PM 4 Basics of Portfolio Planning and Construction
Before any money is committed, an adviser needs to understand who the client is: the goals being pursued, the resources available, the personal or institutional circumstances that shape the decision, and the limits that apply. Broad investor types share common features, yet almost every client carries some distinctive requirement, so the planning has to be individualized. Portfolio planning is the program that is worked out in advance of building the portfolio, and the document that records it is the investment policy statement, or IPS.
The IPS anchors the whole portfolio management process. Success here means the client reaching the goals that matter to them by means they are comfortable with, in terms of the risks accepted and any other concerns. Put simply, the IPS communicates a plan for achieving investment success, and without a full grasp of the client’s situation that success is unlikely.
The document is normally drafted after a fact-finding conversation with the client, which may lean on a questionnaire that draws out risk tolerance and specific expectations. For an institution the fact finding can widen into asset and liability reviews, an assessment of liquidity needs, and a range of tax, legal, and other matters. A growing share of clients also attach a policy on responsible investing, the umbrella term for principles that address environmental, social, and governance (ESG) themes an investor wants weighed both when deciding to invest in a company and throughout the holding period. Sustainable investing is used in much the same spirit, with the emphasis on factoring sustainability into the process. Throughout this lesson we assume the responsible investing policy is folded into the IPS itself.
Why the document matters
Keeping a well-constructed IPS for every client should be routine. The manager builds the portfolio by reference to it and consults it when judging whether a particular investment suits the client. In some places the requirement is not merely good practice but law. UK pension schemes, for instance, must maintain a statement of investment principles under Section 35 of the Pensions Act 1995, and that statement is in essence an IPS. Regulators also press firms to know their customers, and the European Union Markets in Financial Instruments Directive (MiFID) obliges firms to sort clients into categories (eligible counterparties, institutional clients, or retail clients), with the category fixing the protections and limits that apply by law. For an institution like a pension plan or endowment, the IPS can also lay out governance arrangements, for example how the investment committee appoints and reviews managers and how much discretion those managers hold.
An IPS is not written once and forgotten. It should be revisited on a regular cycle so that it still fits the client. The UK Pensions Regulator, for example, suggests a scheme review its statement of investment principles at least every three years. Beyond the calendar, a review is triggered whenever the manager learns of a material shift in the client’s situation, or when the client’s own objectives, time horizon, or liquidity needs shift.
The major components
No single standard format exists, but many statements share a common skeleton:
- Introduction, describing the client.
- Statement of Purpose, stating why the IPS exists.
- Statement of Duties and Responsibilities, setting out what the client, the custodian, and the managers each owe.
- Procedures, covering how the document is kept current and how contingencies are handled.
- Investment Objectives, explaining what the client is investing to achieve.
- Investment Constraints, listing the factors that limit the pursuit of those objectives.
- Investment Guidelines, describing how policy is carried out, such as permitted use of leverage and derivatives and any asset types excluded.
- Evaluation and Review, on how results are fed back.
- Appendices, usually the strategic asset allocation (the policy portfolio) and the rebalancing policy.
Of these, the objectives and the constraints sections are tied most closely to the client’s distinctive needs and are the most important for planning. An IPS built around just these two is said to follow an objectives-and-constraints format, and that is the format the rest of this lesson develops, starting with risk and return objectives. Working through the IPS is the basic mechanism for testing and improving a client’s overall return-and-risk stance: return expectations must be made consistent with risk objectives, and both must sit comfortably within the constraints. Increasingly, investors also fold non-financial, ESG-related convictions into their policies, recognizing that such considerations may in time affect the portfolio’s financial profile as well as express the investor’s values.
The portfolio’s risk must suit the client, so the IPS should state risk tolerance plainly. Risk objectives put numbers on the portfolio risk that matches that tolerance, and quantitatively they come in absolute, relative, or blended form.
