,

CHAPTER 7

BASICS OF PORTFOLIO PLANNING AND CONSTRUCTION

Alistair Byrne, CFA

Edinburgh, U.K.

Frank E. Smudde, CFA

Amsterdam, The Netherlands

LEARNING OUTCOMES

After completing this chapter, you will be able to do the following:

• Explain the reasons for a written investment policy statement (IPS).
• List and explain the major components of an IPS.
• Discuss risk and return objectives, including their preparation.
• Distinguish between the willingness and the ability (capacity) to take risk in analyzing an investor’s financial risk tolerance.
• Describe the investment constraints of liquidity, time horizon, tax concerns, legal and regulatory factors, and unique circumstances and their implications for the choice of portfolio assets.
• Explain the definition and specification of asset classes in relation to asset allocation.
• Discuss the principles of portfolio construction and the role of asset allocation in relation to the IPS.

1. INTRODUCTION

To build a suitable portfolio for a client, investment advisers should first seek to understand the client’s investment goals, resources, circumstances, and constraints. Investors can be categorized into broad groups based on shared characteristics with respect to these factors (e.g., various types of individual investors and institutional investors). Even investors within a given type, however, will invariably have a number of distinctive requirements. In this chapter, we consider in detail the planning for investment success based on an individualized understanding of the client.

This chapter is organized as follows: Section 2 discusses the investment policy statement, a written document that captures the client’s investment objectives and the constraints. Section 3 discusses the portfolio construction process, including the first step of specifying a strategic asset allocation for the client. Section 4 concludes and summarizes the chapter.

2. PORTFOLIO PLANNING

Portfolio planning can be defined as a program developed in advance of constructing a portfolio that is expected to satisfy the client’s investment objectives. The written document governing this process is the investment policy statement (IPS).

2.1. The Investment Policy Statement

The IPS is the starting point of the portfolio management process. Without a full understanding of the client’s situation and requirements, it is unlikely that successful results will be achieved. “Success” can be defined as a client achieving his or her important investment goals using means that he or she is comfortable with (in terms of risks taken and other concerns). The IPS essentially communicates a plan for achieving investment success.

The IPS will be developed following a fact-finding discussion with the client. This fact-finding discussion can include the use of a questionnaire designed to articulate the client’s risk tolerance as well as specific circumstances. In the case of institutional clients, the fact-finding may involve asset–liability management studies, identification of liquidity needs, and a wide range of tax and legal considerations.

The IPS can take a variety of forms. A typical format will include the client’s investment objectives and the constraints that apply to the client’s portfolio.

The client’s objectives are specified in terms of risk tolerance and return requirements. These must be consistent with each other: A client is unlikely to be able to find a portfolio that offers a relatively high expected return without taking on a relatively high level of risk.

The constraints section covers factors that need to be taken into account when constructing a portfolio for the client that meets the objectives. The typical constraint categories are liquidity requirements, time horizon, regulatory requirements, tax status, and unique needs. The constraints may be internal (i.e., set by the client), or external (i.e., set by law or regulation). These are discussed in detail next.

Having a well-constructed IPS for all clients should be standard procedure for a portfolio manager. The portfolio manager should have the IPS close at hand and be able to refer to it to assess the suitability of a particular investment for the client. In some cases, the need for the IPS goes beyond simply being a matter of standard procedure. In some countries, the IPS (or an equivalent document) is a legal or regulatory requirement. For example, U.K. pension schemes must have a statement of investment principles under the Pensions Act 1995 (Section 35), and this statement is in essence an IPS. The U.K. Financial Services Authority also has requirements for investment firms to “know their customers.” The European Union’s Markets in Financial Instruments Directive (“MiFID”) requires firms to assign clients to categories, such as professional clients and retail clients.

In the case of an institution, such as a pension plan or university endowment, the IPS may set out the governance arrangements that apply to the investment funds. For example, this information could cover the investment committee’s approach to appointing and reviewing investment managers for the portfolio, and the discretion that those managers have. The IPS could also set out the institution’s approach to corporate governance, in terms of how it will approach the use of shareholder voting rights and other forms of engagement with corporate management.

The IPS should be reviewed on a regular basis to ensure that it remains consistent with the client’s circumstances and requirements. For example, the U.K. Pensions Regulator suggests that a pension scheme’s statements of investment principles—a form of IPS—should be reviewed at least every three years. The IPS should also be reviewed if the manager becomes aware of a material change in the client’s circumstances, or on the initiative of the client when his or her objectives, time horizon, or liquidity needs change.

2.2. Major Components of an IPS

There is no single standard format for an IPS. Many IPSs, however, include the following sections:

• Introduction. This section describes the client.
• Statement of Purpose. This section states the purpose of the IPS.
• Statement of Duties and Responsibilities. This section details the duties and responsibilities of the client, the custodian of the client’s assets, and the investment managers.
• Procedures. This section explains the steps to take to keep the IPS current and the procedures to follow to respond to various contingencies.
• Investment Objectives. This section explains the client’s objectives in investing.
• Investment Constraints. This section presents the factors that constrain the client in seeking to achieve the investment objectives.
• Investment Guidelines. This section provides information about how policy should be executed (e.g., on the permissible use of leverage and derivatives) and on specific types of assets excluded from investment, if any.
• Evaluation and Review. This section provides guidance on obtaining feedback on investment results.
• Appendices: (A) Strategic Asset Allocation (B) Rebalancing Policy. Many investors specify a strategic asset allocation (SAA), also known as the policy portfolio, which is the baseline allocation of portfolio assets to asset classes in view of the investor’s investment objectives and the investor’s policy with respect to rebalancing asset class weights.

The sections that are most closely linked to the client’s distinctive needs, and probably the most important from a planning perspective, are those dealing with investment objectives and constraints. An IPS focusing on these two elements has been called an IPS in an “objectives and constraints” format.

