This article investigated the implications of distribution free investment strategies on constructing portfolio of individual hedge funds. The author proposes a dynamic, performance-adaptive asset allocation model that allows to optimally diversify across multiple hedge funds styles. The approach to be followed is related to the adaptive allocation of resources between elementary concurrent from the perspective of the theory of sequential investment strategies. The methodological frame gives up the common global stationary hypothesis and approximates locally, a nonstationary paradigm. The approach is then evaluated in a multivariate basis, by examining the performances of several time evolving portfolio strategies in a sample of 16 funds. Empirical experiments are conducted across 5 different hedge fund categories as classified by the Hedge Funds Research and Barclay CTAs databases. We find that the dynamic performance-adaptive allocation strategies amongst hedge funds yield superior annualized average performances, compared to various alternative benchmarks. These findings are robust to different hedge fund restriction provisions such as, lockup and redemption periods.