Absolute and relative risk objectives
An absolute objective is self-standing and unrelated to how markets perform: a wish to avoid any loss of capital, or to cap the loss at a set percentage in any twelve-month period. Taken literally, a no-loss objective could be met by taking no risk at all, for example by holding an insured certificate of deposit at a sound bank. Once risky assets are involved, though, the objective works better as a probability statement. A desire not to lose more than 4 percent in any twelve-month period becomes an objective that, with 95 percent probability, the portfolio will not lose more than 4 percent over any such period. Absolute risk is captured by return variance or standard deviation, and by value at risk, which states the minimum loss expected over a period at a given probability.
A relative objective measures risk against one or more benchmarks taken to represent an appropriate standard. Large-cap UK equities might be judged against an index such as the FTSE 100, large-cap US equities against the S&P 500, and cash-like holdings against a rate such as the Treasury bill rate. The natural measure of risk relative to a benchmark is tracking risk, also called tracking error, the standard deviation of the differences between the portfolio’s returns and the benchmark’s. Relative objectives suit situations where a client’s wealth is divided into partial mandates.
Some clients weigh assets and liabilities together. When the size, timing, or relative certainty of future obligations is known, the IPS can be shaped to meet them, an approach called liability-driven investment (LDI). Life insurers, defined benefit pension plans, and an individual’s post-retirement budget all fit this mold. A pension plan, for example, must meet payments as they fall due, so its risk objective is to minimize the chance of failing to do so, and a matching return objective might be to beat the discount rate used to value the liabilities over a multi-year horizon. When a policy portfolio of long-term asset weights and hedge ratios is in place, the risk objective may be framed as holding the return inside a plus-or-minus X percent band around a blended benchmark, again read most usefully as a probability, such as a 95 percent chance of staying within X percent of the benchmark over a stated period.
A Japanese institutional investor holds a portfolio worth ¥10 billion. Her first risk objective is a wish not to lose more than ¥1 billion over the coming twelve months. Her second is to earn returns within 4 percent of the return on the TOPIX index, which is her benchmark.
Ability versus willingness to take risk
Overall risk tolerance combines two things: the client’s ability to bear risk and the client’s risk attitude, or willingness to take risk. Above-average readings on both give above-average risk tolerance; below-average readings on both give below-average tolerance. When the two disagree, the situation needs resolving, as the grid below shows.
| Ability to bear risk | Willingness below average | Willingness above average |
|---|---|---|
| Below average | Below-average risk tolerance | Resolution needed |
| Above average | Resolution needed | Above-average risk tolerance |
Ability to bear risk rests mainly on objective factors: time horizon, expected income, and wealth relative to liabilities. An investor with a twenty-year horizon can, other things equal, bear more risk than one with a two-year horizon, because two decades leave far more room to recover losses or adjust course. Likewise, an investor whose assets sit comfortably above their liabilities, or a pension plan running a large surplus, can bear more risk than one whose wealth and future spending are finely balanced.
Willingness to take risk is more subjective, shaped by the client’s psychology and current circumstances. Traits such as personality type, self-esteem, and independent thinking appear to correlate with risk attitude: some people are at ease with financial risk while others find it distressing. There is no agreed way to measure it, but it can be gauged through discussion or a psychometric questionnaire. One validated five-item instrument developed by John Grable and collaborators asks respondents to react to statements such as investing being too difficult to understand, or safety mattering more than returns, with answers coded 1 through 4 and summed. Scores run from a low of 5 to a high of 20, with higher scores signaling greater risk tolerance. Across two random samples of faculty and staff at large US universities (n = 406), the average came to 12.86, the standard deviation to 3.01, and the median to 13.
When ability and willingness align, the adviser’s job is straightforward. When ability is low but willingness is high, overall tolerance should be judged low. When ability is high but willingness is low, the adviser may counsel the client, laying out why the ability is high and what forgoing return costs, but should not try to change a willingness that is genuine rather than a miscalculation, since reshaping personality is outside the adviser’s role. The prudent rule is to settle on the lower of the two factors and document the decision. The same conflict arises with institutions, where different stakeholders may disagree; when a trustee owes a fiduciary duty to pension beneficiaries and their interests clash with the sponsor’s, the trustee must act for the beneficiaries.