In the following sections, we discuss the investment objectives and constraints format of an IPS beginning with risk and return objectives. We follow a tradition of CFA Institute presentations in discussing risk objectives first. The process of developing the IPS is the basic mechanism for evaluating and trying to improve an investor’s overall expected return–risk stance. In a portfolio context, “investors have learned to appreciate that their objective is not to manage reward but to control and manage risk.”1 Stated another way, return objectives and expectations must be tailored to be consistent with risk objectives. The risk and return objectives must also be consistent with the constraints that apply to the portfolio.

2.2.1. Risk Objectives

When constructing a portfolio for a client, it is important to ensure that the risk of the portfolio is suitable for the client. The IPS should state clearly the risk tolerance of the client.

Risk objectives are specifications for portfolio risk that reflect the risk tolerance of the client. Quantitative risk objectives can be absolute or relative or a combination of the two.

Examples of an absolute risk objective would be a desire not to suffer any loss of capital or not to lose more than a given percent of capital in any 12-month period. Note that these objectives are not related to investment market performance, good or bad, and are absolute in the sense of being self-standing. The fulfillment of such objectives could be achieved by not taking any risk; for example, by investing in an insured bank certificate of deposit at a creditworthy bank. If investments in risky assets are undertaken, however, such statements would need to be restated as a probability statement to be operational (i.e., practically useful). For example, the desire to not lose more than 4 percent of capital in any 12-month period might be restated as an objective that with 95 percent probability the portfolio not lose more than 4 percent in any 12-month period. Measures of absolute risk include the variance or standard deviation of returns and value at risk.2

Some clients may choose to express relative risk objectives, which relate risk relative to one or more benchmarks perceived to represent appropriate risk standards. For example, investments in large-cap U.K. equities could be benchmarked to an equity market index, such as the FTSE 100 Index. The S&P 500 Index could be used as a benchmark for large-cap U.S. equities, or, for investments with cash-like characteristics, the benchmark could be an interest rate such as LIBOR or a Treasury bill rate. For risk relative to a benchmark, the relevant measure is tracking risk, or tracking error.3

For institutional clients, the benchmark may be linked to some form of liability the institution has. For example, a pension plan must meet the pension payments as they come due and the risk objective will be to minimize the probability that it will fail to do so. (A related return objective might be to outperform the discount rate used in finding the present value of liabilities over a multiyear time horizon.)

When a policy portfolio (that is, a specified set of long-term asset class weightings) is used, the risk objective may be expressed as a desire for the portfolio return to be within a band of plus or minus X percent of the benchmark return calculated by assigning an index or benchmark to represent each asset class present in the policy portfolio. Again, this objective has to be interpreted as a statement of probability; for example, a 95 percent probability that the portfolio return will be within X percent of the benchmark return over a stated time period. Example 7-1 reviews this material.

A client’s overall risk tolerance is a function of the client’s ability to bear (accept) risk and his or her “risk attitude,” which might be considered as the client’s willingness to take risk. For ease of expression, from this point on we will refer to ability to bear risk and willingness to take risk as the two components of risk tolerance. Above-average ability to bear risk and above-average willingness to take risk imply above-average risk tolerance. Below-average ability to bear risk and below-average willingness to take risk imply below-average risk tolerance.

The ability to bear risk is measured mainly in terms of objective factors, such as time horizon, expected income, and the level of wealth relative to liabilities. For example, an investor with a 20-year time horizon can be considered to have a greater ability to bear risk, other things being equal, than an investor with a two-year horizon. This difference is because over 20 years there is more scope for losses to be recovered or other adjustments to circumstances to be made than there is over two years.

Similarly, an investor whose assets are comfortably in excess of their liabilities has more ability to bear risk than an investor whose wealth and expected future expenditure are more closely balanced. For example, a wealthy individual who can sustain a comfortable lifestyle after a very substantial investment loss has a relatively high ability to bear risk. A pension plan that has a large surplus of assets over liabilities has a relatively high ability to bear risk.

Risk attitude, or willingness to take risk, is a more subjective factor based on the client’s psychology and perhaps also his or her current circumstances. Although the list of factors that are related to an individual’s risk attitude remains open to debate, it is believed that some psychological factors, such as personality type, self-esteem, and inclination to independent thinking, are correlated with risk attitude. Some individuals are comfortable taking financial and investment risk, whereas others find it distressing. Although there is no single agreed-upon method for measuring risk tolerance, a willingness to take risk may be gauged by discussing risk with the client or by asking the client to complete a psychometric questionnaire. For example, financial planning academic John Grable and collaborators have developed 13-item and 5-item risk attitude questionnaires that have undergone some level of technical validation. The five-item questionnaire is shown in Exhibit 7-1.

EXHIBIT 7-1 A Five-Item Risk Assessment Instrument

1. Investing is too difficult to understand.

A. Strongly agree

B. Tend to agree

C. Tend to disagree

D. Strongly disagree

2. I am more comfortable putting my money in a bank account than in the stock market.

A. Strongly agree

B. Tend to agree

C. Tend to disagree

D. Strongly disagree

3. When I think of the word “risk” the term “loss” comes to mind immediately.

A. Strongly agree

B. Tend to agree

C. Tend to disagree

D. Strongly disagree

4. Making money in stocks and bonds is based on luck.

A. Strongly agree

B. Tend to agree

C. Tend to disagree

D. Strongly disagree

5. In terms of investing, safety is more important than returns.

A. Strongly agree

B. Tend to agree

C. Tend to disagree

D. Strongly disagree

Source: Grable and Joo (2004).

The responses, (A), (B), (C), and (D), are coded 1, 2, 3, and 4, respectively, and summed. The lowest score is 5 and the highest score is 20, with higher scores indicating greater risk tolerance. For two random samples drawn from the faculty and staff of large U.S. universities (n = 406), the mean score was 12.86 with a standard deviation of 3.01 and a median (i.e., most frequently observed) score of 13.

Note that a question, such as the first one in Exhibit 7-1, indicates that risk attitude may be associated with nonpsychological factors (such as level of financial knowledge and understanding and decision-making style) as well as psychological factors.

