Investment Policy, Asset Allocation and Portfolio Construction

Introduction

Portfolio construction converts an investor’s goals and constraints into a disciplined investment program. Security selection matters, but the allocation among equities, fixed income, cash, property and other assets usually explains much of a portfolio’s variation in returns and risk. Allocation decisions should therefore begin with an investment policy statement (IPS), not with forecasts of the next winning security.

An IPS defines objectives, risk tolerance, time horizon, liquidity needs, taxes, legal restrictions and governance responsibilities. These inputs guide strategic asset allocation, permissible tactical adjustments, portfolio optimization, implementation and rebalancing. Scenario analysis and stress testing then assess whether the proposed portfolio remains acceptable under adverse conditions.

Learning Objectives

After studying this chapter, students should be able to:

• prepare an IPS for an individual or institutional investor;
• distinguish risk capacity from risk willingness;
• translate objectives and constraints into an asset allocation;
• apply strategic, tactical and dynamic allocation approaches;
• calculate portfolio return, variance and minimum-variance weights;
• recognize the conditions and limitations of optimization models;
• design rebalancing rules that account for costs and taxes;
• conduct scenario analysis and stress testing; and
• evaluate governance, regulatory and ethical considerations in portfolio construction.

1. The Investment Policy Statement

An investment policy statement is a written agreement governing how assets will be managed. It promotes consistency, accountability and continuity, especially when markets become volatile or decision-makers change.

A sound IPS should be specific enough to guide decisions but flexible enough to remain useful as capital-market conditions change.

1.1 Major components

1.2 Return objectives

A return objective may be absolute, relative or liability-based. An absolute objective might require a 6% nominal annual return. A relative objective could seek to exceed a specified benchmark by 1% per year. A liability-based objective seeks sufficient return to finance future obligations.

If an investor requires real return r_real and expects inflation i, the exact nominal return is:

Thus:

For a 4% real return and 2% inflation, the required nominal return is (1.04)(1.02)-1=0.0608, or 6.08%. Simply adding the rates gives 6%, which is an approximation that omits the interaction term.

Objectives should also specify whether returns are measured before or after fees, taxes and inflation.

1.3 Risk tolerance: capacity and willingness

Risk tolerance has two distinct dimensions:

• Risk capacity is the investor’s financial ability to sustain losses without jeopardizing essential goals.
• Risk willingness is the investor’s psychological readiness to accept volatility and loss.

The effective tolerance should generally reflect the more restrictive dimension. A wealthy investor may possess high capacity but low willingness. A young investor may be willing to take risk but have low capacity because of near-term obligations or unstable income.

Risk should not be described only by standard deviation. An IPS may also specify maximum drawdown, shortfall probability, loss under a defined stress scenario, tracking error or the probability of failing to meet liabilities.

1.4 Time horizon and liquidity

The time horizon is the period over which assets can remain invested before they are needed. Many investors have several horizons. A household may need emergency funds immediately, education funds in five years and retirement assets in thirty years.

Liquidity constraints include expected withdrawals, collateral requirements, insurance claims, pension payments and emergency reserves. Assets needed soon should not depend on the sale of volatile or illiquid investments at unfavorable prices.

A long horizon may increase risk capacity, but it does not automatically justify high equity exposure. The certainty and timing of liabilities, employment risk and tolerance for interim losses remain important.

1.5 Tax, legal and unique constraints

Taxes affect the timing and location of investments. Interest, dividends and realized capital gains may face different tax rates. Portfolio turnover can accelerate tax liabilities, while tax-advantaged accounts may be appropriate for tax-inefficient assets. Tax rules vary by jurisdiction and must be verified rather than assumed.

Legal and regulatory constraints may include fiduciary duties, concentration limits, solvency rules, eligible-asset requirements, prudent-person standards and restrictions on derivatives or leverage. Insurers and pension funds must also coordinate investments with the duration, currency and uncertainty of their liabilities.

Unique circumstances include responsible-investment requirements, prohibited industries, legacy holdings, family control positions, currency exposure and restrictions arising from employment.

2. From the IPS to an Investable Opportunity Set

The investment team converts IPS requirements into eligible asset classes, benchmarks and allocation ranges. An asset class should normally exhibit internally similar economic characteristics, behave differently from other classes and be represented by an investable benchmark.

Common categories include domestic and international equities, government and corporate bonds, inflation-linked securities, cash, property and selected alternative assets. Private assets may improve diversification but create valuation, liquidity and governance challenges.

Capital-market assumptions are estimates of expected return, volatility and correlation. They should be stated for a consistent horizon, currency and return basis. Nominal and real returns, or pre-tax and after-tax returns, should not be mixed in one optimization.