Two siblings consult the same adviser. Henri Gascon is a 30-year-old energy trader in Paris, married with a five-year-old son. His salary of €250,000 more than covers the family’s outgoings, he owns his apartment outright and holds €1,000,000 of savings, his job feels secure, and he is financially knowledgeable and confident that equities will do well over the long term. On the questionnaire he strongly disagrees that stock and bond gains come down to luck and that safety beats returns. He expects to fund retirement from his savings starting at age 50. His brother Jacques is a 40-year-old self-employed consultant, divorced with four children aged 12 to 16. His earnings average €40,000 but swing widely, he pays €10,000 a year to his ex-wife and children, and he carries a €100,000 mortgage against only €10,000 of savings. Jacques is also knowledgeable and expects very high equity returns, and on the questionnaire he strongly disagrees that a bank account is more comfortable than the stock market and that risk brings loss to mind. He expects most of his savings to support the children at university.
Return objectives
Return objectives, like risk objectives, can be absolute or relative. An absolute objective names a particular rate, stated in nominal terms or in real (inflation-adjusted) terms. A relative objective is set against a benchmark, such as an equity index or a cash rate like the market reference rate, perhaps as a wish to beat the benchmark by one percentage point a year. Some institutions aim relative to a peer group, for instance an endowment targeting the top half of similar funds or a private equity mandate targeting the top quartile. Peer targets are problematic, because little may be known about rivals’ strategies or return methods, not everyone can be above average, and a good benchmark should be investable, which a peer group generally is not.
Returns can be quoted before or after fees, and before or after tax, with the fee basis made clear to both sides; for a taxable investor, returns are best quoted and assessed after tax. A return objective may also be a required return, the amount needed to reach a future goal such as a target retirement income. Above all, the objective has to be realistic. It must be consistent with the risk objective, since high returns rarely come without high risk, and consistent with the market environment: a 15 percent nominal return might be attainable when inflation runs at 10 percent but not when it runs at 3 percent. Client and manager should also agree on whether the target is nominal, which is convenient to measure, or real, which usually maps better to the underlying goal. Where expectations are unrealistic, the adviser must counsel the client on what the market and their risk tolerance actually allow.
Marie Gascon wants to retire in twenty years, at age 50. Her salary meets current and expected spending, but she cannot add to her pension fund. She sets aside €100,000 of savings as a cash emergency fund and invests the remaining €900,000 for retirement. She estimates that €2,000,000 in today’s money, before tax, will fund her retirement income. The adviser expects inflation to average 2 percent a year over the twenty years. In France, pension contributions and pension fund returns are tax-exempt, while distributions are taxed on retirement.
Five categories of constraint shape which assets a portfolio can hold: liquidity, time horizon, tax concerns, legal and regulatory factors, and unique circumstances. Constraints may be internal, set by the client, or external, imposed by law or regulation.
Liquidity
The IPS should state the likely need to withdraw cash. For an individual this might be healthcare costs or tuition; for an institution it might be endowment spending rules, claims coming due at a property and casualty insurer, or benefit payments at a pension fund or life insurer. Where a need exists, part of the portfolio should be set aside to cover it, invested in assets that are liquid, meaning easily turned into cash, and low in risk at the moment the cash is needed, so their value is known with reasonable certainty. A bond maturing when school fees fall due is a simple example. Insurers illustrate the point at scale: the portfolio of the US insurer Progressive Corporation, shown below, leans heavily on fixed income, much of it liquid or short-dated, so that unpredictable automobile claims can be paid.
| Asset category | Allocation |
|---|---|
| Fixed maturities | 76.9% |
| Short-term investments | 10.4% |
| Common equities | 10.2% |
| Nonredeemable preferred stocks | 2.5% |
Source: Progressive Corporation, 2018 Second Quarter Report.
Time horizon
The IPS should state how long the client is investing, whether that is the accumulation period before assets are drawn down or the period until circumstances are likely to change. A 55-year-old planning to retire at 65 has a ten-year horizon; the portfolio may not be sold at 65, but its structure will likely change as income starts to be drawn. Horizon shapes the choice of investments. Illiquid or risky holdings tend to be unsuitable for a short horizon, since there is little time to recover losses, but they can suit a longer horizon, especially where the extra risk is expected to earn higher returns.
