS-IFRS), specifically 75,52%, while 9,79%, (CFS-GAAP), 6,29%(IFS-GAAP) and 8,39% (IFS-IFRS). It is also important to mention that in the market group structure, it is observed that 73,4% belong to Madrid Stock Exchange and 22,4% to Eurostoxx50.

Descriptive analysis and hypothesis verification

Group 1. Profitability Ratios When examining the results, it is observed that mean values of variables are heterogeneous, even, negative values are shown for the IFS-GAAP subgroup, similarly occurred to the variables: return on shareholders funds (X1/2005: -2,492 and X1/2007: -26,082), return on capital employed (X2/2005: -69,791 and X2/2007: -11,267) and return on total assets (X3/2005: -13,015 and X3/2007: -3,883). Additionally, it is necessary to mention that the variable gross margin (X6) from this subgroup does not have available data in any of the analyzed years. In relation to the results of the descriptive statistics, it is proved that there are distributions far from normality, with high values of asymmetry and kurtosis, as it is the case of the CFS-GAAP group from the variables: return on shareholders funds (X1/2006: -3,092 and 10,812), profit margin (X5/2006: -3,197 and 11,456) and EBIT margin (X7/2006: -2,864 and 9,150). The kruskal-wallis population equality test has been considered to contrast the hypotheses related to the relationship of each one of the defined ratios and the group type to show financial information. The hypothesis to be proved is that the mean values of the considered ratios do not show differences, according to the financial reporting standards applied to present financial statements. Regarding H1 contrast results of mean ranks for this period, it is deduced that there are differences in population averages (p < 0,05) in the financial presentation for the variables: Return on shareholders funds (X1/2007: 0,021), Return on capital employed (X2/2005: 0,007), and Return on total assets (X3/2005: 0,012).

While in the other analyzed ratios, there are not significant differences. Therefore, the differences found in some ratios population averages prove that financial information quality can be affected by the used financial reporting standards. Group 2. Operational Ratios In relation to the descriptive statistics of operational ratios during the period 2005-2007, it was observed that dispersion is really high. This proves that mean values vary in the studied variables. There are also distributions far from normality, with high values of asymmetry and kurtosis, as in the group (CFS-IFRS) with the variables: stock turnover (X11/2005: 4,631 and 26,072; X11/2006: 5,568 and 37,406 and X11/2007: 4,890 and 28,913), net assets turnover (X9/2005: 4,444 and 30,605), interest cover (X10/2005: 5,249 and 35,238) and collection period (X12/2006: 2,868 and 15,622). The H2 contrast proves that there are significant differences in population averages (p < 0,05) in the financial information presentation for the variables: collection period (X12/2005: 0,029) and credit period (X13/2005: 0,001 and X13/2006: 0,002 and X13/2007:0); while the rest of variables do not show significant differences. Therefore, from the obtained results, it can be deduced that financial reporting standards can influence financial information quality. Group 3. Structure Ratios The exam of the descriptive analysis results, first of all, indicates that mean values vary in the studied variables.

Distributions far from normality are also proved, such as in the (CFS-IFRS) group, the asymmetry of the variable current ratio (X15) in the years 2005 (2,163) and 2007 (6,831). Similarly occurred to the kurtosis of the same variable X15 in the years 2005 (8,768) and 2007 (54,570). The H3 contrast results allowed to determine that there are differences in population averages (p < 0,05) in the financial information presentation of current ratio (X15/2005: 0,001, X15/2006: 0 and X15/2007: 0,003), liquidity ratio (X16/2005: 0, X16/2006: 0 and X16/2007: 0,001), shareholders liquidity ratio (X17/2005: 0,004) and gearing (X19/2005: 0,001); meanwhile, there are not significant differences in other years. Therefore, this result allows verifying that financial information quality is affected by the existence of differences in population averages in some of the analyzed variables. Group 4. Per employee ratios The descriptive statistics for this group of ratios reveals that mean values vary in the studied variables, and, distributions far from normality are observed. Additionally, high values of asymmetry and kurtosis are detected in some variables. This can be observed, for example, in the asymmetries of (CFS-IFRS) group from the profit per employee (X20/2005: 3,256, X20/2006: 2,788 and X20/2007: 7,972), operating revenue per employee (X21/2005: 2,865, X21/2006: 2,517 and X21/2007: 7,460), and, shareholders funds per employee (X24/2005: 2,306, X24/2006: 2,092 and X24/2007: 5,035). H4 is confirmed and it is evident that there are differences in population averages (p < 0,05) in the presentation of financial information, specifically for the variables: average cost per employee (X23/2005: 0,018 and X23/2006: 0,026) and total assets per employee (X26/2005: 0,026). Furthermore, it is proved that the other variables do not show significant differences. Therefore, the results allow concluding that financial information quality is affected by the applied financial reporting standard, due to the differences in population averages in one of the variables.

