Abstract
The study applies Full-scale optimization model and Mean-Variance Model to construct asset allocation. In addition to using risky asset, we also add risk-free asset to compare the utility of the two approach in the sample and out of the sample. We use 1-year deposit, FTSE 100, and FTSE 250 as our investment portfolio and the period is from 2004 to 2012. Although Markowitz’s investment portfolio still plays an important role nowadays, many scholars suspect that Mean Variance Model can not describe different type of return and utility function. They suggest use Full Scale Optimization to allocate asset and the empirical results indicate that Full Scale optimization is better. However, our empirical results which is used different-period data indicates that in-sample data and out-of sample data influence the results. In other word, we have different results. Our results are as follows. First, Full Scale Optimization Model tends to increase weight of risky asset and high return asset compare to Mean Variance Model. Second, Mean Variance Model does not include the third and higher moment. Therefore, when the return of out-of-sample data is negative, the performance of Full Scale optimization may be worse than Mean Variance Model. Third, Adding risk-free asset can be more diversification.