• To maximize the risk-adjusted total return of a diversified portfolio of hedge fund sub-indices by actively overweighting and underweighting the components of the strategic portfolio

• Innovative Solution: Alpha to be generated over and above Credit Suisse/Tremont investable hedge-fund index based on statistical evidence of the predictability in hedge fund sub-index returns
• Transparent Quantitative Process: A disciplined approach based on robust quantitative processes from initial long-term derivation of long-term optimal weights to the derivation of tactical alpha tilts
• Optimal Portfolio Construction: Equilibrium portfolio weights to be devised so as to meet investor’s objectives i.e. low correlation to traditional asset classes, maximum expected information ratio, low drawdown risks, or a combination of non-mutually exclusive objectives and constraints.
• Market and economic intuition: Signals driving alpha tilts to be based on systematic risk factors, consistent with market and economic intuition

• Application of correction techniques to overcome biases associated with hedge fund sub-indices
• Non-linear dependence techniques to be applied to generate optimal equilibrium hedge fund basket weights (Rank Correlations, Copulas)
• Equilibrium portfolio allocations to be derived incorporating several predefined risk criteria and to be assessed with multiple methodologies (Reverse Mean-Variance optimization, Value at Risk (VaR) and modified VaR, Copula/Extreme Value Theory optimizations)
• Optimal allocation to include all predefined investor specifications and risk preferences

2. Tactical views and alpha generation: An innovative and transparent approach to focus on higher risk-adjusted opportunities

• Consistent incorporation of multiple dimensions of information including significant exogenous indicators and systematic risk factors
• Parsimonious number of signals are applied to all relevant sub hedge-fund indices
• Signals’ sensitivities to be dynamically adjusted according to their explanatory power
• Direction and confidence of resulting tactical views expressed in expected mean and standard error

• Tactical views to be integrated at the final portfolio optimization step, where strategic and tactical views are to be apportioned according to model derived associated confidence levels
• Final optimization: To include all investor preferences (orthogonality to traditional asset classes, long-only and no leverage restrictions).
• Flexible control of position turnover via a transaction penalty parameter
• Risks associated with final solution to be consistent with following predefined bounds: maximum acceptable loss assessed via stress-tests, VaR, mVaR, and EVT parameters

• Initial objectives and constraints
• The objective is to consistently out-perform the benchmark portfolio while staying within the prescribed risk limits 
• No short selling or leverage is permitted in the overall portfolio
  

• Coverage includes CS/Tremont’s 10 main hedge fund indices:
• Convertible Arbitrage Index 
• Dedicated Short Bias Index 
• Emerging Market Index 
• Equity Market Neutral Index 
• Event Driven Index 
• Fixed Income Arbitrage Index 
• Global Macro Index 
• Long/Short Equity Index 
• Managed Futures Index 
• Multi-Strategy Index
  

• The Alpha² model selects from 95 exogenous signals:
• Approach based on statistically sounds methodologies
• Exogenous signals and systematic risk factors are consistent with market and economic intuition
• Capability to expand the approach to additional hedge fund sub-sector indices
  

• The Alpha² model offers an innovative and transparent approach to focus on profitable relative trades, allowing consistent incorporation of multiple dimensions of information advantage into the portfolio
• At the individual hedge fund sub-sector strategy level: 
• A parsimonious number of signals are chosen 
• Weights of signals are dynamically adjusted 
• Direction and confidence of profitable investment views are derived 
• Across hedge fund sub-sector strategies: 
• Strategic equilibrium weights and tactical investment views are blended through a Bayesian framework based on their relative information advantage 
• Expected returns for all hedge fund sub-sectors are derived alongside their associated confidence levels
  
• Portfolio construction: 
• Tactical over/under-weights are determined based on return and risk tradeoff, given the benchmark, constraints, turnover and transaction costs considerations