O aumento do valor de mercado e da liquidez em bolsa das ações e a expansão da cobertura de analistas e da participação de investidores institucionais são apontados na literatura como metas centrais dos programas de Relações com Investidores - RI. Os resultados de estudos internacionais apontaram relação significativa entre características dos programas de RI das companhias e a consecução destas metas. Entretanto, para o cenário brasileiro, as pesquisas sobre relações com investidores em sua maioria tomam o formato de questionários e visam avaliar a estrutura e o valor das RI na percepção dos profissionais da área e do mercado. Pouco realmente se sabe sobre os efeitos das RI no comportamento das ações das empresas brasileiras. Portanto, o objetivo deste estudo é verificar se a estrutura e as atividades de relações com investidores são realmente fatores determinantes da valorização das ações, liquidez em bolsa, cobertura de analistas e participação de investidores institucionais para as companhias abertas brasileiras. Os resultados encontrados apontam que as características dos departamentos de RI tem impacto indireto na liquidez das ações, através da expansão da cobertura de analistas e do número de investidores institucionais. O número de funcionários do departamento de RI apareceu como determinante importante tanto da cobertura de analistas quanto da participação de investidores institucionais. Já a qualidade das relações com investidores e o fato da empresa possuir um diretor exclusivo de RI apareceram como fatores explicativos importantes da cobertura de analistas. A principal contribuição deste trabalho é a de analisar uma questão até então negligenciada pela literatura acadêmica brasileira, mas que pode ter implicações importantes para uma das atividades vitais das companhias: seu relacionamento com os investidores. / The increase in the stocks’ market value, liquidity, analyst following and participation of institutional investors are mentioned in the literature as central goals of Investor Relations programs. International studies have shown the relationship between characteristics of IR programs and the achievement of these goals. However, to the Brazilian context, research on investor relations mostly take the form of questionnaires and aim to evaluate the structure and value of IR in the perception of IR professionals and the market. Little is actually known about the effects of IR on the behavior of the stocks of Brazilian companies. Therefore, the aim of this study is to verify if the organization and activities of investor relations are actually determinants of stock appreciation, stock liquidity, analyst following and institutional ownership for Brazilian companies. The results show that the characteristics of the IR department have indirect effect on the liquidity of the stocks through the expansion of analyst coverage and institutional ownership. The number of employees working in the IR department emerged as important determinant of both analyst following and institutional ownership. The quality of investor relations program and the fact that the company has an exclusive director of IR emerged as important explanatory factors of analyst following. The main contribution of this paper is to analyze an issue so far neglected by the Brazilian academic literature, but which may have important implications for one of the vital activities of the companies: their relationship with investors.

O aumento do valor de mercado e da liquidez em bolsa das ações e a expansão da cobertura de analistas e da participação de investidores institucionais são apontados na literatura como metas centrais dos programas de Relações com Investidores - RI. Os resultados de estudos internacionais apontaram relação significativa entre características dos programas de RI das companhias e a consecução destas metas. Entretanto, para o cenário brasileiro, as pesquisas sobre relações com investidores em sua maioria tomam o formato de questionários e visam avaliar a estrutura e o valor das RI na percepção dos profissionais da área e do mercado. Pouco realmente se sabe sobre os efeitos das RI no comportamento das ações das empresas brasileiras. Portanto, o objetivo deste estudo é verificar se a estrutura e as atividades de relações com investidores são realmente fatores determinantes da valorização das ações, liquidez em bolsa, cobertura de analistas e participação de investidores institucionais para as companhias abertas brasileiras. Os resultados encontrados apontam que as características dos departamentos de RI tem impacto indireto na liquidez das ações, através da expansão da cobertura de analistas e do número de investidores institucionais. O número de funcionários do departamento de RI apareceu como determinante importante tanto da cobertura de analistas quanto da participação de investidores institucionais. Já a qualidade das relações com investidores e o fato da empresa possuir um diretor exclusivo de RI apareceram como fatores explicativos importantes da cobertura de analistas. A principal contribuição deste trabalho é a de analisar uma questão até então negligenciada pela literatura acadêmica brasileira, mas que pode ter implicações importantes para uma das atividades vitais das companhias: seu relacionamento com os investidores. / The increase in the stocks’ market value, liquidity, analyst following and participation of institutional investors are mentioned in the literature as central goals of Investor Relations programs. International studies have shown the relationship between characteristics of IR programs and the achievement of these goals. However, to the Brazilian context, research on investor relations mostly take the form of questionnaires and aim to evaluate the structure and value of IR in the perception of IR professionals and the market. Little is actually known about the effects of IR on the behavior of the stocks of Brazilian companies. Therefore, the aim of this study is to verify if the organization and activities of investor relations are actually determinants of stock appreciation, stock liquidity, analyst following and institutional ownership for Brazilian companies. The results show that the characteristics of the IR department have indirect effect on the liquidity of the stocks through the expansion of analyst coverage and institutional ownership. The number of employees working in the IR department emerged as important determinant of both analyst following and institutional ownership. The quality of investor relations program and the fact that the company has an exclusive director of IR emerged as important explanatory factors of analyst following. The main contribution of this paper is to analyze an issue so far neglected by the Brazilian academic literature, but which may have important implications for one of the vital activities of the companies: their relationship with investors.