Tax concerns
Tax status differs across investors. Some pay tax on returns and some, such as many pension funds, do not. Where income (dividends and interest) and capital gains are taxed differently, income is usually taxed more heavily; gains may face a lower rate or be partly exempt, and gains are often taxed only when realized while income is taxed as earned, giving a time-value benefit to deferring gains. The portfolio should generally reflect this. A taxable investor may favor holdings that deliver capital gains rather than income, and a US taxable investor may include municipal bonds, whose interest, unlike that on Treasuries and corporates, is tax-exempt. A tax-exempt investor, a pension fund for example, hardly cares about the form returns take.
Legal and regulatory factors
The IPS should record any legal or regulatory limits that bind the portfolio. In some countries, institutional investors such as pension funds face caps on portfolio composition, for instance limits on equities, on other risky assets, or on overseas holdings, as well as self-investment limits on securities issued by the plan sponsor. The United States places no cap on pension fund asset allocation, though several other countries do, as the examples below show. A separate legal constraint arises when an individual holds material nonpublic information: a company director, for example, may be barred from trading the firm’s shares in the run-up to results, and the IPS should flag this so the stock is not traded by accident.
| Country | Listed equity | Real estate | Government bonds | Corporate bonds | Foreign currency exposure |
|---|---|---|---|---|---|
| Switzerland | 50% | 30% | 100% | 100% | Unhedged 30% |
| Japan | 100% | Not permitted | 100% | 100% | No limits |
| South Africa | 75% | 25% | 100% | 75% | 25% |
Source: OECD Survey of Investment Regulations of Pension Funds, July 2018.
Unique circumstances and ESG concerns
This section captures anything else, including beliefs and values, that may materially affect what the portfolio holds. Faith or moral values can rule out certain investments: an investor who wants to comply with Shari’a, the Islamic law, steers clear of businesses and instruments inconsistent with it, such as casinos and conventional interest-bearing bonds, because gambling and lending at interest are prohibited. Objections may be secular too, aimed at products such as weapons, tobacco, or gambling, or at practices such as environmental harm or poor labor standards, and they can exclude whole companies, countries, or security types from both the investable universe and the benchmark. Investing along such lines is often called socially responsible investing (SRI).
ESG approaches lack a settled vocabulary, but six generic types are commonly named:
- Negative screening: leaving out companies or sectors because of what they do or the environmental and social harm they cause.
- Positive screening: favoring sectors or companies that satisfy chosen ESG criteria, usually strong ESG performance relative to peers.
- ESG integration: routinely factoring material ESG issues into allocation, selection, and construction choices.
- Thematic investing: putting money into themes or assets linked to ESG factors.
- Engagement and active ownership: using shareholder power to steer corporate behavior toward ESG goals alongside financial returns.
- Impact investing: investing with the intent to create positive, measurable social and environmental impact together with a financial return.
Unique circumstances also cover the parts of a client’s wealth outside the manager’s control. An employee whose salary and retirement provision already depend on one company, and who may hold its shares and options, might decide the portfolio should avoid buying more of that stock. An entrepreneur may not want the portfolio invested in competitors or in businesses whose risks mirror her own venture. More broadly, when a client’s income leans on one industry or asset class, sound diversification argues for underweighting it: a stockbroker, whose earning power falls when equities fall, should think twice about a heavy equity weighting, and employees should be wary of concentrated positions in their employer’s shares, since trouble at the firm could cost them both their job and much of their portfolio.
Continuing the Marie Gascon case, the adviser drafts the constraints for her retirement portfolio. She expects to keep working at the oil company on a high income; the couple plan no more children and expect to fund their son’s university from salary. Her €100,000 emergency reserve is held separately, and her €900,000 of retirement savings sits in a French defined-contribution pension: contributions come from gross income, fund returns are tax-exempt, and payments to the retiree are taxed as income. Classify each constraint.
Knowing the client is not optional. In the European Union, MiFID II requires financial intermediaries to carry out substantial fact finding, not only for full wealth management or an IPS but also in lighter relationships, whether advisory (the client decides after consulting the adviser) or execution-only (the client decides independently). The fact finding belongs at the very start of the relationship and gathers both the facts of the client’s situation and a discussion of goals and requirements.
