Investors want alpha that is: 1) positive, 2) reliable and 3) preferably uncorrelated to underlying benchmarks and other active strategies. Pure alpha must be achieved without any persistent risk-factor tilts, such as beta above one or a small size or value tilt vs. benchmark. Investors know that producing such alpha is challenging and that it takes specialized skills to achieve. Thus, they expect rigor in the investment process and transparency in the objectives, including targets for the alpha and risks (tracking error) of the active strategy overall and each segment or layer of the strategy. Knowing how much alpha and risk should come from stock picking vs. sector tilts or market timing, or anything else, helps investors build a portfolio of active strategies.
Alpha is more valuable when reliable, so sophisticated investors focus on the information ratio. The information ratio is alpha divided by tracking error. Thus, it gauges alpha vs. reliability as tracking error is the standard deviation or reliability of alpha. An information ratio above the long-term sharpe ratio of the underlying benchmark is very valuable. Note that a low tracking error doesn't mean that alpha is normally low. It just indicates the active risk via alpha volatility.
We think active equity strategies should be based on fundamental analysis of macro and micro issues that determine value and influence total-return potential in a conceptually robust way. We do not rely on technical or just empirical analysis. We believe that fundamental analysis is most effective and efficient when implemented in a well-organized research framework and decision-making process. A process that includes quantitative tools that organize data to help identify and forecast the drivers of value. Such quantitative tools include many information databases and metrics organized for screening purposes, factor-based ranking, discounted-cash-flow (DCF) models, etc.
Every deviation from benchmark should be for a well-justified fundamental reason that aligns with where that deviation occurs from top to bottom. Beta deviations for macro reasons and relative asset-class preferences, sector deviations for sector-level reasons and relative sector preferences, and securities-weight deviations for micro and relative stock-specific preferences. Every overweight comes with an underweight, so both sides of any deviation must be justified.
Micro reasons include industry trends and company-specific operating and financial forecasts. Macro reasons include views on the economy, inflation, interest rates, currencies, commodities, capital expenditure (capex) and credit conditions, etc. Macro considerations, including the relative valuations of other asset classes and key asset-class sub-segments, tend to determine desired asset-class or beta exposure and also equity-sector and other macro factor tilts in the portfolio. Micro considerations, including the relative valuation of securities, tend to determine the stocks selected within sectors and within other macro-factor clusters. We refrain from sector or macro-factor tilts for bottom-up reasons, but this can be justified if the micro assessment of large-weight companies warrants it.
Bottom-up and top-down methods work best in their respective areas of benchmark deviation and will complement each other when they meet in the middle. Understand all risks in a portfolio from both its top-down and bottom-up architecture. Quantitative portfolio-construction tools help ensure that all tilts from the top to the bottom of the benchmark are as intended and within risk budgets.
Both systematic and discretionary active strategies seek alpha vs. benchmarks. Like our discretionary strategies, our systematic strategies are based on fundamental frameworks. Systematic investing can be thought of as the codification of a fundamental portfolio manager's (PM) decision-making process. Systematic strategies can be well-suited for investing in core and value categories, whereas more PM discretion is often required with growth stocks. Systematic strategies are usually better suited for well-diversified portfolios from large eligible investment universes of mature industry stocks with long operating and valuation histories. More concentrated portfolios tend to require more PM discretion. Discretionary strategies are usually better suited with smaller and newer companies with short operating and valuation histories.
Active share is not a reliable indication of likely alpha or even active risk, in our view. We believe 'where different from benchmark' is more important than 'how much different'. Different in the right places counts. Active management should be selective and deliberate in its deviations from benchmark. The two ways to raise active share are to underweight big-benchmark-weight stocks or to be concentrated. But underweighting big-weight stocks must be fundamentally justified, not automatic. Concentration is a two-edged sword. It also usually adds volatility to performance and makes it harder for investors to distinguish luck from skill. It makes sense to budget tilt sizes vs. benchmark weights on conviction, but it is wise to balance concentration and contrarianism.