Portfolio Buy and Sell Decisions
The model portfolio is constructed from the securities that are ranked in the top 10% of our quantitative model at the time of purchase and viewed favorably by our fundamental analysts. When a holding, regardless of absolute gain or loss, drops in quantitative ranking below the 40th percentile, it becomes a candidate for replacement. These buy and sell thresholds have been in place since the inception of our strategies and are not arbitrary. Our research indicates that the maximum return to our clients, after commissions and execution costs, is achieved at these thresholds.
Throughout the history of our composite, our allocation to economic sectors has been very stable. Our process adds alpha through stock selection, rather than sector “bets” against the benchmark. Thus, we have established a target number of positions for each of the sectors included in our universe. Our target for each sector is designed to create a weight that is in proportion to the weighting of the equivalent benchmark sector. In our Large Cap Concentrated portfolio we focus on the number of positions in each particular benchmark sector as opposed to its market cap weighting. This methodology has resulted in relatively stable and highly predictable sector exposures over time.
All client portfolios are managed versus a “model” portfolio that reflects the buy and sell decisions of our investment team. The professionals on our Trading desk have discretion to place orders in the location and manner that maximizes the speed at which we obtain positions, while minimizing total execution costs.
Using analytical tools that are distinct from those of portfolio management, the risk management process assesses the risks embedded in the strategy. We closely monitor the levels and trends of fundamental factor exposures, economic sector exposures, and the composition and magnitude of residual risk versus various benchmarks and indices. This risk management process has resulted in stable sector and style exposures, and a realized beta significantly lower than the market.