This dissertation includes three chapters in which the first two are on return predictability and the third is on yield curve and yield factors. The abstract of each of them is as follows: 1), This paper proposes using capital gains instead of total returns in return predictability tests. Total return predictability can be inferred from capital gain predictability since total returns with dividends are highly correlated with returns based on capital gains only. An exact linear relationship exists among log dividend growth, log capital gain and log dividend price ratio. This exact linear relationship has similar implication as the Campbell-Shiller (1988) linear approximation but is more precise and easier for predictability tests. I verify the standard empirical findings on return predictability using capital gain predictability. Separation of price change and dividend change also leads to a new finding: shocks to dividend growth is shown to have significant positive correlation with shocks to dividend price ratio in the vector autoregressive regression (VAR) rather than close to zero as shown in previous literature. 2), This paper tests the return predictability of the cyclical and trend components in the log dividend price ratio. The log dividend ratio is found to have a near-unit root trend factor if the expectation of the future discount factor is highly persistent. We use Bayesian analysis and the Kalman filter to extract the strictly stationary and near-random-walk components in the log dividend price ratio. The extracted cyclical process can predict one-year ahead total returns during the post-war period and one-year ahead dividend growth rates during the pre-war and war period with notable R^2. We also demonstrate a reverse of predictability: returns become more predictable while dividend growth rates become more unpredictable. 3), This paper examines the fourth principal component of the yields matrix, which is largely ignored in macro-finance forecasting applications, in the context of predicting excess bond returns. Using yields data from the Fama-Bliss and the Federal Reserve, we present the significant in-sample and out-of-sample predictive power of models including the fourth yield factor. Additionally, the "return-forecasting factor" in Cochrane and Piazzesi (2005) is shown to be a restricted linear combination of all yield factors and to be highly correlated with the second and fourth factors. We interpret the fourth yield factor as a factor representing "S-shape" (the shape of a sigmoid curve) and demonstrate the connection between the S-shape factor and the yield curve.