3. Portfolio Return and Risk

For a portfolio of n assets, expected return is the weighted average of asset expected returns:

Here, w_i is the proportion invested in asset i, E(R_i) is its expected return per period and ∑w_i=1 for a fully invested portfolio. Negative weights indicate short positions when permitted.

Portfolio variance incorporates both individual variances and covariances:

In this expression, σ_p² is portfolio return variance, w is the weight vector and Σ is the covariance matrix. All returns and covariance estimates must use the same periodicity. Portfolio standard deviation is σ_p=√(σ_p²).

For two assets:

The correlation coefficient ρ₁₂ lies between −1 and +1. Diversification arises when assets are not perfectly positively correlated, although correlations may increase during crises.

3.1 Global minimum-variance portfolio for two assets

Where w₂=1-w₁, differentiating portfolio variance with respect to w₁ and setting the derivative to zero gives the unconstrained minimum-variance weight:

The denominator equals Var(R₁-R₂). It is positive unless the two returns differ only by a constant with probability one. However, a positive denominator alone does not guarantee an admissible interior solution. The calculated weight must also lie within the permitted bounds. Under long-only constraints, an interior solution requires 0<w₁^*<1. If it lies outside the feasible interval, the constrained minimum occurs at a boundary.

Worked example. Equity has volatility 18%, bonds have volatility 8%, and their correlation is 0.20. Their covariance is:

The unconstrained equity weight is:

The bond weight is 89.35%. Because both weights lie between zero and one, this is also the long-only minimum-variance solution. Substitution gives a portfolio standard deviation of approximately 7.76%, slightly below the 8% volatility of bonds because of diversification.

4. Strategic Asset Allocation

Strategic asset allocation (SAA) establishes long-term target weights and permissible ranges. It reflects enduring objectives, liabilities, risk tolerance and long-term capital-market assumptions rather than short-term forecasts.

A common mean–variance formulation is:

Here, μ is the expected-return vector, λ>0 represents preference for return relative to variance, and w_L and w_U are lower and upper bounds. Other constraints may address liquidity, turnover, currency, carbon exposure, leverage or liability duration.

Alternatively, an investor may minimize variance subject to a required return:

R_T is the target expected return. Feasibility must be checked: an excessively high target may be impossible under the imposed bounds.

Figure: A strategic portfolio should be feasible and consistent with the investor’s preferred risk–return trade-off.

4.1 Limits of mean–variance optimization

Optimization is highly sensitive to expected-return estimates. Small changes in inputs may produce large changes in weights. Historical estimates may also be unreliable when market structures change.

Practical responses include:

• imposing realistic weight and turnover constraints;
• using shrinkage or robust covariance estimates;
• testing several plausible capital-market assumptions;
• resampling or averaging allocations across models;
• incorporating liabilities and downside risk; and
• applying judgment through documented governance rather than manipulating inputs to obtain a desired answer.

Risk contribution analysis can reveal hidden concentration. The marginal contribution of asset i to portfolio volatility is:

Its total contribution is w_iMCR_i, and the contributions sum to σ_p, provided σ_p>0. A small capital weight can make a large risk contribution when the asset is volatile or highly correlated with the rest of the portfolio.

5. Tactical and Dynamic Asset Allocation

5.1 Tactical asset allocation

Tactical asset allocation (TAA) involves temporary deviations from strategic targets to exploit perceived short- or medium-term opportunities. For example, a manager may modestly overweight government bonds when recession risk appears underpriced.

TAA requires explicit limits, a decision horizon, benchmarks and rules for closing positions. It creates forecasting risk, transaction costs, tax consequences and potential conflict with the IPS. Performance should be separated from the return generated by the strategic policy.

5.2 Dynamic asset allocation

Dynamic allocation changes exposure systematically as portfolio wealth, market values or liabilities change. Two contrasting rules are constant-mix strategies and constant proportion portfolio insurance (CPPI).

A constant-mix strategy maintains a fixed risky-asset proportion. It buys risky assets after declines and sells after increases. It can benefit from oscillating markets but performs poorly relative to buy-and-hold in strongly trending markets.

CPPI seeks to preserve a floor while retaining upside participation. Define:

• V_t: portfolio value at time t;
• F_t: desired floor value at time t;
• C_t=V_t-F_t: cushion;
• m: non-negative multiplier;
• E_max: maximum permitted risky exposure.

The risky exposure is:

The zero lower bound is essential. If V_t<F_t, the formula assigns zero—not negative—risky exposure unless a separate short-selling strategy is explicitly intended. For an unleveraged portfolio, E_max≤V_t. The remaining value is invested in the defensive asset.