The information needed spans family and employment circumstances as well as finances. For an individual it may extend to a spouse or other family members and to the health of the client and any dependents. For an institution it means understanding the key stakeholders and what each of them wants. The exercise can be informal or run through structured interviews, questionnaires, and data analysis, and many advisers now capture it electronically in systems that store the data and generate tailored reports. Careful recordkeeping matters throughout, and can prove decisive if any part of the relationship is later disputed.
With the IPS complete, the manager can build the portfolio, and the traditional first step is the strategic asset allocation (SAA). The SAA is stated as percentage weights across asset classes, where an asset class gathers assets that share characteristics, attributes, and risk-return behavior. Formally, the SAA is the blend of exposures to IPS-permitted asset classes that is expected to meet the client’s long-term objectives, given their risk profile and constraints. It may also include a policy for hedging risks not captured by the asset weights, such as foreign currency exposure, interest rate risk from an asset-liability mismatch, or inflation risk, often implemented through overlay portfolios of derivatives.
The emphasis on the SAA follows from two principles. First, over the long term most of a portfolio’s change in value traces back to its systematic risk. Systematic risk stems from the economic system itself, the business cycle for instance, and cannot be diversified away, unlike nonsystematic risk, the risk unique to particular assets, which can be offset by holding others. Second, the returns to groups of similar assets predictably reflect exposures to certain systematic factors, for example long-term debt claims responding to unexpected changes in interest rates. The SAA is therefore how the investor takes on the systematic risks of the asset classes, in proportions that fit the objectives.
Building the SAA draws on the IPS and on capital market expectations: the investor’s views on how asset classes may perform in risk and return, however broadly or narrowly those classes are drawn. Traditionally these expectations are quantified as expected returns, return standard deviations, and pairwise correlations across asset classes. An asset class’s expected return is thought of as a risk-free rate plus one or more risk premiums; in practice expected returns come from historical estimates, economic analysis, and valuation models, while standard deviations and correlations usually rest on historical data and risk models. Combining these expectations with the client’s objectives produces the SAA that is expected to reach those objectives under normal market conditions.
For a long time the main asset classes were taken to be cash, equities, bonds split into government and corporate, and real estate. That roster has widened to take in hedge funds, private equity, commodities, and high-yield and emerging market bonds, with items such as art and intellectual property added by investors willing to accept some illiquidity for a more innovative stance. Grouping newer classes together with hedge funds and private equity under the heading alternative investments is now standard.
Defining asset classes
Because the SAA is built up class by class, how the classes are defined matters, since it sets how much control the investor has over the eventual risk and return. Splitting bonds into government and corporate, then corporate into investment grade and high yield and government into domestic and foreign, creates four bond categories, each with its own risk-return expectations and correlations that a manager can model explicitly. Equities can be split similarly, between developed and emerging markets, domestic and international, or large-cap and small-cap. Some regulatory regimes even mandate asset class definitions, forcing managers to articulate expectations for the specified classes; a broader grouping instead leaves the split between sub-categories to the managers responsible for each class.
Good asset class definitions follow a few criteria. A class should hold relatively homogeneous assets yet diversify against other classes: put statistically, expectations for risk and return should look alike inside a class, with paired correlations high within it and lower against other classes. The classes should be mutually exclusive and together approximate the investable universe closely enough that the SAA has considered all the real alternatives.
| US eq. | Emerg. mkt | European eq. | Japanese eq. | Commod. | Euro gov’t bonds | US Treasuries | |
|---|---|---|---|---|---|---|---|
| US equities | 1.00 | 0.78 | 0.88 | 0.59 | 0.32 | 0.08 | −0.37 |
| Emerging markets | 0.78 | 1.00 | 0.84 | 0.64 | 0.46 | 0.21 | −0.24 |
| European equities | 0.88 | 0.84 | 1.00 | 0.64 | 0.43 | 0.16 | −0.28 |
| Japanese equities | 0.59 | 0.64 | 0.64 | 1.00 | 0.32 | 0.24 | −0.18 |
| Commodities | 0.32 | 0.46 | 0.43 | 0.32 | 1.00 | 0.13 | −0.18 |
| European gov’t bonds | 0.08 | 0.21 | 0.16 | 0.24 | 0.13 | 1.00 | 0.45 |
| US Treasuries | −0.37 | −0.24 | −0.28 | −0.18 | −0.18 | 0.45 | 1.00 |
| Volatility | 14.3% | 21.6% | 18.4% | 15.6% | 22.3% | 4.9% | 4.4% |
Selected rows and columns from the source matrix; unhedged, in USD. Source: MSCI, Bloomberg, S&P.
