Quant Optimum 
Constructing a robust hedge fund portfolio always involves the task of finding an optimal asset allocation satisfying the given investment objectives. Portfolio optimization may incorporate a broad range of risk metrics as well as various constrains (leverage, liquidity, category limits and so on). Going far beyond the conventional CAPM optimization framework, Quant delivers industry’s genetic optimization system for hedge fund portfolios.
StructureTypical Tasks
Features
Why It MattersAn often asked question is: "Why does Quant Optimum include genetic algorithms traditionally used for the NASA research type projects? What is wrong with common Excelbased optimization?" The answer derives from nonnormality of hedge fund distributions of returns.
Since the meanvariance theory and, consequently, the wellknown quadratic optimization methods, are hardly applicable to hedge funds, we have to optimize the VaRbased objective functions better adapting distribution nonnormalities. However, because the VaR presents a nonconvex multiextreme function, the simplest optimization routines, widely used for classic portfolio optimization, become inappropriate (click here for the explanation). The genetic optimization algorithms, in turn, present one of the best known solutions for multiextreme optimization. Another approach to avoid the complexity of multiextreme optimization is to use the alternative VaR metrics like MVaR or CVaR, which effectively leads to simple local optimization. However, these metrics are not free from drawbacks (see the Knowledgebase for the details). Addressing the explained problems, Quant Optimum delivers both the global VaR optimization framework and the quadratic solutions based on the alternative VaR metrics. Objective Functions (examples)
