基于序关系重构的恒定再平衡投资组合

CRP（Constant Rebalancing Portfolios，即恒定再平衡投资组合）是一种重要的投资组合，它要求资产收益率序列必须满足独立同分布条件，这个限定条件影响了CRP在实际金融投资中的应用。为了使CRP发挥与其理论基础相称的实际效果，对真实金融数据进行基于序关系的重构，使得重构后的数据接近于满足独立同分布条件。在重构后的数据上构建最优CRP，这种策略称为rankCRP。证明了在适当的条件下rankCRP具有最优性。运用国际国内多个真实市场的大型历史数据进行检验，检验结果表明rankCRP在真实市场的表现优于CRP，远远超越市场指数，甚至战胜涨幅最大的证券。RankCRP在理论上的最优性表明，通过适当的重构，CRP也可以适用于非独立同分布收益率时间序列，拓宽了CRP理论的适用领域。RankCRP在实践上的有效性对金融投资实务有重要意义。

Rank-Reconstruction Based Constant Rebalanced Portfolios

The Constant Rebalanced Portfolio (CRP) is an asset allocation algorithm which keeps the same distribution of wealth among a set of assets along a period of time. The best CRP is optimal when the stock return time series is I.I.D.. However the unrealistic I.I.D. condition set the CRP back from application. By means of reconstructing the real financial data to be I.I.D., it is possible to apply the CRP to investment industry. The optimality of rank-reconstruction based Constant Rebalanced Portfolios (rankCRP proposed in this paper) is proved theoretically. And the empirical study on several markets shows the effective of rankCRP. The study sheds light on the relationship between CRP and non I.I.D. time series.