The carried out descriptive analysis has made possible evaluating the wide variability of ratios analyzed in different periods. This has proved that dispersion ranks vary according to the variables of the study. The results with IFRS application showed significant differences in population averages (p < 0,05) in some variables, such as Return on shareholders funds (X1), Return on capital employed (X2) and Return on total assets (X3). As a result, the financial information quality has been affected. In relation to the hypotheses contrast, it has been possible to prove that information quality is affected by the financial reporting standards used in the four empiric contrasts (H1, H2, H3 and H4). This results are similar to other studies (Henry et al., 2009 and Lantto and Sahlström, 2009).

Principal component analysis (PCA) for ratios for the year 2007

In this section, the principal component analysis technique is used to make tests to reduce variables (ratios), and, to determine independent ratios categories, for avoiding multicollinearity in the results of the analysis. The group of ratios, analyzed with the application of the multivariate technique for the year 2007, is explained in the methods section. Therefore, the main purpose of the analysis is to obtain a data matrix from mxn, m rows (companies) by n columns (ratios). Once all variables were introduced, it was proved that the obtained data offered a non-positively defined matrix, this prevents from applying PCA. Therefore, those with lost or disproportionate values in relation to the others were excluded. As a result, using the matrix of reproduced and residual correlations, and making an individual analysis for each variable, it was detected that the variables: X6, X14 (due to lost or not available values), X10, X11 and X22, in some cases, have high correlation values or extremely low ones. At the same time, they show high residual differences. Indeed, these variables affected the results of the analysis, and, therefore, this can cause the matrix to be indefinite, suggesting its progressive elimination.

Consequently, when variables are eliminated, it is observed that the obtained new matrix is positively defined and the analysis is carried out. In relation to the results of the previous tests, excellent data suitability is found, making PCA possible. Specifically, with a high Kaiser-Meyer-Olkin sampling suitability average of (0,63), and a Bartlett sphericity test with a significant high value of (A”¡2 = 3492,06 and p = 0,000), it can be proved from the analysis that the significance is perfect. Regarding the communalities obtained for each variable, once the extraction was carried out, it can be proved that all the variables show communalities higher than zero, with high values that approximately vary from 0,62 to 0,95, so the variance proportion is explained by the factors generated in the analysis. Regarding, the extraction method shown in table 4, it can be said that the factor solution that adjusts the best to the observed correlations converges on six factors. Therefore, with an accumulated explained variance of 87%, and considering that the first two factors explain around 44% of the total variance, these values are high enough to determine that six is a number of appropriate factors to explain the group of variables.

Extraction method: Principal Component Analysis.

The following step was to obtain a simpler interpretation of factors using the promax rotation method. Therefore, it can be observed that the solution converged on 6 iterations. The first factor that explains 29,19% of the variance, shows strong weighing in the variables related to the following ratios: X20 (profit per employee, 0,932), X21 (operating revenue per employee, 0,945), X23 (average cost per employee, 0,877), X24 (shareholders funds per employee, 0,933), X25 (working capital per employee, 0,702) and X26 (total assets per employees, 0,885). Consequently, it is reasonable to define this component as a dimension of "employees potential". It is also important to mention that the factor loadings of variables are shown with clear and high saturations in this factor. The second factor that explains 18, 19% of the total variance, shows its most significant positive loading factors in variables X1 (return on shareholders funds, 0,937), X2 (return on capital employed, 0,912) and X3 (return on total assets, 0,929); while the lowest loadings are observed in variables X4 (cash flow by sales volume 0,667) and X5 (profit margin, 0,864). Therefore, this factor gives an idea of the profitability, so it is convenient to label it as "profitability potential". When contrasting these results with Pinches et al. (1973) results, it can be noticed that variables X1 and X3 grouped in the factor "Return on investment", and variable X4, in the factor labeled "capital intensiveness". Meanwhile, according to Stevens (1973) variables X3 and X5 are in agreement with the factor "profitability". The third factor is integrated by variables X15 (current ratio, 0,965), X16 (liquidity ratio, 0,959) and X17 (shareholders liquidity ratio, 0,933). Additionally, it saturates with high and positive factor loadings, higher than 0,75, explaining 16,38% total variability of ratios. This component is labeled "operability potential". It should be mentioned that in Pinches et al. (1973), variables X15 and X16 are correlated in the factor "short-term liquidity".