The adviser needs to examine whether a client’s ability to accept risk is consistent with the client’s willingness to take risk. For example, a wealthy investor with a 20-year time horizon, who is thus able to take risk, may also be comfortable taking risk; in this case the factors are consistent. If the wealthy investor has a low willingness to take risk, there would be a conflict.

In the institutional context, there could also be conflict between ability and willingness to take risk. In addition, different stakeholders within the institution may take different views. For example, the trustees of a well-funded pension plan may desire a low-risk approach to safeguard the funding of the scheme and beneficiaries of the scheme may take a similar view. The sponsor, however, may wish a higher-risk/higher-return approach in an attempt to reduce future funding costs. When a trustee bears a fiduciary responsibility to pension beneficiaries and the interests of the pension sponsor and the pension beneficiaries conflict, the trustee should act in the best interests of the beneficiaries.

When ability to take risk and willingness to take risk are consistent, the investment adviser’s task is the simplest. When ability to take risk is below average and willingness to take risk is above average, the investor’s risk tolerance should be assessed as below average overall. When ability to take risk is above average but willingness is below average, the portfolio manager or adviser may seek to counsel the client and explain the conflict and its implications. For example, the adviser could outline the reasons why the client is considered to have a high ability to take risk and explain the likely consequences, in terms of reduced expected return, of not taking risk. The investment adviser, however, should not aim to change a client’s willingness to take risk that is not a result of a miscalculation or misperception. Modification of elements of personality is not within the purview of the investment adviser’s role. The prudent approach is to reach a conclusion about risk tolerance consistent with the lower of the two factors (ability and willingness) and to document the decisions made.

Example 7-2 is the first of a set that follows the analysis of a private wealth management client through the preparation of the major elements of an IPS.

2.2.2. Return Objectives

A client’s return objectives can be stated in a number of ways. Similar to risk objectives, return objectives may be stated on an absolute or a relative basis.

As an example of an absolute objective, the client may want to achieve a particular percentage rate of return, for example, X percent. This could be a nominal rate of return or be expressed in real (inflation-adjusted) terms.

Alternatively, the return objective can be stated on a relative basis, for example, relative to a benchmark return. The benchmark could be an equity market index, such as the S&P 500 or the FTSE 100, or a cash rate of interest such as LIBOR. LIBOR might be appropriate when the investor has some liability that is linked to that rate; for example, a bank that has a particular cost of funding linked to LIBOR. A relative return objective might be stated as, for example, a desire to outperform the benchmark index by one percentage point per year.

Some institutions also set their return objective relative to a peer group or universe of managers; for example, an endowment aiming for a return that is in the top 50 percent of returns of similar institutions, or a private equity mandate aiming for returns in the top quartile among the private equity universe. This objective can be problematic when limited information is known about the investment strategies or the returns calculation methodology being used by peers, and we must bear in mind the impossibility of all institutions being “above average.” Furthermore, a good benchmark should be investable—that is, able to be replicated by the investor—and a peer benchmark typically does not meet that criterion.

In each case, the return requirement can be stated before or after fees. Care should be taken that the fee basis used is clear and understood by both the manager and client. The return can also be stated on either a pre- or post-tax basis when the investor is required to pay tax. For a taxable investor, the baseline is to state and analyze returns on an after-tax basis.

The return objective could be a required return—that is, the amount the investor needs to earn to meet a particular future goal—such as a certain level of retirement income.

The manager or adviser must ensure that the return objective is realistic. Care should be taken that client and manager are in agreement on whether the return objective is nominal (which is more convenient for measurement purposes) or real (i.e., inflation-adjusted, which usually relates better to the objective). It must be consistent with the client’s risk objective (high expected returns are unlikely to be possible without high levels of risk) and also with the current economic and market environment. For example, 15 percent nominal returns might be possible when inflation is 10 percent, but will be unlikely when inflation is 3 percent.

When a client has unrealistic return expectations, the manager or adviser will need to counsel them about what is achievable in the current market environment and within the client’s tolerance for risk.

In the following sections, we analyze five major types of constraints on portfolio selection: liquidity, time horizon, tax concerns, legal and regulatory factors, and unique circumstances.

2.2.3. Liquidity

The IPS should state what the likely requirements are to withdraw funds from the portfolio. Examples for an individual investor would be outlays for covering health care payments or tuition fees. For institutions, it could be spending rules and requirements for endowment funds, the existence of claims coming due in the case of property and casualty insurance, or benefit payments for pension funds and life insurance companies.

When the client does have such a requirement, the manager should allocate part of the portfolio to cover the liability. This part of the portfolio will be invested in assets that are liquid—that is, easily converted to cash—and low risk at the point in time the liquidity need is actually present (e.g., a bond maturing at the time when private education expenses will be incurred), so that their value is known with reasonable certainty. For example, the asset allocation in the insurance portfolios of Finnish insurer Sampo (see Exhibit 7-2) shows a large allocation to fixed-income investments, some of which are either highly liquid or have a short maturity. These investments enable the company, in the case of property and casualty insurance, to pay out on potentially “lumpy” claims of which the timing is unpredictable, and in the case of life insurance, to pay out on life benefits, the size and timing of which are more predictable and can therefore be matched with the maturity profile of the fixed-income portfolio.

2.2.4. Time Horizon

The IPS should state the time horizon over which the investor is investing. It may be the period over which the portfolio is accumulating before any assets need to be withdrawn; it could also be the period until the client’s circumstances are likely to change. For example, a 50-year-old pension plan investor hoping to retire at age 60 has a 10-year horizon. The portfolio may not be liquidated at age 60, but its structure may need to change, for example, as the investor begins to draw an income from the fund.

The time horizon of the investor will affect the nature of investments used in the portfolio. Illiquid or risky investments may be unsuitable for an investor with a short time horizon because the investor may not have enough time to recover from investment losses, for example. Such investments, however, may be suitable for an investor with a longer horizon, especially if the risky investments are expected to have higher returns.