The matrix shows how tightly the equity classes move together, with European and US equities correlated 0.88, while equities correlate far less with bonds and are actually negative against US Treasuries (US equities −0.37, emerging markets −0.24, European −0.28, Japanese −0.18). Among equity classes, US and Japanese equities are least alike at 0.59, against 0.78 or higher for the others. On correlation alone only emerging markets stand out clearly from European equities, which raises a fair question: why keep subdividing equities at all? Part of the answer is regulatory, and part is organizational. Many managers specialize in one niche, such as emerging market equities or US small-cap, so aligning the class definitions with the products actually available in the market simplifies matters. Do remember that correlations drift over time, and figures like these may be specific to the sample period.
How the pieces combine
An SAA’s risk-return profile turns on the expected returns and risks of each class and on how those classes correlate. In general, adding classes with low correlation improves the trade-off, offering more return for similar risk. Risk-averse investors therefore lean heavily on government bonds and cash, while those more able and willing to take risk hold more equities and alternatives. Classes are usually represented by benchmarks from providers such as FTSE, MSCI, or Bloomberg; when a negative screen or a best-in-class policy narrows the choice, the standard off-the-shelf benchmarks may no longer fit, and separate indices reflecting the exclusions can be used instead. The table below shows a real policy portfolio, that of ABP, the pension fund for Dutch government-sector employees, which had €405 billion under management as of the first quarter of 2018 and aims for a pension of 70 percent of average career real income.
| Asset class | Weight |
|---|---|
| Equities, developed countries | 27% |
| Equities, emerging markets | 9% |
| Total equity | 36% |
| Real estate | 10% |
| Private equity | 5% |
| Hedge funds | 4% |
| Commodities | 5% |
| Infrastructure | 3% |
| Total alternatives | 27% |
| Government bonds | 13% |
| Corporate bonds | 13% |
| Inflation-linked bonds | 8% |
| Emerging market bonds | 3% |
| Total fixed income | 37% |
| Total | 100% |
Source: ABP Quarterly Report Q1 2018.
Reconciling objectives with the market
In theory, objectives and capital market expectations are reconciled through optimization. Investors are assumed to prefer lower risk among allocations with similar returns and higher return among allocations with similar risk, which can be written as a utility function that rises with expected return and falls with variance:
The allocation with the highest expected utility is optimal for that investor. Plotting all risk-return combinations that yield a given level of utility traces an indifference curve, along which the investor is equally content. Capital market expectations, meanwhile, turn into an efficient frontier. A multi-asset portfolio’s expected return is the weighted sum of the class expected returns,
and its risk depends on the weights and the covariances among classes,
where the covariance between two classes is the product of their correlation and their standard deviations,
Because a portfolio’s risk falls when its assets have low correlation, many portfolios can share the same expected return at different risk levels. The line joining the lowest-risk portfolio for each level of expected return, above the minimum-variance portfolio, is the efficient frontier. Adding low-correlation assets with adequate return pushes the frontier upward, as does a rise in expected returns while volatility and correlation assumptions hold. The optimal allocation is where the efficient frontier just touches the highest attainable indifference curve, the point of tangency.
This framework is illustrative rather than the exact procedure used in practice. An IPS does not usually reduce a client to a single utility function; it gives threshold levels for risk and return plus extra constraints the simple model cannot hold. The model is also single-period, whereas the constraints in a real IPS make multi-period analysis more appropriate, and multi-period problems are often better handled by simulation.