Example. A CU10 million portfolio has a CU9 million floor, multiplier 3 and risky-exposure cap of CU10 million. The cushion is CU1 million, so risky exposure is CU3 million. If the risky asset immediately loses 20% while the defensive asset is unchanged, portfolio value becomes CU9.4 million. The new cushion is CU0.4 million and risky exposure is reduced to CU1.2 million.

CPPI does not guarantee the floor. A sudden market gap, delayed trading, transaction costs, illiquidity or default of the defensive asset can cause a breach. The floor must also be defined consistently—for example, as a current amount or the present value of a future guarantee.

6. Portfolio Construction in Practice

Portfolio construction integrates model output with implementation realities. A practical sequence is:

1. verify objectives and constraints;
2. define eligible assets and benchmarks;
3. establish capital-market assumptions;
4. generate feasible allocations;
5. test risk, liquidity and liability compatibility;
6. select instruments and implementation vehicles;
7. document rebalancing and monitoring rules; and
8. obtain appropriate approval.

Implementation choices include active or passive management, pooled funds or direct securities, currency hedging and derivative overlays. Derivatives may efficiently adjust exposure but require collateral, valuation expertise and counterparty controls.

6.1 Goal-based and liability-aware construction

A single mean–variance portfolio may be unsuitable when goals differ significantly. Goal-based investing assigns separate sub-portfolios to emergency reserves, near-term expenditures and long-term growth. Institutional investors apply related techniques through cash-flow matching, duration matching and liability-driven investing.

Matching duration reduces sensitivity of the surplus to small parallel interest-rate changes, but it does not eliminate yield-curve, credit, inflation or liquidity risk.

7. Rebalancing Policy

Market movements cause actual weights to depart from strategic targets. Rebalancing restores intended risk exposures by buying underweight assets and selling overweight assets.

Three common approaches are:

• Calendar rebalancing: trade at fixed intervals, such as annually.
• Tolerance-band rebalancing: trade when a weight breaches an allowed range.
• Hybrid rebalancing: monitor periodically but trade only after a band is breached.

Suppose a CU1,000 portfolio begins at 60% equity and 40% bonds. Equity rises to CU720 while bonds fall to CU360, giving total value CU1,080. Current weights are 66.67% and 33.33%. Restoring the 60:40 target requires equity of CU648 and bonds of CU432. The investor therefore sells CU72 of equity and buys CU72 of bonds, ignoring costs and taxes.

Frequent rebalancing maintains tighter risk control but increases trading costs and potentially taxable gains. Cash flows, dividends and new contributions should first be directed toward underweight assets. Material tax consequences should be evaluated before sales.

8. Scenario Analysis and Stress Testing

Scenario analysis estimates portfolio outcomes under coherent combinations of economic and market conditions. Stress testing focuses on severe adverse events, including events outside the historical sample.

For scenario s, portfolio return is:

If mutually exclusive scenarios have probabilities p_s summing to one, probability-weighted expected return is:

Consider a portfolio with 60% equity, 30% bonds and 10% cash:

The probability-weighted return is 0.25(10.1%)+0.50(5.6%)+0.25(-12.4%)=2.225%. The recession result is especially important if the IPS permits no more than a 10% one-year loss; the allocation would then fail that policy test even if its average expected return seemed satisfactory.

Useful stresses include equity crashes, interest-rate shocks, credit-spread widening, inflation surprises, currency depreciation, liquidity freezes and simultaneous correlation increases. Reverse stress testing begins with an unacceptable outcome and identifies combinations of events that could produce it.

Scenarios are not forecasts. Their value lies in exposing vulnerabilities and encouraging contingency planning.

9. Illustrative Case: Constructing a Household Portfolio

Background

This fictional case concerns Arun, aged 40, with CU1,000,000 of investable assets. He requires CU60,000 as an emergency reserve, CU200,000 for a home payment in four years and long-term retirement growth. He has moderate risk willingness and states that a one-year loss above 15% would be unacceptable. The figures below ignore personal tax details; professional implementation would use jurisdiction-specific after-tax assumptions.

Problem

A single aggressive portfolio could endanger the four-year goal. Conversely, placing all assets in defensive securities would probably fail to support long-term purchasing power.

Analysis and approach

Assume a secure four-year instrument yields 4% annually after applicable costs and taxes. The current amount required to finance the home payment is:

The emergency reserve is placed in liquid cash. This leaves approximately CU769,040 for retirement. An initial 65% retirement-equity allocation caused the total portfolio to lose more than 15% under a stress of −30% equity, −5% bonds and 0% cash.