Financial Derivatives are one of the most popular and emerging innovations in the field of financial engineering. Since their inception, there has been a phenomenal growth in the volumes of derivatives traded all over the world. Financial markets are known to be extremely volatile and derivatives provide a way of eliminating or reducing the risks involved in these markets. Since these instruments derive their value from some underlying asset, trading in these instruments is bound to affect the underlying assets. Thus it becomes important to examine what these effects are and whether they have been favourable or detrimental to the underlying stock markets specially when there has been an explosive growth of these financial derivatives all over the world. This issue gains more importance in the case of emerging markets like India as they try to be more competitive and efficient as the developed Western markets. This thesis mainly deals with looking at this impact on the Indian stock markets. The Indian markets still being very new in this area, not many studies have been reported here related to this issue. The main focus of this thesis is to provide some more evidence on the impact of one kind of derivative instrument, namely options on different characteristics of underlying stocks in the Indian stock market. The thesis has the following objectives: • To examine the impact of option introduction on the price of underlying stocks in National Stock Exchange (NSE). • To examine the impact of option introduction on the volatility of underlying stocks in NSE • To examine the impact of option introduction on liquidity of underlying stocks in NSE NSE introduced derivatives beginning with index futures on June 12, 2000, followed by index options on June 4, 2001, options on individual securities on July 2, 2001 and finally futures on individual securities on November 9, 2001. Due to the temporal proximity of the introduction of index options and individual options, there exists a possibility of an interaction of these two effects. This problem is solved by a judiciously chosen sampling design. In particular, three groups of stocks are considered. The first group consists of stocks on which options were first introduced on 2nd July 2001 and thus would exhibit a combined effect of the two events if any. The second group consists of stocks on which options were introduced much later and therefore would show effects of individual option introduction if any. The third group comprises of nonoptioned stocks whose returns are considered around the date of index option introduction and thus would show effects of index option alone if any. To separate the two effects an ANOVA/ Logistic Regression model is used. An objective selection of the event and estimation windows is done using a Bayesian Change Point Analysis. The first part of the thesis looks at the effect of option introduction on the price of underlying stocks. A standard event study methodology as has been used in the literature is employed for this purpose. The study does not find any significant effect of option introduction on the prices. The second part of the thesis deals with the effect on volatility. Volatility is measured as the risk of a stock and as is done in the literature, three kinds of risk are looked at: total risk, systematic risk and the unsystematic risk. In case of the total risk, an F-test and an Ansari Bradley test is used to check for changes in the variance and scale parameters of market-adjusted continuously compounded returns of the stocks before and after option introduction. The results of these tests are recorded as a categorical variable taking on the value 0 for no change and 1 for a change and a Binomial Logistic Regression is used to separate the effects of the two events. Furthermore, after recording the results of the above mentioned tests as a categorical variable with three categories (0, 1, -1), a Multinomial Logistic Regression is also used in order to estimate the direction of the change (increase, decrease or no change). The ratios of after to before total risks are also analyzed using an ANOVA model. The systematic risk is measured using three kinds of betas – OLS betas, Scholes-Williams betas and Fowler-Rorke betas. The differences in the before and after betas of every stock are modelled using an ANOVA model in order to separate the two effects as well as the interaction effect. The unsystematic risk is estimated by the conditional variances and the unconditional