Rainer Gottschalk has sold his regional home-construction company to a national builder and taken a job as regional manager there. His salary covers short- and medium-term needs, and he wants to invest the sale proceeds to retire well in twenty years. Because his income already depends on the real estate market, he will not invest in real estate, and on family values he excludes tobacco manufacturers and retailers from his equities. His adviser sets a return objective of 5 percent with a standard deviation of 10 percent. Market expectations for the three permitted classes are: European equities 6.0 percent return and 15.0 percent standard deviation, emerging market equities 8.0 percent and 20.1 percent, and European government bonds 2.0 percent and 7.8 percent. Excluding tobacco is estimated to lower expected European equity return by 0.2 percent and raise its standard deviation by 0.1 percent, with negligible effect on emerging markets and on correlations.
The strategic asset allocation is not yet an actual portfolio; it is the first step in putting a strategy to work. Quantitatively minded managers usually move next to risk budgeting: deciding the total amount of risk to run and then dividing it across the sources of return. Because the IPS already fixed the total risk, the concern here is the split.
Beyond the systematic exposures set in the SAA, returns come from two further sources. Tactical asset allocation is a deliberate, temporary deviation from the policy weights, made to add value on a view about near-term class returns, for instance overweighting equities when they are expected to lead. Security selection tries to beat a class benchmark by picking securities expected to outperform, funding the purchase by selling one expected to lag. Both add uncertainty over and above the policy portfolio, so the IPS should set risk limits and target payoffs for all three activities: strategic allocation, tactical allocation, and security selection.
Active versus passive
For each class, risk budgeting forces a choice between active management, using security selection to generate return, and passive management. Unlike systematic exposure, security selection carries no long-run reward for risk: it is a zero-sum game in which every investor in a class competes for a limited set of mispriced assets, so gross returns across all participants average to the market return. Because active managers trade more and pay for ideas, the average active manager underperforms the market net of costs. This does not mean no manager can beat a benchmark with some consistency, nor that every passive manager matches the index, since a high-turnover index raises trading costs for the passive manager too.
How much value security selection can add depends on the manager’s skill and on how informationally efficient the market is. In an efficient market, prices quickly reflect new information, which requires many investors trading on rational expectations with similar models and equal access to information; US large-cap equities are a good example. Some regional bond and equity markets lack the systems to spread information promptly, leaving inefficiencies that a skilled manager can exploit. Sometimes the active-versus-passive choice is made implicitly by including a class at all: markets such as non-listed real estate and infrastructure are so illiquid that gaining diversified exposure is impossible without security selection.
Drift and rebalancing
As returns accrue, the class weights drift away from the policy weights. Periodically, or once a threshold deviation (the bandwidth or corridor) is breached, the portfolio should be rebalanced back toward policy. The set of rules governing this is the rebalancing policy, and even without a formal risk budget it is an important piece of risk management.
A European charity starts the year with weights exactly at policy: European equities 30 percent, international equities 15 percent, European government bonds 20 percent, corporate bonds 20 percent, and cash 15 percent, each with a corridor of plus or minus 2 percent. Over the first six months the class returns are 15.0 percent, 10.0 percent, 0.5 percent, 1.5 percent, and 1.0 percent respectively, lifting the ending weights to 32.4, 15.5, 18.9, 19.1, and 14.2 percent.
| Asset class | Policy weight I | Starting weight II | Difference III | Period return IV | Contribution III × IV |
|---|---|---|---|---|---|
| European equities | 30.0% | 32.0% | 2.0% | −9.0% | −0.18% |
| International equities | 15.0% | 15.5% | 0.5% | −6.0% | −0.03% |
| European gov’t bonds | 20.0% | 18.9% | −1.1% | 4.0% | −0.05% |
| Corporate bonds | 20.0% | 19.1% | −0.9% | 4.0% | −0.04% |
| Cash and money market | 15.0% | 14.6% | −0.4% | 2.0% | −0.01% |
| Total | 100.0% | −2.0% | −0.30% |
Once weights and managers are in place, execution continues by monitoring the managers and the results of both tactical and strategic allocation. When weights move outside their corridors, money shifts from oversized classes to those that have fallen short, and the managers and the SAA itself are reviewed against the monitoring results. Changing capital market expectations, or changes in the client’s circumstances and objectives, can then feed back into an adjusted SAA.