The retirement allocation was therefore revised to 54% equity, 41% bonds and 5% cash. The total portfolio becomes approximately:

Rounding explains the minor total difference. Under the defined stress, estimated loss is:

The allocation therefore remains within the stated 15% stress-loss limit, although actual losses could differ.

Assumed long-run nominal returns of 8% for equity, 4.5% for bonds and 2% for cash produce an expected portfolio return of approximately 5.71%. If retirement planning requires 6.08%, increasing risk would conflict with the stress constraint. The appropriate response is not to disguise this conflict through optimistic assumptions. Arun may need to raise contributions, delay retirement, reduce the spending goal or reconsider the loss limit.

Solution and governance

The IPS establishes the calculated allocation, ±5 percentage-point asset-class bands, quarterly monitoring and annual formal review. Near-term goal assets cannot be used to fund tactical equity positions. Rebalancing first uses new contributions and cash flows to reduce costs and taxes.

Learning points

• Separate near-term goals from long-term risk capital.
• Test allocations against explicit loss limits rather than relying only on volatility.
• Treat optimization as decision support, not as a substitute for judgment.
• Resolve infeasible objectives by changing goals, savings or constraints—not by manipulating forecasts.

10. Governance, Ethics and Regulation

Portfolio decisions should be traceable to approved policy. Responsibilities for assumption setting, execution, custody, valuation, risk monitoring and performance measurement should be separated where practical. Conflicts of interest, commissions and related-party transactions require disclosure and control.

Fiduciaries should act with loyalty, care and impartiality. Ethical portfolio management also requires fair client risk assessment, honest communication of uncertainty and avoidance of misleading back-tests. Sustainability preferences should be translated into measurable rules rather than vague labels.

Regulation varies across jurisdictions and investor types. Banks, insurers, pension funds and collective investment schemes may face different capital, liquidity, valuation and concentration requirements. Compliance is a binding constraint, not an optimization preference.

11. Technology, AI and Contemporary Developments

Portfolio platforms can integrate holdings, tax lots, factor exposures, liabilities and scenario analysis. Artificial intelligence may assist with forecasting, document review, anomaly detection and personalized reporting. However, AI-generated assumptions can inherit biased data, produce unstable recommendations and lack transparent economic reasoning.

Human oversight remains necessary. Inputs, model versions and overrides should be documented. Sensitive client data require privacy and cybersecurity controls. Current developments include direct indexing, tax-aware optimization, climate scenario analysis, private-market allocation and greater use of robust rather than purely historical estimates.

12. Challenges and Best Practices

Major challenges include estimation error, changing correlations, illiquidity, model risk, transaction costs, taxes and behavioral responses to losses. Expected returns are especially difficult to estimate, and apparently precise allocations may create false confidence.

Best practice is to use transparent assumptions, multiple models, realistic constraints and adverse scenarios. Policy ranges should be wide enough to avoid unnecessary trading but narrow enough to preserve intended risk. The IPS should be reviewed after major changes in liabilities, wealth, regulation or client circumstances—not merely after poor market performance.

Chapter Summary

An investment policy statement translates objectives, risk tolerance, time horizon, liquidity needs and tax, legal and unique constraints into an accountable investment program. Strategic allocation establishes long-term exposures, tactical allocation permits controlled temporary deviations, and dynamic strategies alter exposure systematically as market values or cushions change.

Mean–variance analysis quantifies the interaction of return, variance and covariance, but optimized weights are valid only when assumptions and feasibility conditions are satisfied. In particular, an unconstrained minimum-variance solution must fall within the permitted bounds to be an interior constrained solution. Portfolio construction must also account for liabilities, liquidity, taxes, implementation costs and governance.

Rebalancing maintains intended risk exposures, while scenario analysis and stress testing reveal vulnerabilities that average forecasts may conceal. Effective portfolio management combines quantitative discipline with professional judgment, ethical conduct and continuous monitoring.

Key Terms

Asset allocation: Distribution of capital among asset classes.

Capital-market assumptions: Estimates of asset returns, volatility and correlations.

Constant-mix strategy: Dynamic strategy that restores a fixed risky-asset proportion.

CPPI: Strategy linking risky exposure to a multiplier times the portfolio cushion, subject to exposure bounds.

Cushion: Difference between portfolio value and its designated floor.

Efficient frontier: Feasible portfolios offering the highest expected return for each risk level.

Investment policy statement: Written document governing objectives, constraints and investment procedures.

Liquidity constraint: Requirement to meet cash needs without unacceptable loss or delay.

Mean–variance optimization: Selection of portfolio weights using expected returns, variances and covariances.