2.2.5. Tax Concerns

Tax status varies among investors. Some investors will be subject to taxation on investment returns and some will not. For example, in many countries returns to pension funds are exempt from tax. Some investors will face various rates of tax on income (dividends and interest payments) than they do on capital gains (associated with increases in asset prices). Typically, when there is a differential, income is taxed more highly than gains. Gains may be subject to a lower rate of tax or part or all of the gain may be exempt from taxation. Furthermore, income may be taxed as it is earned, whereas gains may be taxed when they are realized. Hence, in such cases there is a time value of money benefit in the deferment of taxation of gains relative to income.

In many cases, the portfolio should reflect the tax status of the client. For example, a taxable investor may wish to hold a portfolio that emphasizes capital gains and receives little income. A taxable investor based in the United States is also likely to consider including U.S. municipal bonds (“munis”) in his or her portfolio because interest income from munis, unlike from treasuries and corporate bonds, is exempt from taxes. A tax-exempt investor, such as a pension fund, will be relatively indifferent to the form of returns.

2.2.6. Legal and Regulatory Factors

The IPS should state any legal and regulatory restrictions that constrain how the portfolio is invested.

In some countries, such institutional investors as pension funds are subject to restrictions on the composition of the portfolio. For example, there may be a limit on the proportion of equities or other risky assets in the portfolio, or on the proportion of the portfolio that may be invested overseas. The United States has no limits on pension fund asset allocation but some countries do, examples of which are shown in Exhibit 7-3. Pension funds also often face restrictions on the percentage of assets that can be invested in securities issued by the plan sponsor, so called self-investment limits.

EXHIBIT 7-3 Examples of Pension Fund Investment Restrictions

Source: OECD Survey of Investment Regulations of Pension Funds, July 2008.

When an individual has access to material nonpublic information about a particular security, this situation may also form a constraint. For example, the directors of a public company may need to refrain from trading the company’s stock at certain points of the year before financial results are published. The IPS should note this constraint so that the portfolio manager does not inadvertently trade the stock on the client’s behalf.

2.2.7. Unique Circumstances

This section of the IPS should cover any other aspect of the client’s circumstances that is likely to have a material impact on the composition of the portfolio. A client may have considerations derived from his or her religion or ethical values that could constrain investment choices. For instance, a Muslim investor seeking compliance with Shari’a (the Islamic law) will avoid investing in businesses and financial instruments inconsistent with Shari’a, such as casinos and bonds, because Shari’a prohibits gambling and lending money on interest. Similarly, a Christian investor may wish to avoid investments that he or she believes are inconsistent with their faith.

Whether rooted in religious beliefs or not, a client may have personal objections to certain products (e.g., pornography, weapons, tobacco, gambling) or practices (e.g., environmental impact of business activities, human impact of government policies, labor standards), which could lead to the exclusion of certain companies, countries, or types of securities (e.g., interest-bearing debt) from the investable universe as well as the client’s benchmark. Such considerations are often referred to as ESG (environmental, social, governance), and investing in accordance with such considerations is referred to as SRI (socially responsible investing).

When the portfolio represents only part of the client’s total wealth, there may be aspects or portions of wealth not under the control of the manager that have implications for the portfolio. For example, an employee of a public company whose labor income and retirement income provision are reliant on that company and who may have substantial investment exposure to the company through employee share options and stock holdings, may decide that their portfolio should not invest additional amounts in that stock. An entrepreneur may be reluctant to see his or her portfolio invested in the shares of competing businesses or in any business that has risk exposures aligned with his or her entrepreneurial venture.

A client’s income may rely on a particular industry or asset class. Appropriate diversification requires that industry or asset class to be deemphasized in the client’s investments. For example, a stockbroker should consider having a relatively low weighting in equities, as his skills and thus income-generating ability are worth less when equities do not perform well. Employees should similarly be wary of having concentrated share positions in the equity of the company they work for. If the employer encounters difficulties, not only may the employee lose his or her job, but their investment portfolio could also suffer a significant loss of value.

2.3. Gathering Client Information

As noted previously, it is important for portfolio managers and investment advisers to know their clients. For example, Dutch securities industry practice requires financial intermediaries to undertake substantial fact finding. This is required not only in the case of full-service wealth management or in the context of an IPS, but also in “lighter” forms of financial intermediation, such as advisory relationships (in which clients make investment decisions after consultation with their investment adviser or broker) or execution-only relationships (in which the client makes his investment decisions independently).

An exercise in fact-finding about the customer should take place at the beginning of the client relationship. This will involve gathering information about the client’s circumstances as well as discussing the client’s objectives and requirements.

Important data to gather from a client should cover family and employment situations as well as financial information. If the client is an individual, it may also be necessary to know about the situation and requirements of the client’s spouse or other family members. The health of the client and his or her dependents is also relevant information. In an institutional relationship, it will be important to know about key stakeholders in the organization and what their perspective and requirements are. Information gathering may be done in an informal way or may involve structured interviews or questionnaires or analysis of data. Many advisers will capture data electronically and use special systems that record data and produce customized reports.

Good recordkeeping is very important, and may be crucial in a case in which any aspect of the client relationship comes into dispute at a later stage.

Example 7-8, the final one based on Henri Gascon, shows how the information obtained from the fact-finding exercises might be incorporated into the objectives and constraints section of an IPS.

3. PORTFOLIO CONSTRUCTION

Once the IPS has been compiled, the investment manager can construct a suitable portfolio. Strategic asset allocation is a traditional focus of the first steps in portfolio construction. The strategic asset allocation is stated in terms of percent allocations to asset classes. An asset class is a category of assets that have similar characteristics, attributes, and risk/return relationships. The strategic asset allocation (SAA) is the set of exposures to IPS-permissible asset classes that is expected to achieve the client’s long-term objectives given the client’s investment constraints.

The focus on the SAA is the result of a number of important investment principles. One principle is that a portfolio’s systematic risk accounts for most of its change in value over the long term. Systematic risk is risk related to the economic system (e.g., risk related to business cycle) that cannot be eliminated by holding a diversified portfolio. This risk is different from the particular risks of individual securities, which may be avoided by holding other securities with offsetting risks. A second principle is that the returns to groups of similar assets (e.g., long-term debt claims) predictably reflect exposures to certain sets of systematic factors (e.g., for the debt claims, unexpected changes in the inflation rate). Thus, the SAA is a means of providing the investor with exposure to the systematic risks of asset classes in proportions that meet the risk and return objectives.