The fourth factor, with an explained variance percentage of 11,03%, would be associated to variables X7 (EBIT margin, 0,866), X8 (EBITDA margin, 0,788) and X9 (net assets turnover, -0,664). As a result, considering the variables characteristics, this component has been labeled "investment potential". Besides, assuming a linear combination of these variables, it can be observed that variable X9 shows a weighing of 0,664 in opposite direction to the rest of variables from this factor. On the other hand, in relation to variable X7, Pinches et al. (1973) group it in the factor "Capital Intensiveness", Stevens (1973) in the factor "profitability". The fifth factor that explains 7,87% of the variance, would be associated to variables X12 (collection period, 0,910) and X13 (credit period, 0,780). This factor is labeled "credit potential". Finally, the sixth factor with an explained variance percentage of approximately 4,97%, would be associated to variables X18 (solvency ratio, 0,873) and X19 (gearing, -0,894), and it is labeled "potentiality of high operability". It can also be noticed that the variable "gearing", with a factor loading of 0,894, goes in an opposite direction in relation to the other identified variable in the component. It should be emphasized that variable X19 groups in the factor "leverage" according to Stevens (1973). The principal component analysis, carried out from the ratios corresponding to the year 2007, has allowed knowing the main relationships among variables, generating groups from the correlations between them. It is necessary to state that high factor loadings were observed in the performance of the components "employees potential" and "profits potential". Regarding the other factors, with lower values in the factor loadings; they do not contradict the found solution. However, it affirmatively indicates the polarization among the factors with higher or smaller explanatory capacity of data variance.

Conclusions

The carried out analyses made possible achieving the objectives of this study, which are related to the analysis of financial information quality, evaluating whether financial and economic ratios are affected by changes in financial standards applied by companies (Local GAAP or IFRS). At the same time, it was possible to determine a group of factors that allow proving the capacity of ratios to measure the quality of accounting information, as well as, facilitating the information analysis process to different groups of users. Additionally, the sample of selected companies belongs to Madrid Stock Exchange and Eurostoxx50. Financial statements from the period 2005-2007 were used. At the same time, a group of indicators (ratios) were considered to carry out the respective statistical measurements. The results obtained from the empiric contrast, once the medians were compared between the applied financial reporting standards for the financial statements presentation and the ratios from the period (2005-2007), show that in all the hypotheses (H1, H2, H3 and H4) there are differences in population averages in the financial presentation. Therefore, it is proved that financial information quality is affected by financial reporting standards. These results lead to think that one of the reasons for this performance is possibly attributed to the idea that one financial reporting standard is better than the other.

It could also be considered the interpretations that are made out of regulations, so that they do not affect the quality of disclosed data. This could justify the obtained results. The results of the principal component analysis (PCA) for the year 2007, allowed to group variables in six factors related with "employees potential", "operability potential", "investment potential", "credit potential" and "high operability potential". Therefore, the applied multivariate technique (PCA) has enabled to empirically check the capacity of ratios to measure financial information quality. Additionally, the found factors are only an alternative to group variables to make sense out of the interpretation of ratios from a theoretical point of view. Consequently, the results could be an alternative way for accounting users to analyze accounting information. Finally, the findings suggest continuous improvements in financial reporting standards applied to financial statements; so that, companies operations disclosed data fulfill some quality requirements, on behalf of accounting users. At the same time, it should be allowed that report analyses through accounting ratios, must be provided with interpretations as true as possible of the entityA´s financial and economic reality.

Cite this page