Testing the household Euler equation with consumption data faces econometric challenges caused by large measurement errors in the data and a short time span. We adopt a framework to test the Euler equation with stock market data to alleviate the measurement error and short time span issues. Utilizing a data-driven group panel data method, we identify a heterogeneous pattern of Euler equation failure among different groups of listed firms. The identified degree of Euler equation failure is significantly related to firm characteristics that are associated with famous stock anomalies. We show that the correlations between the degree of Euler equation failure and firm characteristics provide a new set of stylized facts that can help us distinguish between different economic theories on Euler equation failures and asset pricing anomalies, and identify challenges facing current theories.

We incorporate econometrics approach into panel data methods to examine the impact of the board competence on the performance in Chinese banks. By investigating the biographical background of directors in the 20 largest banks during the period 2008 to 2016, our estimate shows that the board of directors play a prominent role in the performance. Especially, both the education background and the management experience have negative impact on the performance.

The pairs (or repeat) transactions index model is the most widely used method for constructing housing market index and for estimating individual house prices. However, this model is unable to provide price level information on the housing market and encounters several statistical problems when individual houses have more than two transactions. In order to overcome the shortcomings of pairs transactions model, we propose a multi transactions model with the panel data approach to estimate market index and to predict house prices. The empirical results based on 5,000 houses in Howard County Maryland demonstrate that the proposed multi transactions model: (1) can provide price level information on the housing market; (2) produces market index with smaller standard errors; (3) is more robust in terms of difference between the sub-sample estimate and full sample estimate; (4) has smaller index revision volatilities; and (5) performs better in the out-of-sample test on the prediction of individual house prices. Thus, the multi transactions model is recommended for constructing housing market index and assessing individual house prices.