The process of formulating a strategic asset allocation is based on the IPS, already discussed, and capital market expectations, introduced in Section 3.1. How to make the strategic asset allocation operational with a rebalancing policy and a translation into actual investment portfolios will be described in Section 3.3. Section 3.4 lists some alternatives to the approach chosen and describes some portfolio construction techniques.

3.1. Capital Market Expectations

Capital market expectations are the investor’s expectations concerning the risk and return prospects of asset classes, however broadly or narrowly the investor defines those asset classes.4 When associated with the client’s investment objectives, the result is the strategic asset allocation that is expected to allow the client to achieve his or her investment objectives (at least under normal capital market conditions).

Traditionally, capital market expectations are quantified in terms of asset class expected returns, standard deviation of returns, and correlations among pairs of asset classes. Formally, the expected return of an asset class consists of the risk-free rate and one or more risk premium(s) associated with the asset class. Expected returns are in practice developed in a variety of ways, including the use of historical estimates, economic analysis, and various kinds of valuation models. Standard deviations and correlation estimates are frequently based on historical data.

3.2. The Strategic Asset Allocation

Traditionally, investors have distinguished cash, equities, bonds, and real estate as the major asset classes. In recent years, this list has been expanded with private equity, hedge funds, and commodities. In addition, such assets as art and intellectual property rights may be considered asset classes for those investors prepared to take a more innovative approach and to accept some illiquidity. Combining such new asset classes as well as hedge funds and private equity under the header “alternative investments” has become accepted practice.

As the strategic asset allocation is built up by asset classes, the decision about how to define those asset classes is an important one. Defining the asset classes also determines the extent to which the investor controls the risk and return characteristics of the eventual investment portfolio. For example, separating bonds into government bonds and corporate bonds, and then further separating corporate bonds into investment grade and noninvestment grade (high yield) and government bonds into domestic and foreign government bonds, creates four bond categories for which risk–return expectations can be expressed and correlations with other asset classes (and, in an asset–liability management context, with the liabilities) can be estimated. An asset allocator who wants to explicitly consider the risk–return characteristics of those bond categories in the strategic asset allocation may choose to treat them as distinct asset classes. Similarly, in equities some investors distinguish between emerging market and developed market equities, between domestic and international equities, or between large-cap and small-cap equities. In some regulatory environments for institutional investors, asset class definitions are mandatory, thereby forcing asset allocators to articulate risk–return expectations (and apply risk management) on the asset classes specified.

When defining asset classes, a number of criteria apply. Intuitively, an asset class should contain relatively homogeneous assets while providing diversification relative to other asset classes. In statistical terms, risk and return expectations should be similar and paired correlations of assets should be relatively high within an asset class but should be lower versus assets in other asset classes.5 Also, the asset classes, while being mutually exclusive, should add up to a sufficient approximation of the relevant investable universe. Applying these criteria ensures that the strategic asset allocation process has considered all available investment alternatives.

Using correlation as a metric, Example 7-9 tends to indicate that only emerging markets were well differentiated from European equities. So, why do investors still often subdivide equities? Apart from any regulatory reasons, one explanation might be that this decomposition into smaller asset classes corresponds to the way the asset allocation is structured in portfolios. Many investment managers have expertise exclusively in specific areas of the market, such as emerging market equities, U.S. small-cap equity, or international investment-grade credit. Bringing the asset class definitions of the asset allocation in line with investment products actually available in the market may simplify matters from an organizational perspective.

The risk–return profile of the strategic asset allocation depends on the expected returns and risks of the individual asset classes, as well as the correlation between those asset classes. In general, adding assets classes with low correlation improves the risk–return trade-off (more return for similar risk), as long as the stand-alone risk of such asset classes does not exceed its diversification effect. Typically, the strategic asset allocation for risk-averse investors will have a large weight in government bonds and cash, whereas those with more willingness and ability to take risk will have more of their assets in risky asset classes, such as equities and many types of alternative investments.

A strategic asset allocation results from combining the constraints and objectives articulated in the IPS and long-term capital market expectations regarding the asset classes. The strategic asset allocation or policy portfolio will subsequently be implemented into real portfolios. Exhibit 7-6 illustrates conceptually how investment objectives and constraints and long-term capital market expectations combine into a policy portfolio.

In some frameworks used in practice, the asset allocation is an integral part of the investment policy statement. This presentation, however, keeps the asset allocation separate from the investment policy statement because clients’ investment objectives and constraints qualitatively differ in nature from capital market expectations, thus requiring different types of analysis, different sources of information, and different review cycles.

How will capital market expectations and investment objectives and constraints eventually translate into a strategic asset allocation? In general, investors choosing from a range of asset allocations with similar returns would prefer those with lower risk. Choosing from allocations with similar levels of risk, investors would prefer those with the highest return. Formally, investors’ risk and return objectives can be described as a utility function, in which utility increases with higher expected returns and lower risk. The assumption that expected utility is increasing in expected return and decreasing in variance of return could yield an expected utility equation, such as that shown in Equation 7.1.6

where

This utility function expresses a positive relationship between utility and expected portfolio return (i.e., higher expected return increases utility, all else equal) and a negative relationship between utility and volatility of portfolio return as measured by the variance of portfolio returns. The stronger the negative relationship, the greater the investor’s risk aversion. The portfolio is understood to represent a particular asset allocation. The asset allocation providing the highest expected utility is the one that is optimal for the investor given his or her risk aversion.

For different values of Up, a line can be plotted that links those combinations of risk and expected return that produces that level of utility: an indifference curve. An investor would attain equal utility from all risk/return combinations on that curve.