variances of ARMA and ARMA-GARCH models fitted to market model excess returns. The ANOVA model is used here as well. In addition to this, the before and after ARCH and GARCH coefficients of GARCH (1, 1) models fitted to the excess returns are also compared using the ANOVA model. The results indicate that individual options are leading to a decline in total risk however index options are causing an increase in total risk. The interaction effect is significant in this case thereby causing an increase in total risk in the Group I stocks. The OLS betas indicate that individual option introduction seems to have increased the systematic risk. The Scholes-Williams betas indicate that index option introduction seems to have increased the systematic risk. The Fowler Rorke betas on the other hand, do not show any significant impact of individual option or index option introduction. For all the three betas index options introduction seems to have no effect on the systematic risk. Though the interaction effect seems to be significant in all the three cases, it however does not significantly affect the systematic risk in Group I stocks. As regards the unsystematic risk, both the conditional and unconditional variances of ARMA models show a significant reduction for individual option introduction but index options do not have any significant impact on either one of these measures. In case of unconditional variances of ARMA-GARCH models, none of the effects come out as significant. While comparing the news and persistence coefficients of GARCH (1, 1) models, the news coefficients indicate that the due to index option introduction, stocks are becoming more efficient in terms of absorbing the news more rapidly. No significant effect of either event is found on the persistence coefficients. The last part of the thesis deals with the liquidity issue. Liquidity has been measured using two measures – relative volume (based on daily data) and implicit bid-ask spread given by Roll (1984) (calculated from intra-day data). In case of the liquidity measures, the Logistic Regression models are used i.e. a categorical variable with two or three categories obtained from the results of a Wilcoxon Rank Sum test for comparing the median volume and spread before and after option introduction, is used. It is found that for the relative volume, individual option introduction has led to a favourable effect in terms of increasing the volume post introduction of options; however index options seem to have had a negative effect. As for the spread, index options seem to have had a stabilizing influence on the underlying stocks than the individual options.

Made available in DSpace on 2010-04-20T20:48:21Z (GMT). No. of bitstreams: 3 98397.pdf.jpg: 24214 bytes, checksum: 90792e5ee1e13b6da84a2d989e957018 (MD5) 98397.pdf: 400450 bytes, checksum: 0fb67fdc5e1861a8aa530eac58cc2e7e (MD5) 98397.pdf.txt: 224072 bytes, checksum: ab99c41276e2da8d41ff995a72e7de5e (MD5) Previous issue date: 2005-05-11T00:00:00Z / O presente estudo teve como objetivo explicar o comportamento da volatilidade dos retornos das principais ações negociadas na Bovespa no período compreendido entre janeiro de 1995 e setembro de 2003. O trabalho buscou contribuir de diversas maneiras para o estudo da volatilidade dos retornos das ações no mercado brasileiro. Primeiro, fazendo uma exposição abrangente das diversas teorias e modelos que têm sido desenvolvidos para explorar os fatores determinantes da volatilidade dos retornos das ações. Dentre as teorias exploradas, o estudo trabalhou com a teoria da alavancagem, a teoria da retroalimentação da volatilidade, o modelo das diferenças de opinião entre os agentes econômicos e os modelos de preço–volume. Outra contribuição importante do trabalho foi utilizar uma amostra ampla de 35 ações com níveis aceitáveis de liquidez, utilizando a metodologia própria do estudo. Além disso, o estudo buscou agregar aos modelos econométricos não apenas variáveis microeconômicas mas também as macroeconômicas (conjunturais). Finalmente, o trabalho discutiu os resultados dos modelos à luz das especificidades do mercado acionário brasileiro. Dentre os resultados encontrados, foi possível observar que a volatilidade dos retornos das ações é impulsionada pelo grau de alavancagem financeira das companhias. No Brasil, a volatilidade dos retornos das ações segue o mesmo caráter persistente encontrado em mercados desenvolvidos, como o dos Estados Unidos. O giro dos negócios com as ações tende a alimentar a