DescriptionMy dissertation comprises of three essays: 1) Accounting information and financial derivatives: a literature review 2) The Effect of Option Transaction Costs on Informed Trading in the Option Market around Earnings Announcements; and 3) The Effects of Credit Default Swaps trading on Analyst Forecast Properties. The first essay surveys the previous researches on accounting information and financial derivatives. The financial derivate instruments we mainly focus on are stock option and credit default swaps. Then we also identify some research gaps for future research. The second essay investigates the effect of transaction costs related to trading options on the directional and volatility informed trading in the option market. We find that both forms of informed trading are significantly stronger among firms with lower option bid-ask spread. Importantly, the effect of transaction costs is significant around earnings announcements, but not significant (on average) around randomly chosen dates with no events of consequence. This suggests that transaction costs play a particularly important role during information intensive periods. Trading strategies based on directional informed trading and option transaction costs earn monthly abnormal returns of 1.39% to 1.91%. The third essay investigates whether the initiation of credit default swaps (CDSs) trading can affect analysts' forecast properties. Using a difference-in-difference research design, we find that the onset of CDS trading help analysts to increase forecast accuracy, which is consistent with notion that a new financial market facilitate information discovery and dissemination. This effect is more pronounced for firms with greater information asymmetry and higher leverage. We also find that CDS initiation can depress analysts' strategic forecast optimism. Relying on several proxies for analysts' strategic optimism, we find that the depressing effect is more pronounced for subsamples with higher optimism level. In addition, we find that the depressing effect is stronger when bad news is realized ex post in the earnings announcement date.