“Efficient Portfolio Management”: A Guide To Effective Investment Supervision Processes
In practice, it is essential to take an integrated approach to clients’ wealth management and investment supervision activities. The integration is based upon expressly recognizing the three major components of this process: wealth planning, investment supervision, and portfolio accounting and reporting. Each component is now examined in turn.
I The Investment Process: Wealth Planning
Although often a convenient starting point for any investment relationship, the planning phase is really a continuous and dynamic process that requires ongoing review and refinement as client needs and objectives change. It is during this phase that investment supervisors assist their clients in a determination of client goals, objectives, and constraints. The importance of this process should not be underestimated. Proper planning develops the inputs upon which all other portfolio management decisions will rely.
Specifically, investment supervisors must be interested in learning about their client’s views and preferences regarding the risk to be taken. These include determining the degree of tolerance for portfolio drawdown (peak-to-trough intervals of loss) and conditional value-at-risk (point-estimate maximal loss).
Other metrics of interest often include the degree of aversion to skew (exposure to directional market bias) and kurtosis (exposure to extreme events). The purpose of these inquiries is to closely discern the degree of client risk tolerance (where risk manifests itself in multiple dimensions).
This concentrated focus on the specific articulation of risk tolerance differentiates “best practice” approaches and is a critical input to the portfolio management and supervision process outlined below.
Effective investment supervisors also assess the organization of their client’s investment-level entities in order to determine how such goals and objectives are shared across different constituents. Other planning activities, including tax optimization (and related entity allocations), income and expense modeling, and asset and liability matching must also be conducted in order to more fully appreciate the requirements of the investment portfolio.
The initial goal of this process is to help clients formulate a comprehensive investment policy statement, including the articulation of risk tolerance, which will ultimately guide the asset allocation and portfolio selection decisions.
II The Investment Process: Portfolio Management and Supervision
Once an investment policy statement has been articulated, it must be translated into a consistent and actionable portfolio management process. Ultimately, a diversified portfolio of investments in the “traditional” (e.g., cash, fixed income, equities), and “alternative” (e.g., hedge funds, private equity, real estate) asset classes is built through a deep and intensive due diligence process.
Numerous academic and empirical studies confirm the philosophy that the asset allocation decision is one of the primary determinants of portfolio return variance—both across time and across investment managers (For example, Brinson, et. al., 1991 and Ibbotson and Kaplan, 2000). Security selection and market timing, though also influential, are generally secondary considerations that have marginal contributions relative to the overall asset allocation.
Effective asset allocation optimizes the power of diversification and ensures that an investment portfolio maximizes the return generated for a given level of risk. As a result, getting the “top-down” decision correct is of critical importance.
Unfortunately, applications of the technique by many consultants, investment managers, and brokers often fail to account for varying degrees of market efficiency, skew (bias or event risk), and kurtosis (extreme events) often present among different asset classes. Relying solely upon returns and standard deviations can lead to sub-optimal conclusions, especially since neither vector is particularly robust with respect to time. In addition, the popularly used linear mean-variance optimization models tend to produce results that produce inherently unstable and grossly imbalanced portfolios that are highly sensitive to estimation error (Michaud, 1989).
Such narrow reliance among investment consultants and money managers is not uncommon; however, it often leads to an overestimation of expected return or an underestimation of risk and is the root cause of many forms of investor disappointment. It is also a causal factor in many of the very notable fund “blow ups” that have been witnessed in recent times.
An “Efficient Portfolio Management (EPM)” Methodology
The EPM methodology explicitly addresses such shortcomings by taking on the issue of risk directly. The goals of the EPM are to:
Explicitly determine how much risk to take and what forms of such risk are acceptable; and
Ensure that the portfolio is maximally compensated for taking on those risks.
Under the EPM framework, investment exposure (the sources of risks and, correspondingly, returns) comes in two forms, namely: systematic and nonsystematic. The EPM process is a synthesis of strategic asset allocation (for the purposes of determining systematic allocations) and active risk budgeting (for the purposes of determining nonsystematic allocations).
A brief discussion of each follows below.
Allocations to Systematic Investment Assets
Systematic investment assets typically relate to broad, well-diversified market exposures (also generically referred to as “beta”). In order to determine the optimal systematic exposure to various asset classes, EPM relies upon the process of strategic asset allocation.
The process begins by identifying the investment universe of candidate assets. These may include, for instance, the following classes (and their related subsets): cash, fixed income, equities, commodities, and real estate. In order to establish the initial (neutral) portfolio weights, EPM draws upon the general findings of the Capital Asset Pricing Model (CAPM), as described by Sharpe (1964) and Lintner (1965).
According to the CAPM equilibrium, the optimal asset class weights are directly related to the relative aggregate market capitalizations of each such asset class. That is, in equilibrium, the portfolio weights for the “market portfolio” in aggregate also become the optimal weights for individual portfolios. EPM refers to these weights as the “Equilibrium Portfolio,” and it serves as the initial “center of gravity” for the strategic asset allocation process.
As prescribed by Black and Litterman (1992), EPM next “reverse-optimizes” the Equilibrium Portfolio to ascertain its implied views of asset class risk premiums. Stated differently, if one believes that the Equilibrium Portfolio is in fact the optimal portfolio (as suggested by CAPM), what must the views on relative asset class risk premiums be in order to satisfy mean-variance optimality conditions?
The analysis of such implied views is analogous to the more popular application of using implied volatility to evaluate options pricing. In that approach, it is assumed that the market price represents both known and estimated variables that affect valuation. However, given the difficulty of objectively estimating expected volatility, many investors consequently take the market price as a given and instead determine what it must imply about expected volatility. Then, investors can make individual judgments about whether this implied volatility is in fact reasonable.
Similarly, investment supervisors can compare the implied views of the Equilibrium Portfolio with their own views of relative asset class premiums. Importantly, under this approach, it is not necessary for investment supervisors to make forecasts of the absolute returns of every asset class; they need only make relative risk premium assessments.
Where differences emerge between investment supervisor (or client) views and the market-implied views, EPM combines them through a conditional, Bayesian-weighted adjustment in order to capture the corresponding degree-of-confidence in each view. The adjustment may also affect other asset classes, as the process attempts to maintain the consistency of the covariance relationships among such classes. The resulting portfolio weights form a “passive risk portfolio.” This approach tends to minimize the forecast errors and unrealistic portfolio tilts that often plague standard mean-variance optimization.
Since the passive risk portfolio primarily addresses broad market, systematic (beta) exposure, its implementation is best conducted through low-cost, tax-efficient passive investment instruments. Such composition for broad market systematic exposure is consistent not only with CAPM, but also with the general findings of Fama (1970) and Brinson, et. al. (1991).
Allocations to Nonsystematic Investment Assets
Nonsystematic investment assets generally relate to idiosyncratic exposures (also referred to as “alpha”) that are not correlated to the systematic universe. In equilibrium, and relying upon the principles of diversification, such idiosyncratic exposure should not be expected to produce excess return. Therefore, expected alpha return in equilibrium is equal to zero.
In reality, however, the markets are often not in a state of CAPM equilibrium, especially in the short to medium term. As a result, it