volatilidade dos papéis. Um giro maior de negócios pode ser resultante de um incremento na quantidade de informações que chegam ao mercado, o que provoca revisões nos preços das ações e estimula um aumento na volatilidade dos retornos. Quando aliado às divergências de opinião entre os investidores, um giro maior também pode alimentar a volatilidade dos retornos. Diferentemente do previsto pelas teorias, os resultados do presente estudo não corroboraram o argumento de que existe maior volatilidade dos retornos das ações em momentos de retornos negativos. Os resultados indicaram ainda clara influência de algumas variáveis macroeconômicas sobre a volatilidade dos retornos das ações. Uma redução no Produto Interno Bruto do Brasil tende a incrementar a volatilidade dos retornos das ações, possivelmente fruto de uma perspectiva de instabilidade econômica ou recessão. A volatilidade dos retornos no mercado acionário também aumentou quando da alteração do regime cambial implementada no início de 1999. O estudo mostrou que não parece haver uma dependência significante entre os mercados de ações do Brasil e dos Estados Unidos. Finalmente, a inclusão de variáveis relacionadas com características específicas do mercado acionário brasileiro mostrou resultados interessantes. Parece existir uma relação direta entre volatilidade dos retornos e ações preferenciais, oriunda possivelmente do alto giro dos negócios com esse tipo de ação ou da ausência do direito de voto por parte dos investidores, aumentando o risco do ativo. Contudo, testes adicionais são necessários para confirmar tal associação. No que diz respeito à adoção de melhores práticas de governança por parte das companhias, os resultados parecem corroborar o argumento de que a preocupação com a boa governança tende a reduzir o risco percebido pelos investidores com relação às ações. / This study aimed to explain the behavior of stock return volatility concerning the main stocks negotiated in São Paulo Stock Exchange (Bovespa) from January, 1995 through September, 2003. The study wanted to contribute in several ways for the understanding of stock return volatility in Brazilian market. First, it shows the theories and models developed to explain the determining factors of stock return volatility. Among those theories, this work explored leverage theory, volatility feedback theory, differences of opinion models and price– volume models. Another important contribution of this study was to use a sample of 35 stocks with minimum levels of liquidity using the study own methodology. Additionally, microeconomic and macroeconomic variables were added to the econometric models. Finally, this study discussed the results and their relation with the idiosyncrasies of Brazilian stock market. Among the results, it was observed that stock return volatility is influenced by the company leverage. In Brazil, stock return volatility follows the persistent behavior found in developed capital markets. The stock turnover tends to increase stock volatility. Maybe the high stock turnover is caused by new information increasing and revisions in stock prices. Those revisions stimulate stock return volatility. When the stock turnover occurs jointly with differences of opinion among investors, a higher turnover may produce stock return volatility. Contrary to the theory, the results didn’t corroborate the argument that stock return volatility is bigger when the return is negative. Additionally, results indicated influence of macroeconomic variables over the stock return volatility. A reduction in Brazilian GNP is associated with the increment in return volatility, possibly because it signalizes the perspective of recession. Also, stock return volatility increased in 1999 after the implementation of new exchange regime in Brazil. The study results showed that Brazilian stock market is not dependent of U.S. stock market. Finally, the inclusion of variables concerning the idiosyncrasies of Brazilian stock market showed interesting results. It seems that there is a positive association between stock return volatility and preferred stocks, maybe because Brazilian preferred stocks have higher turnover than common stocks and the lack of voting rights concerning preferred stocks increases the risk for investors. However, additional tests are necessary to confirm that association. Concerning the adoption of good corporate governance practices, the results of this study corroborate the argument that better corporate governance reduces stock risk.

