This thesis is to explore the associations between two major factors and the pricing policy of the Depositary Receipt (DR) from the past experiences of Taiwanese listed companies. We first apply regression models to investigate the relationship between the first factor, the percentage change of the stock prices in that year of Taiwan Stock Exchange (TWSE) and the premium/discount in pricing DR. Then, we examine the connection between the second factor, the percentage change of the issuing companies¡¦ stock prices within the past 180 calendar days in TWSE prior to the offering of DR and it¡¦ final pricing. We also interviewed several financial experts to collect the industry norms on the pricing strategies of DR in order to help the future issuers meet their expectations. We have concluded the following: 1. With the significance level of 95%, there is no linear relationship between the final pricing of DR and the percentage change of the stock prices of TWSE. 2. With the significance level of 95%, there is also no linear relationship between the final pricing of DR and the percentage change of the issuer¡¦s stock prices for the past 180 days. 3. With the significance level of 95%, there exists very significant negative linear relationship between the final pricing of DR and the percentage change of the issuer¡¦s stock prices for the past 180 days, if we exclude all the cases of enjoying premium in DR pricing and all the stock prices that have a percentage increase over 170% prior to the offering of DR. 4. The best pricing policy for a company to issue a DR is to sign a contract with the bookrunner in which the limited discount percentage of DR¡¦s offering is guaranteed.

[pt] O objetivo deste trabalho é analisar a eficiência na precificação de três dos Exchange-Traded Funds brasileiros mais líquidos (BOVA11, SMAL11 e PIBB11), buscando determinar se eles seguem com bastante proximidade os índices que procuram replicar, comparando com o que é observado na literatura internacional no que diz respeito ao desempenho de ETFs estrangeiros. Inicialmente verificou-se a estratégia de replicação adotada, bem como a qualidade dessa replicação a partir da avaliação do tracking error observado nesses fundos. Em seguida buscou-se avaliar se há algum desvio na precificação entre o preço de negociação e o valor patrimonial líquido (NAV) do respectivo ETF, ou seja, se o ativo está sendo negociado, na média, com prêmio ou desconto. Por fim, foi analisada a persistência dos prêmios ou descontos encontrados, isto é, quanto tempo leva até o preço de mercado e o NAV voltarem ao equilíbrio. Os resultados encontrados mostram que os fundos BOVA11 e PIBB11 adotam uma estratégia de full replication, enquanto o SMAL11 apresenta uma estratégia de otimização. O tracking error encontrado está em linha com aqueles observados em ETFs europeus e os três fundos estudados são negociados, na média, com desconto. Finalmente, tanto BOVA11 e PIBB11 levam sete dias para voltarem ao equilíbrio, bastante acima da média observada na literatura internacional, enquanto o SMAL11 leva apenas dois dias para isso, o que é inesperado já que é o fundo menos líquido dentre os analisados. Os resultados indicam que as ferramentas de arbitragem não estão sendo utilizadas de maneira eficiente. / [en] The objective of this study is to analyze the pricing efficiency of three of the most liquid brazilian Exchange-Traded Funds (BOVA11, SMAL11 and PIBB11) and determine if they follow closely the indexes they try to replicate, comparing with the international literature regarding the foreign ETFs performance. Firstly, this study verifies which strategy is adopted by each fund, as well as the quality of this replication by evaluating the tracking error observed in these funds. Then it is analyzed if there is any deviation between the trading price and the net asset value (NAV) of the respective ETF, that is, if the security is being traded, on average, with a premium or discount. Finally, it is evaluated the persistence of those premiums and discounts found, that is, how much time it takes until the trading price and the NAV go back to equilibrium. The results showed that both BOVA11 and PIBB11 adopt a full replication strategy, while SMAL11 presents an optimization strategy. The tracking error found is in line with those observed in European ETFs and the three funds are traded, on average, with a discount. Finally, both BOVA11 and PIBB11 take seven days to go back to equilibrium, while SMAL11 only takes two days, an unexpected result as this is the least liquid fund of the three that are part of this study. Therefore, the arbitrage tools are not being used efficiently.