Strategies:
| Strategy | Benchmark |
| Large Cap Value Quantitative Equity |
Russell 1000 Large Cap Value Index |
| Large Cap Core Quantitative Equity |
Russell 1000 Large Cap Index |
| Large Cap Growth Equity |
Russell 1000 Large Cap Growth Index |
| Mid Cap Core Quantitative Equity |
Russell 800 Mid Cap Index |
Target Client:
Tax-Exempt and Low Tax Bracket Investors.
Investment Objective:
To generate a total return, net of investment management and transaction costs, in excess of the relevant benchmark index over rolling three to five year periods.
|
|
Investment Philosophy:
- We believe that behavioral and fundamental economic, industry and company factors account for the majority of short-term stock price movements relevant to a benchmark (alpha).
- Our research attempts to identify those primary factors that have a consistent, causal relationship with alpha (both positive and negative).
- We construct portfolios of stocks that have positive and negative exposures to these primary factors in order to consistently generate positive alpha. We also seek to control our tracking error and to maximize our information ratio (risk adjusted excess return) and our batting average (an alpha generating consistency measure).
- In order to consistently maximize short-term alpha, net of management and transaction fees, we employ strategies that result in annual portfolio turnover in excess of 100%. Accordingly, these strategies are not suited for taxable investors.
- We use short-term and long-term fundamental sources of alpha. Our dual alpha forecasting process produces more consistent and less volatile alpha.
- Alpha is good, but risk-adjusted alpha is better. We attempt to minimize unintended fundamental exposures and maximize our information ratio.
- Obtaining high batting and information ratios requires timely trading. Poor execution reduces alpha.
- Unexplainable performance reflects unintended fundamental exposures. Mean-reversion is a powerful reality.
|
Investment Process:
Define Investable Universe
- Size: Our large, mid and small capitalization universes are dynamically determined based on standardized scores relative to an average market capitalization. This approach is more robust over time than period-sensitive “hard dollar” boundaries.
- Liquidity: Low priced and illiquid companies are screened-out.
- Quality: Companies that display significant variations between cash and accrued financials relative to their industry peers are deemed to be of lower quality and are screened out.
Identify Industry Specific Sources of Alpha
- Industry specific factors are behavioral and fundamental company characteristics that have consistently explained, both fundamentally and statistically, equity returns over a meaningful historical time period. Utilizing historical factor returns and cross-correlations, we employ a complementary mix of factors that, based on inherent causal relationships, are able to consistently generate alpha.
Create Industry Specific Alpha Forecasts
- Our industry-specific models identify and exploit the most pertinent information from each of the industry specific alpha forecasts to create a robust, final alpha forecast.
Portfolio Construction
- Our core portfolios are diversified across fifty-five industries, are industry-neutral weighted, and run tracking errors ranging from 3% to 6%.
- Turnover exceeds 100% per year. Achieving high monthly batting and information ratios necessitates high turnover rates.
Trading Execution
- We minimize trading and market impact costs by using algorithmic trading programs. Algorithmic trading reduces market-impact costs, has lower commissions, and facilitates trading numerous orders simultaneously.
- Post-trade analysis enables us to evaluate our trade execution and that of our brokers. This continuous next-day analysis provides us with a better understanding of real-world frictional costs and opportunities for improvement.
Performance Attribution and Risk Management
- Performance is closely monitored by reviewing daily performance and factor attribution reports to ensure our portfolios are behaving as intended.
|