Abstract Title: Direct and indirect effect of notes – for actors in a stock market Background: Notes make up a significant part of the company´s annual report, but does this information have any major impact? On one side, notes should lead to less information asymmetry, which positively affects the stock market and financial analysts forecasting precision as well as this relationship depends on different contexts. The question is also asked if all information presented in the notes is too extensive, which creates information overload for financial analysts. Purpose: The purpose is to explain the effects of more information in accounting notes for actors in equity markets in different contexts that have the same accounting standard. Method: The study applies a cross-sectional design together with a deductive approach to investigate whether the amount of note information has a relationship with financial analysts forecasting precision, stock volatility and stock liquidity. For these relationships, moderating effects will also be tested. The hypotheses arose through theories of asymmetric information, effective market hypothesis, cognitive load theory and system-orientated theories. Conclusion: The study rejects the relationship between note information and errors in financial analyst forecast, stock volatility as well as stock liquidity. However, the results show a significant positive relationship between note information and error in forecasting precision. Financial analyst forecasting precision show that the relationship is affected by three different contexts. Despite the relationship between note information and stock liquidity were rejected when tested on the overall sample is the relationship affected by two different contexts. / Sammanfattning Titel: Direkta och indirekta effekter av noter - För aktörer på en aktiemarknad Bakgrund: Noter utgör en väsentlig del av företags årsredovisning men har den informationen någon större påverkan? På ena sidan bör noter leda till mindre informationsasymmetri, vilket påverkar aktiemarknaden och finansiella analytikers prognosprecision positivt samt att detta samband även beror på olika kontexter. Frågan ställs även ifall informationen som presenteras i notavsnittet är för omfattande, vilket skapar informationsöverbelastning för finansiella analytiker. Syfte: Syftet är att förklara effekten av ökad delgivande av information i redovisningsnoter på aktörer på aktiemarknaden i olika kontexter som tillämpar samma redovisningsstandard. Metod: Studien tillämpar en tvärrsnittsdesign tillsammans med en deduktiv ansats för att undersöka ifall mängden notinformation har något samband med finansiella analytikers prognosprecision, aktievolatilitet och aktielikviditet För dessa samband testas även modererande effekter. Hypoteserna uppstod genom teorierna asymmetrisk information, effektiva marknadshypotesen, kognitiv belastningsteori och systemorienterade teorier. Slutsats: Studien förkastar sambandet mellan notinformation och fel i prognosprecision, aktievolatilitet och aktielikviditet. Däremot visar resultatet ett positivt signifikant samband mellan notinformation och fel i prognosprecision. Finansiella analytikers prognosprecision uppvisar även att sambandet påverkas av tre olika kontexter. Trots att sambandet mellan notinformation och aktielikviditet förkastas när den testas på en övergripande nivå, påverkas sambandet av två olika kontexter.

本研究目的為透過實證分析研究2005年至2008年台灣上市櫃公司股票流動性與企業價值兩者之關係，企圖瞭解當考慮其他已證實會影響企業價值之因素，如：公司治理、資訊揭露程度、公司特有風險等因素後，股票流動性是否依然對企業價值存在影響力，即本研究將在控制公司治理程度、資訊揭露程度及公司特有風險等因素下，以周轉率作為股票流動性之代理變數、以Tobin’s Q作為企業價值之代理變數，分別使用普通最小平方法與兩階段最小平方法，檢視股票流動性是否依然影響企業價值。除此之外，亦參考Vivian W. Fang, Thomas H. Noe, Sheri Tice[2009]的研究方法，將代表企業價值之Tobin’s Q拆解成其他三個變數：營業利益對股價比、權益佔資產比及資產報酬率，取代企業價值成為應變數，再以相同研究模型對股票流動性進行實證分析，以求瞭解股票流動性影響企業價值之背後原因。最後，將研究樣本分別依照不同程度之公司治理、不同程度之資訊透明度、不同程度之公司特有風險分成低、中、高三組，以普通最小平方法分析在不同程度之控制變數下，股票流動性對企業價值之影響是否不同。本研究結果如下：第一、股票流動性越高之企業，其企業價值亦越高。第二、股票流動性高的企業，擁有較低之負債比與較高之資產報酬率，是企業價值高的中間影響因子。第三、在不同的董監持股比率、董監質押比率、法人持股比率、獨立董監比率、資訊揭露程度及公司特有風險下，流動性對企業價值的影響力是不同的。公司治理越差、資訊揭露程度越低、公司特有風險越高的公司，其流動性對企業價值越具顯著影響力。 / This paper investigates the relationship between stock liquidity and firm performance in Taiwan by using both of Ordinary Least Squares method and two-stage least squares method. To access whether stock liquidity improves, harms, or has no effect on firm performance, a proxy for Tobin’s Q is regressed on the liquidity measure, stock turnover, and several control variables, for example, corporate governance, information transparency, idiosyncratic risk, etc. Next, In order to gain further insight into the source of higher firm performance for stocks with high liquidity, based on Vivian W. Fang, Thomas H. Noe, Sheri Tice (2009), I break Tobin’s Q into three components: price-to-operating earnings, financial leverage, and operating profitability and then use these three variables as dependent variables to be regressed respectively on the liquidity measure in the same regression model. Besides, I also examine whether the effect of high liquidity on firm performance is different for firms with different level of each control variable. At last, this study shows that liquidity positively affects firm performance and proves that stocks with high liquidity not only have better firm performance, but also have more equity in their capital structure and higher operating profitability levels. This study also finds that the effect of stock liquidity on firm performance is significantly different for firms with different level of control variables.