Minimum-variance portfolio: Feasible portfolio with the lowest return variance.

Rebalancing: Trading that restores portfolio weights toward policy targets.

Risk capacity: Financial ability to bear investment loss.

Risk contribution: Amount of total portfolio risk attributable to an asset or position.

Risk willingness: Psychological readiness to accept uncertainty and loss.

Scenario analysis: Evaluation of portfolio results under coherent possible conditions.

Strategic asset allocation: Long-term target allocation based on objectives and constraints.

Stress testing: Evaluation under severe adverse market conditions.

Tactical asset allocation: Temporary deviation from strategic weights based on market views.

Review Questions

Short Answer Questions

1. What is the primary purpose of an investment policy statement?
2. Distinguish risk capacity from risk willingness.
3. Why should portfolio liquidity be evaluated relative to liabilities?
4. What is the difference between strategic and tactical asset allocation?
5. Why can CPPI fail to preserve its floor during a market gap?

Long Answer Questions

1. Explain how objectives, time horizon, liquidity, tax and legal constraints influence asset allocation.
2. Discuss the assumptions, uses and limitations of mean–variance optimization.
3. Compare calendar, tolerance-band and hybrid rebalancing policies.
4. Explain how scenario analysis and stress testing support portfolio governance.
5. Evaluate the benefits and risks of using AI and optimization technology in portfolio construction.

Analytical / Application Questions

1. An investor allocates 50% to an asset returning 9%, 30% to an asset returning 5% and 20% to cash returning 2%. Calculate expected portfolio return and discuss whether the calculation measures downside risk.
2. A CU8 million CPPI portfolio has a CU6.5 million floor, multiplier 4 and maximum risky exposure of CU5 million. Calculate risky exposure. Recalculate it if portfolio value falls below the floor.
3. Design an IPS outline for an insurance company with predictable claims, catastrophe exposure, solvency requirements and restrictions on illiquid assets.

Multiple Choice Questions with Answers

Question 1

Which item is a financial ability rather than a psychological preference?

A. Risk willingness

B. Risk capacity

C. Tactical conviction

D. Market sentiment

Correct Answer: B. Risk capacity

Question 2

A portfolio invests 60% in an asset with an expected return of 8% and 40% in an asset with an expected return of 3%. What is its expected return?

A. 4.8%

B. 5.0%

C. 6.0%

D. 11.0%

Correct Answer: C. 6.0%

Question 3

Two assets have volatilities of 20% and 10% and correlation of 0.30. What is their covariance?

A. 0.006

B. 0.030

C. 0.060

D. 0.600

Correct Answer: A. 0.006

Question 4

For a long-only two-asset minimum-variance portfolio, when is the unconstrained stationary solution an admissible interior solution?

A. Whenever the expected returns are positive

B. Whenever correlation is below one

C. Whenever the denominator in the weight formula is non-negative

D. When the denominator is positive and the calculated weight lies strictly between zero and one

Correct Answer: D. When the denominator is positive and the calculated weight lies strictly between zero and one

Question 5

A CPPI portfolio has value CU12 million, floor CU10 million, multiplier 3 and maximum risky exposure CU12 million. What is its risky exposure?

A. CU2 million

B. CU6 million

C. CU10 million

D. CU12 million

Correct Answer: B. CU6 million

Question 6

Under the chapter's CPPI rule, what risky exposure applies when portfolio value is below the floor and no separate short-selling strategy is intended?

A. Zero

B. The negative cushion

C. The full portfolio value

D. The floor value

Correct Answer: A. Zero

Question 7

Which statement best describes tactical asset allocation?

A. Permanent selection of long-term policy weights

B. Temporary deviation from strategic weights based on market views

C. Automatic maintenance of a fixed portfolio floor

D. Matching every liability with one security

Correct Answer: B. Temporary deviation from strategic weights based on market views

Question 8

A CU1,080 portfolio is worth CU720 in equity and CU360 in bonds. How much equity must be sold to restore a 60% equity target, ignoring costs?

A. CU36

B. CU60

C. CU72

D. CU120

Correct Answer: C. CU72

Question 9

What does reverse stress testing primarily identify?

A. The most likely expected return

B. Events capable of producing a predefined unacceptable outcome

C. The historical minimum-variance portfolio

D. The cheapest rebalancing frequency

Correct Answer: B. Events capable of producing a predefined unacceptable outcome

Question 10

If the required real return is 4% and expected inflation is 2%, what is the exact required nominal return?

A. 2.00%

B. 6.00%

C. 6.08%

D. 8.00%

Correct Answer: C. 6.08%