Empirical study on performance attribution of the national social security fund's stock portfolios

The National Social Security Fund's stock portfolios are dynamic, yet timely measurement of individual manager's market timing ability and security selection ability is critical to maximize performance. Attribution analysis is the most effective way to clarify the causes of investment performance in dynamic stock portfolios. In order to keep inherent differences and asset pricing information's direct transfer, a holding proportion was introduced as the information variable. Using GMM methods to estimate the equation in order to reduce the impact of the residuals distribution that caused by using discrete data model. The results showed that regardless of the degree of fitness or the variable's significance, the total conditional model has better ability to explain performance.