Submitted by João Patah (joaopatah@uol.com.br) on 2012-02-13T23:00:13Z No. of bitstreams: 1 Dissertação final - entregue.pdf: 712466 bytes, checksum: 84c7a6ea77cae324255c3817c658d668 (MD5) / Rejected by Gisele Isaura Hannickel (gisele.hannickel@fgv.br), reason: Prezado João, Após a folha da ficha catalográfica, deve constar a folha de assinaturas (com o nome dos membros que compuseram sua banca). Modelo no site da Biblioteca: http://bibliotecadigital.fgv.br/site/bkab/normalizacao Att, Secretaria de Registro on 2012-02-14T11:24:34Z (GMT) / Submitted by João Patah (joaopatah@uol.com.br) on 2012-02-14T12:17:17Z No. of bitstreams: 1 Dissertação final - entregue.pdf: 714165 bytes, checksum: ccfbe05ea31a2bf5542b1f78fe329c13 (MD5) / Approved for entry into archive by Gisele Isaura Hannickel (gisele.hannickel@fgv.br) on 2012-02-14T12:57:03Z (GMT) No. of bitstreams: 1 Dissertação final - entregue.pdf: 714165 bytes, checksum: ccfbe05ea31a2bf5542b1f78fe329c13 (MD5) / Made available in DSpace on 2012-02-14T13:09:50Z (GMT). No. of bitstreams: 1 Dissertação final - entregue.pdf: 714165 bytes, checksum: ccfbe05ea31a2bf5542b1f78fe329c13 (MD5) Previous issue date: 2012-01-30 / Os efeitos dos diferentes graus de concentração do controle e de governança corporativa das empresas no desempenho das ações têm crescido em interesse no Brasil. Este estudo utiliza a concentração do poder de voto e um índice de governança corporativa para analisar se há impacto destes fatores na liquidez das ações, o que poderia ser um canal de sua influência nos diferenciais de retorno. Becht (1999) conclui que maior concentração de controle reduz a liquidez, Attig el al. (2006) argumentam que a separação do poder de voto e do direito sobre o fluxo de caixa levam à redução da qualidade de informações e expropriação de minoritários, reduzindo a liquidez das ações, e Chung, Elder e Kim (2010) concluem que melhoria da governança interna tem impacto positivo na liquidez. Os resultados obtidos neste estudo sugerem que, para uma amostra de empresas brasileiras, o indicador de governança corporativa não influencia a liquidez com significância estatística. A maior liquidez das ações está em empresas com concentração de controle entre 30% e 50%, com menores preços de negociação, abertura de capital recente e/ou listadas no Novo Mercado. / The effects of different degrees of control concentration and corporate governance on stock returns have rising interest in Brazil. This study makes use of the voting power concentration and of a corporate governance index to examine whether these factors affect stock liquidity, which can be the channel of influence in excess returns. Becht (1999) concluded that concentration of voting power reduces liquidity, Attig el al. (2006) argue that deviations between ultimate control and ownership result in reduced quality of information and extraction of value from minority shareholders, reducing liquidity, and Chung, Elder and Kim (2010) concluded that better corporate governance positively impacts liquidity. The statistical results of this study suggest that, for a sample of Brazilian companies, the corporate governance index does not affect liquidity. Superior liquidity is found when concentration of voting power is between 30% and 50%, with lower share prices, recently listed and on the Novo Mercado segment.

