Flexible Asset Allocation (FAA)
Developed by Keller & Van Putten · Generalized Momentum · Med Risk
Flexible Asset Allocation was published by Wouter Keller and Hugo van Putten in their 2012 SSRN paper (#2193735), introducing a multi-factor ranking system that expanded the concept of momentum-based asset selection beyond pure price returns. While earlier strategies ranked assets solely by trailing return — implicitly assuming that the asset with the highest past return has the highest expected future return — FAA recognized that return momentum is only one dimension of what makes an asset attractive. Volatility and correlation with other holdings also carry meaningful information about an asset's forward-looking risk-reward profile.
Keller and Van Putten's composite scoring function combines three factors: trailing return (momentum), trailing volatility, and trailing correlation with the other assets in the universe. Each factor receives a configurable weight, with return receiving the highest emphasis, volatility the second, and correlation the third. Assets with high returns, low volatility, and low correlation with other candidates score highest — a natural preference for assets that are trending well, doing so with contained risk, and providing diversification benefits to the overall portfolio.
The strategy applies this composite ranking to a seven-asset universe covering US equities (VTI), international developed (VEA), emerging markets (VWO), commodities (GSG), real estate (VNQ), aggregate bonds (BND), and short-term Treasuries (SHY). The top three by composite score receive equal weight, with an absolute momentum filter that removes any selected asset whose four-month return is negative. This short four-month lookback for both ranking and filtering makes FAA substantially more responsive than strategies using twelve-month returns, at the cost of higher turnover and greater sensitivity to short-term market noise.
FAA represents an early attempt at what would later be called multi-factor tactical allocation — using information beyond price momentum to improve the quality of asset selection in systematic portfolios. Its influence can be seen in subsequent strategies like EAA, which took the multi-factor concept further by introducing more sophisticated mathematical formulations for combining return, volatility, and correlation information.
How It Works
Multi-Factor Composite Scoring
Each month, all seven assets are evaluated across three factors. First, trailing four-month return provides the primary momentum signal — assets with stronger recent returns receive higher return rank scores. Second, trailing four-month volatility (measured as standard deviation of daily returns) provides a risk adjustment — assets with lower volatility receive higher stability rank scores. Third, trailing four-month average correlation with all other assets in the universe provides a diversification signal — assets with lower correlation to the field receive higher diversification rank scores.
These three rank scores are combined using configurable weights: 1.0 for return rank, 0.8 for volatility rank, and 0.6 for correlation rank. The weighting hierarchy ensures that momentum remains the primary selection criterion while volatility and correlation provide meaningful secondary adjustments. An asset with moderate returns but exceptionally low volatility and correlation can outrank a higher-returning asset that is volatile and highly correlated with other candidates.
Selection and Absolute Momentum Filter
The top three assets by composite score are selected for equal-weight allocation. Before receiving their allocation, each selected asset must pass an absolute momentum check: its trailing four-month return must be positive. Assets that fail this test have their allocation redirected to the fallback position (SHY), reducing the portfolio's risk exposure proportionally to how many of the selected assets are in negative momentum territory.
The four-month lookback for both the composite scoring and the absolute filter is notably shorter than the twelve-month standard used by most tactical strategies. This shorter horizon makes FAA more responsive to recent market shifts but also more sensitive to short-term noise. During choppy markets, the four-month returns can flip between positive and negative more frequently than twelve-month returns, producing higher turnover as assets cycle in and out of the portfolio.
The Multi-Factor Advantage
The inclusion of volatility and correlation in the ranking addresses a known weakness of pure momentum selection: momentum strategies can concentrate the portfolio in highly correlated, high-volatility assets during trending markets, creating portfolios that look diversified on the surface but carry concentrated factor risk. By penalizing volatile and correlated assets in the ranking, FAA produces portfolios with lower realized risk than equivalent momentum-only approaches.
During market stress, when correlations between risk assets typically spike, the correlation factor naturally penalizes assets that are becoming more correlated with each other, biasing the selection toward assets that retain their diversification properties. This correlation-aware ranking provides a built-in risk management function that operates independently of the absolute momentum filter, creating a secondary layer of protection that emerges from the selection process itself.
Explore Flexible Asset Allocation (FAA)
See the full backtest across 18 years of market data, or run your own what-if scenarios by adjusting all research parameters.