The first is the spread of exchange traded funds (ETFs) alongside algorithm-based advice, or robo-advice. ETFs track an index or sub-index, trade easily, and are relatively cheap, so retail investors can gain fast, inexpensive, liquid exposure across equities, fixed income, and commodities, while robo-advice has cut the cost of assembling a diversified portfolio. The second responds to criticism that modern portfolio theory is too sensitive to small errors in return forecasts and to unstable correlations and volatilities. In answer, some practitioners weight asset classes by their contribution to risk, an approach called risk parity investing. Proponents argue that a conventional portfolio with 60 percent or more in equities carries far more risk than the weight suggests, since equities are much more volatile than bonds; critics counter that the strong showing of risk parity after the 2007 to 2009 global financial crisis owed much to a long decline in interest rates that flattered bonds.
Carrying out a responsible investing policy touches both the strategic asset allocation and the construction that follows. The ESG approaches described earlier need explicit instructions for managers on selecting securities, exercising shareholder rights, and choosing strategies. The forces driving environmental and social integration include the scarcity of natural resources, the physical effects of climate change, shifts in the global economy and demographics, questions of diversity and inclusion, and the growth of social media. Many of these can be expressed through structured, numeric data, such as executive pay, the carbon footprint, staff turnover, absenteeism, and lost-time injuries and fatalities. Disclosure of such data has often been voluntary, but that is shifting as stock exchanges and regulators across developed and emerging markets set sustainability disclosure guidelines. Frameworks such as the Principles for Responsible Investment, the UN Global Compact, and the OECD Guidelines for Multinational Enterprises help asset owners form responsible investing policies, under which they may exclude or engage with companies, or require their managers to weigh these issues.
Narrowing the universe through negative screening affects expected returns and risk, so managers, whether active or passive, prefer to be measured against a benchmark that reflects the reduced universe, and such benchmarks and vehicles are increasingly available, especially in equities. As ESG integration spreads, more asset owners actually expect their managers to beat the regular benchmarks, viewing integration as an enhancement to traditional analysis rather than a wholly new way to invest.
Mountain Materials, a fictitious South African cement maker, is among the largest in the world and a significant emitter of greenhouse gases. South Africa began pricing carbon on its largest emitters in 2019, with an average price across seven pilot markets of between 5 and 15 US dollars per ton of carbon dioxide. Mountain has only ad hoc emissions initiatives and no firmwide program to limit energy use or carbon, and its handling of toxic emissions and of employee health and safety lags industry best practice. Ved Disha, CFA, compares his base-case internal rate of return before and after integrating these environmental and social risks.
| Scenario | Bear case | Base case | Bull case |
|---|---|---|---|
| IRR before integration | −14.0% | 15.0% | 27.0% |
| IRR after integration | −20.0% | 5.0% | 22.0% |
Engagement, themes, and impact
Shareholder engagement needs good cooperation between client and manager, since it is time-consuming and usually serves the client’s interest more than the manager’s. The two must be clear about voting rights, filing shareholder proposals, and talking to management. Engagement and voting may be delegated to the manager under its stewardship policy, handled by a proxy agent on the client’s own policy, or kept by the client directly, whether through individual or collaborative engagement. Collaborative initiatives have grown because collective action more easily wins corporate attention; Climate Action 100+, for example, aims to make the world’s largest corporate emitters act on climate change, targeting over 100 systemically important emitters responsible for two-thirds of yearly global industrial emissions, plus more than 60 further companies well placed to drive the clean energy transition.
Thematic investing, especially in liquid classes, means finding specialist managers who can spot the right opportunities, and because a thematic allocation tilts the whole class toward one theme, the manager should show its effect on the total risk-return profile. Impact investing goes further, selecting opportunities specifically for their intended positive environmental and social effect. The cost and effort of limiting the universe might seem to hurt returns, yet proponents argue that better governance and the avoidance of material risks can help them. Empirical research on ESG in equities remains mixed on the return impact, but adoption keeps rising: nearly 31 trillion US dollars of assets were dedicated to responsible investment mandates at the start of 2018, and ESG integration is now widely used across mainstream funds rather than confined to client-specific accounts.