[pt] Este estudo avaliou o comportamento das diferenças de preços entre ações ordinárias e preferenciais de um conjunto representativo de empresas com ações negociadas na BOVESPA no período 1995-1999, testando o impacto também das mudanças na lei das SAs ocorridas em 1997. Os testes realizados indicaram que no decorrer do período estas mudanças influenciaram as diferenças de preços entre ações ordinárias e preferenciais para a maioria das empresas, passando as ações ordinárias a serem negociadas, em geral, a preços de mercado inferiores às ações preferenciais. A partir da identificação na literatura dos principais fatores que podem causar a diferença de preços entre ações de uma mesma empresa foram realizadas regressões de seção transversa para cada ano do período, assim como para o período como um todo, para testar a capacidade explicativa de modelos contendo variáveis explicativas derivadas da liquidez, da estrutura de capital e da composição acionária de cada empresa, bem como os dividendos pagos a cada classe de ação. Para o período como um todo podemos concluir que os modelos explicam parcialmente as diferenças de preços. Os resultados obtidos mostraram que os modelos apresentam resultados satisfatórios a partir de 1996, sendo que a capacidade explicativa e a confiabilidade são maiores a partir de 1998. As variáveis representando a liquidez e o percentual de ações preferenciais com os controladores,bem como os impactos da mudanças na lei das SAs foram as que mostraram melhor significância estatística no período como um todo. Os coeficientes lineares obtidos nas regressões para as variáveis de liquidez foram positivos, conforme era esperado, representando a correlação destas medidas com as diferenças de preços entre ações ordinárias e preferenciais. Os possíveis impactos de novas alterações na lei das SAs na relação de preços entre ações ordinárias e preferenciais e no mercado como um todo podem representar interessante objeto de pesquisa para novos estudos. / [en] This study evaluates the price differences between voting and non-voting shares of a representative set of Brazilian companies traded on the São Paulo Stock Exchange from 1995 to 1999, assessing also the impacts of the amendments in the Brazilian Corporate Law that take place in 1997. The tests showed that, due to these changes, for most companies a downtrend occurred in the price differences, and, as time passed, the voting share was usually trading at a discount to the non-voting share. After identifying in the literature some key variables that influence these prices differences, we used some cross-section regressions for each year of the period to test the model with explicatory variables related to the liquidity, capital structure and shareholder composition of each company, as well as dividends paid to each class of share. The models tested were statistically significants for the whole period, explaining partially these price differences. The results of the regressions for each annual period showed that, after 1996, the models presented satisfactory results, with better results and degrees of confidence after 1998. The liquidity variables and the variable that represents the controller`s ownership of non-voting shares were the ones that showed the better degree of confidence during the period. The linear coefficients of the liquidity variables were positive, as expected, representing the correlation between these measures and the price differences between voting and non-voting shares. New changes in the corporate law are being studied and may affect the price differences between voting and non-voting shares and the valuation of all the stock market, representing an interesting subject to future studies. / [es] Este estudio evalúa el comportamiento de las diferencias de precios entre acciones ordinarias y preferenciales de un conjunto representativo de empresas con acciones negociadas en BOVESPA en el período comprendido entre los años 1995- 1999. Se evalúa también el impacto de los cambios en la ley de las SAs, ocurridos en 1997. Las pruebas realizadas indicaron que, en el transcurso del período mencionado, dichos cambios influyeron en las diferencias de precios entre acciones ordinarias y preferenciales para la mayor parte de las empresas, y las acciones ordinarias pasaron a ser negociadas, en general, a precios de mercado inferiores a las acciones preferenciales. Se llevó a cabo un estudio bibliográfico para identificar los principales factores que pueden causar diferencias en el precio de las acciones de una misma empresa. Con estos factores se realizó una regresión transversa para cada año del período 1995-1999 y también considerando el período completo. Estas regresiones tienen como objetivo, probar la capacidad explicativa de los modelos que contienen variables explicativas derivadas de la líquidez, de la extructura de capital y de la composición de las acciones de cada empresa, así como los dividendos y pagos a cada clase de acción. Considerando el período completo, podemos concluir que los modelos explican parcialmente las diferencias de precios. Los resultados muestran que los modelos presentan resultados satisfactorios a partir de 1996, siendo que la capacidad explicativa y la confiabilidad son mayores a partir de 1998. Al considerar el período completo, las variables que se mostraron estadísticamente significativas fueron: las variables que representan la líquidez, el porcentaje de acciones preferenciales con los controladores y el impacto de los cambios en la ley de las SAs. Los coeficientes lineales obtenidos em las regresiones para las variables de líquidez fueron positivos, tal y como se esperaba, representando la correlación de estas medidas con la diferencia de precio entre acciones ordinarias y preferenciales. El posible impacto de nuevas alteraciones en la ley Sas en relación a los precios entre acciones ordinarias y preferenciales y en el mercado como un todo pueden representar un interesante objeto de investigación para nuevos estudios.