Capital market expectations, specified in asset classes’ expected returns, standard deviations of return, and correlations, translate into an efficient frontier of portfolios. A multiasset class portfolio’s expected return is given by

where wi equals the weight of asset class i in the portfolio, and its risk is given by

The covariance between the returns on asset classes i and j is given by the product of the correlation between the two asset classes and their standard deviations of return:

where

The resulting portfolios can be represented as a scatter of dots in a chart depicting their risk and expected return. As a portfolio’s risk is a positive function of the risk of its assets and the correlations among them, a portfolio consisting of lowly correlated risky assets has lower risk than one with similarly risky assets with high correlation. It is therefore possible to construct different portfolios with equal expected returns but with different levels of risk. The line that connects those portfolios with the minimal risk for each level of expected return (above that of the minimum variance portfolio) is the efficient frontier. Clearly, the efficient frontier will move “upward” as more lowly correlated assets with sufficient expected return are added to the mix because it lowers the risk in the portfolios. Similarly, when return expectations increase for asset classes while volatility and correlation assumptions remain unchanged, the efficient frontier will move upward because each portfolio is able to generate higher returns for the same level of risk.

Both the efficient frontier and a range of indifference curves can be plotted in the risk–return space. In Exhibit 7-7, the curves that are concave from below represent efficient frontiers associated with different assumed expected returns. The dashed curves are indifference curves. The point where the efficient frontier intersects with the indifference curve with the highest utility attainable (i.e., the point of tangency) represents the optimal asset allocation for the client/investor. In Exhibit 7-7, efficient frontier 1 has a point of tangency with indifference curve 1. Higher levels of utility, such as those associated with indifference curve 0, can apparently not be reached with the assets underlying the efficient frontier. It is clear that when capital market expectations change, this change moves the efficient frontier away from its original location. In the chart, this movement is illustrated by efficient frontier 2, which incorporates different capital market expectations. This new efficient frontier has a point of tangency with indifference curve 2, which is associated with a lower level of expected utility. Because the point of tangency represents the strategic asset allocation, it implies the asset allocation should be adjusted. Similarly, should investment objectives or constraints change, the indifference curves will change their shape and location. This change will again move the point of tangency, and hence change the asset allocation.

This framework describes how investor objectives and capital market expectations should theoretically be reconciled. It will, however, not be the exact procedure that in practice will be followed. First, an IPS does not necessarily translate the client’s investment objectives and constraint into a utility function. Rather, an IPS gives threshold levels for risk and expected return, combined with a number of additional constraints that cannot be captured in this model. Second, the model illustrated is a single-period model, whereas in practice, the constraints from the IPS will make it more appropriate to use multiperiod models. Multiperiod problems can be more effectively addressed using simulation.

3.3. Steps toward an Actual Portfolio

The strategic asset allocation in itself does not yet represent an actual investment portfolio. It is the first step in implementing an investment strategy. For quantitatively oriented portfolio managers, the next step is often risk budgeting.

As used in this chapter, risk budgeting is the process of deciding on the amount of risk to assume in a portfolio (the overall risk budget), and subdividing that risk over the sources of investment return (e.g., strategic asset allocation, tactical asset allocation, and security selection).7 Because the decision about the amount of risk to be taken is made in constructing the IPS, at this stage we are concerned about the subdivision of that risk.

Apart from the exposures to systematic risk factors specified in the strategic asset allocation, the returns of an investment strategy depend on two other sources: tactical asset allocation and security selection. Tactical asset allocation is the decision to deliberately deviate from the policy exposures to systematic risk factors (i.e., the policy weights of asset classes) with the intent to add value based on forecasts of the near-term returns of those asset classes. For instance, an investor may decide to temporarily invest more of the portfolio in equities than the SAA prescribes if the investor anticipates that equities will deliver a higher return over the short term than other asset classes. Security selection is an attempt to generate higher returns than the asset class benchmark by selecting securities with a higher expected return. For example, an investment manager may decide to add more IBM stock in his portfolio than the weight in his equity benchmark if he expects this stock to do better than the benchmark. To fund this purchase, he may sell another stock expected to do worse than either the benchmark or IBM. Obviously, deciding to deviate from policy weights or to select securities aiming to beat the benchmark creates additional uncertainty about returns. This risk is over and above the risk inherent in the policy portfolio. Hence, an investment policy should set risk limits and desired payoffs for each of these three activities.

Risk budgeting implies that the portfolio manager has to choose, for every asset class, whether to deploy security selection as a return generator. This choice is generally referred to as the choice between active or passive management. Contrary to strategic asset allocation, where exposures to sources of systematic risk are selected and sized, security selection is not rewarded with a long-run payoff to risk. Security selection is a zero-sum game: All investors in an asset class are competing with each to identify a typically limited number of assets that are misvalued. In total, the gross returns of all market participants average out to the market return (the reward for taking systematic risk). This implies that the average active investor will match the market return, and that one investor’s gain versus the market return is the other investor’s loss versus the market return. However, because active managers tend to trade more and have to pay people (including themselves) to generate investment ideas or information leading to such ideas, the average active manager will underperform the market, net of costs. This does not imply, however, that there are no skillful investment managers who, with some consistency, beat their benchmarks. Neither does it imply that all passive managers will be able to match the benchmark. The higher the turnover of an index, the more trading costs a passive manager will incur, making the task of effectively mimicking an index more difficult.

The likelihood of adding a significant amount of value from security selection depends on the skills of the manager and the informational efficiency of the market for the asset class his skill relates to. The more efficient an asset class or a subset of that asset class (such as a regional stock, bond, or real estate market or a size category within the stock market), the more skillful an asset manager has to be to add value. Broadly speaking, an efficient market is a market in which prices, on average, very quickly reflect newly available information. That requires a sizeable participation of mean–variance optimizing investors, acting on rational expectations, using the same or similar pricing models, and having equal opportunities to access relevant information. Clearly, the market for U.S. large-capitalization equities would be quite efficient. By contrast, some regional bond and equity markets do not have the technical and regulatory systems for information dissemination that are sufficient to serve all investors on a timely basis. Skilled managers should be able to exploit the resulting inefficiencies.

Sometimes, however, the choice between active and passive management is actually made implicitly when the asset class is included in the asset allocation. The markets for some assets—such as those for nonlisted real estate, art, and infrastructure assets—are so illiquid that it is very difficult to buy a diversified exposure. As a result, there is no way of taking exposure to the asset class without engaging in security selection.

As the portfolio is constructed and its value changes with the returns of the asset classes and securities in which it is invested, the weights of the asset classes will gradually deviate from the policy weights in the strategic asset allocation. This process is referred to as drift. Periodically, or when a certain threshold deviation from the policy weight (the bandwidth) has been breached, the portfolio should be rebalanced back to the policy weights. The set of rules that guide the process of restoring the portfolio’s original exposures to systematic risk factors is known as the rebalancing policy. Even absent a formal risk budget, formulating a rebalancing policy is an important element of risk management.

The process of executing an investment strategy continues with selecting the appropriate manager(s) for each asset class and allocating funds to them. The investment portfolio management process is then well into the execution stage.

The investment managers’ performance will be monitored, as well as the results of the strategic asset allocation. When asset class weights move outside their corridors, money is transferred from the asset classes that have become too large compared with the SAA to those that fall short. Managers as well as the strategic asset allocation will be reviewed on the basis of the outcome of the monitoring process. In addition, capital market expectations may change, as may the circumstances and objectives of the client. These changes could result in an adjustment of the strategic asset allocation.

3.4. Additional Portfolio Organizing Principles

The top-down oriented framework laid out in earlier paragraphs is quite general. Other models of portfolio organization are also used in practice, and these are described briefly here. In addition, this section introduces some portfolio construction concepts in use to better capture the value of active management.

According to some practitioners, a top-down investment process as described earlier has two drawbacks. They both result from the fact that in a top-down process, a multitude of specialist managers may work for the same client within the same asset class. Each of these managers will manage risk versus the client’s benchmark, and because these benchmarks may be similar to or overlapping with those of other managers, the aggregate of all the portfolios within one asset class may be less active than was intended. The resulting investment portfolio may underutilize its risk budget. Another drawback is that as the investment managers each engage in trading over the full extent (including the benchmark exposure) of their portfolio, the aggregate of all portfolios may not be efficient from a capital gains tax point of view. More trading results in more capital gains being realized, increasing the tax bill at the end of the year. To circumvent these issues, the core–satellite approach was developed. In this approach, a majority of the portfolio is invested on a passive or low active risk basis (usually a combination of a bond portfolio and an equity portfolio) while a minority of the assets is managed aggressively in smaller “satellite” portfolios. The aim of the satellite portfolios is to generate a high active return with little regard for benchmark exposure, whereas the core is managed with low turnover to capture the long-term systematic risk premium of its assets on a tax-optimal basis. The aggressive management in satellites can be executed with the objective of market timing (in which case tactical asset tilts are executed through long–short positions in asset class indices or derivatives), as well as security selection (in which case, highly active security selection vehicles act as satellites). A drawback of the core–satellite approach is that the assignment of asset class monies to actual portfolio managers—with their various expected alphas, risk surrounding those alphas, and correlations between those alphas—can be seen as an optimization process in its own right. The outcome of that process is not necessarily consistent with a core–satellite structure (Waring and Siegel, 2003).

4. SUMMARY

In this chapter, we have discussed construction of a client’s investment policy statement, including discussion of risk and return objectives and the various constraints that will apply to the portfolio. We have also discussed the portfolio construction process, with emphasis on the strategic asset allocation decisions that must be made.

• The IPS is the starting point of the portfolio management process. Without a full understanding of the client’s situation and requirements, it is unlikely that successful results will be achieved.
• The IPS can take a variety of forms. A typical format will include the client’s investment objectives and also list the constraints that apply to the client’s portfolio.
• The client’s objectives are specified in terms of risk tolerance and return requirements.
• The constraints section covers factors that need to be considered when constructing a portfolio for the client that meets the objectives. The typical constraint categories are liquidity requirements, time horizon, regulatory requirements, tax status, and unique needs.
• Risk objectives are specifications for portfolio risk that reflect the risk tolerance of the client. Quantitative risk objectives can be absolute or relative or a combination of the two.
• The client’s overall risk tolerance is a function of the client’s ability to accept risk and their “risk attitude,” which can be considered the client’s willingness to take risk.
• The client’s return objectives can be stated on an absolute or a relative basis. As an example of an absolute objective, the client may want to achieve a particular percentage rate of return. Alternatively, the return objective can be stated on a relative basis, for example, relative to a benchmark return.
• The liquidity section of the IPS should state what the client’s requirements are to draw cash from the portfolio.
• The time horizon section of the IPS should state the time horizon over which the investor is investing. This horizon may be the period during which the portfolio is accumulating before any assets need to be withdrawn.
• Tax status varies among investors and a client’s tax status should be stated in the IPS.
• The IPS should state any legal or regulatory restrictions that constrain the investment of the portfolio.
• The unique circumstances section of the IPS should cover any other aspect of a client’s circumstances that is likely to have a material impact on the composition of the portfolio; for example, any religious or ethical preferences.
• Asset classes are the building blocks of an asset allocation. An asset class is a category of assets that have similar characteristics, attributes, and risk/return relationships. Traditionally, investors have distinguished cash, equities, bonds, and real estate as the major asset classes.
• A strategic asset allocation results from combining the constraints and objectives articulated in the IPS and capital market expectations regarding the asset classes.
• As time goes on, a client’s asset allocation will drift from the target allocation, and the amount of allowable drift as well as a rebalancing policy should be formalized.
• In addition to taking systematic risk, an investment committee may choose to take tactical asset allocation risk or security selection risk. The amount of return attributable to these decisions can be measured.

PROBLEMS

1. Which of the following is least important as a reason for a written investment policy statement (IPS)?

A. The IPS may be required by regulation.

B. Having a written IPS is part of best practice for a portfolio manager.

C. Having a written IPS ensures the client’s risk and return objectives can be achieved.

2. Which of the following best describes the underlying rationale for a written investment policy statement (IPS)?

A. A written IPS communicates a plan for trying to achieve investment success.

B. A written IPS provides investment managers with a ready defense against client lawsuits.

C. A written IPS allows investment managers to instruct clients about the proper use and purpose of investments.

3. A written investment policy statement (IPS) is most likely to succeed if:

A. Created by a software program to assure consistent quality.

B. It is a collaborative effort of the client and the portfolio manager.

C. It reflects the investment philosophy of the portfolio manager.

4. The section of the investment policy statement (IPS) that provides information about how the policy may be executed, including investment constraints, is best described as the:

A. Investment Objectives.

B. Investment Guidelines.

C. Statement of Duties and Responsibilities.

5. Which of the following is least likely to be placed in the appendices to an investment policy statement (IPS)?

A. Rebalancing Policy.

B. Strategic Asset Allocation.

C. Statement of Duties and Responsibilities.

6. Which of the following typical topics in an investment policy statement (IPS) is most closely linked to the client’s “distinctive needs”?

A. Procedures.

B. Investment Guidelines.

C. Statement of Duties and Responsibilities.

7. An investment policy statement that includes a return objective of outperforming the FTSE 100 by 120 basis points is best characterized as having a(n):

A. Relative return objective.

B. Absolute return objective.

C. Arbitrage-based return objective.

8. Risk assessment questionnaires for investment management clients are most useful in measuring:

A. Value at risk.

B. Ability to take risk.

C. Willingness to take risk.

9. Which of the following is best characterized as a relative risk objective?

A. Value at risk for the fund will not exceed US$3 million.

B. The fund will not underperform the DAX by more than 250 basis points.

C. The fund will not lose more than €2.5 million in the coming 12-month period.

10. In preparing an investment policy statement, which of the following is most difficult to quantify?

A. Time horizon.

B. Ability to accept risk.

C. Willingness to accept risk.

11. After interviewing a client in order to prepare a written investment policy statement (IPS), you have established the following:

• The client has earnings that vary dramatically between £30,000 and £70,000 (pre-tax) depending on weather patterns in Britain.
• In three of the previous five years, the after-tax income of the client has been less than £20,000.
• The client’s mother is dependent on her son (the client) for approximately £9,000 per year support.
• The client’s own subsistence needs are approximately £12,000 per year.
• The client has more than 10 years experience trading investments including commodity futures, stock options, and selling stock short.
• The client’s responses to a standard risk assessment questionnaire suggest he has above average risk tolerance.

The client is best described as having a:

A. Low ability to take risk, but a high willingness to take risk.

B. High ability to take risk, but a low willingness to take risk.

C. High ability to take risk and a high willingness to take risk.

12. After interviewing a client in order to prepare a written investment policy statement (IPS), you have established the following:

• The client has earnings that have exceeded €120,000 (pre-tax) each year for the past five years.
• She has no dependents.
• The client’s subsistence needs are approximately €45,000 per year.
• The client states that she feels uncomfortable with her lack of understanding of securities markets.
• All of the client’s current savings are invested in short-term securities guaranteed by an agency of her national government.
• The client’s responses to a standard risk assessment questionnaire suggest she has low risk tolerance.

The client is best described as having a:

A. Low ability to take risk, but a high willingness to take risk.

B. High ability to take risk, but a low willingness to take risk.

C. High ability to take risk and a high willingness to take risk.

13. A client who is a 34-year-old widow with two healthy young children (aged 5 and 7) has asked you to help her form an investment policy statement. She has been employed as an administrative assistant in a bureau of her national government for the previous 12 years. She has two primary financial goals—her retirement and providing for the college education of her children. This client’s time horizon is best described as being:

A. Long term.

B. Short term.

C. Medium term.

14. The timing of payouts for property and casualty insurers is unpredictable (“lumpy”) in comparison with the timing of payouts for life insurance companies. Therefore, in general, property and casualty insurers have:

A. Lower liquidity needs than life insurance companies.

B. Greater liquidity needs than life insurance companies.

C. A higher return objective than life insurance companies.

15. A client who is a director of a publicly listed corporation is required by law to refrain from trading that company’s stock at certain points of the year when disclosure of financial results are pending. In preparing a written investment policy statement (IPS) for this client, this restriction on trading:

A. Is irrelevant to the IPS.

B. Should be included in the IPS.

C. Makes it illegal for the portfolio manager to work with this client.

16. Consider the pairwise correlations of monthly returns of the following asset classes:

Based solely on the information in the preceding table, which equity asset class is most sharply distinguished from U.S. equities?

A. Brazilian equities.

B. European equities.

C. East Asian equities.

17. Returns on asset classes are best described as being a function of:

A. The failure of arbitrage.

B. Exposure to the idiosyncratic risks of those asset classes.

C. Exposure to sets of systematic factors relevant to those asset classes.

18. In defining asset classes as part of the strategic asset allocation decision, pairwise correlations within asset classes should generally be:

A. Equal to correlations among asset classes.

B. Lower than correlations among asset classes.

C. Higher than correlations among asset classes.

19. Tactical asset allocation is best described as:

A. Attempts to exploit arbitrage possibilities among asset classes.

B. The decision to deliberately deviate from the policy portfolio.

C. Selecting asset classes with the desired exposures to sources of systematic risk in an investment portfolio.

20. Investing the majority of the portfolio on a passive or low active risk basis while a minority of the assets is managed aggressively in smaller portfolios is best described as:

A. The core–satellite approach.

B. A top-down investment policy.

C. A delta-neutral hedge approach.

1Maginn and Tuttle (1983), p. 23.

2Value at risk is a money measure of the minimum value of losses expected during a specified time period at a given level of probability.

3Tracking risk (sometimes called tracking error) is the standard deviation of the differences between a portfolio’s returns and its benchmark’s returns.

4For an in-depth discussion of this topic see Calverley, Meder, Singer, and Staub (2007).

5A technically precise characterization other than one in terms of pairwise correlations is given by Kritzman (1999).

6Sharpe, Chen, Pinto, and McLeavey (2007).

7Some writers use risk budgeting to refer to allocating an amount or budget of tracking risk to active portfolio managers on a given asset class. See Waring, Whitney, Pirone, and Castille (2000).