A portfolio manager estimates expected returns and optimises the allocation across asset classes, and separately has to hold a client to the agreed plan through a violently volatile market. How does the division of portfolio management into a science and an art assign those two tasks?

Estimating returns and optimising the allocation belongs to the art; holding to the plan under emotional pressure belongs to the science

Estimating returns and optimising the allocation belongs to the science; holding to the plan under emotional pressure belongs to the art

Both belong to the science, because a client's behaviour under stress can itself be modelled and allowed for inside the optimisation

Both belong to the art, because the mathematics only describes decisions the manager has already reached by judgment

One adviser runs personalised accounts for retail clients; elsewhere a team of analysts, portfolio managers and risk officers runs a multi-billion-dollar pension portfolio. What is the relationship between the objectives governing the two mandates?

The institutional portfolio adds transparent communication as a fourth objective, which a retail account meets through its periodic statements instead

The retail account is governed by suitability while the institutional portfolio is governed by efficient implementation, the objectives diverging with the client type

The same four objectives govern both mandates, so the two differ in scale and complexity rather than in the principles applied

The objectives are the same in name but reversed in priority, the institutional mandate putting efficient implementation ahead of understanding the client

A manager has finished assessing the client and has set the target allocation across broad asset classes. Which activities of the portfolio management process come next, and which comes last?

Security selection, then implementation, with performance evaluation last

Implementation, then security selection, with performance evaluation last

Monitoring and rebalancing, then security selection, with implementation last

Security selection, then performance evaluation, with rebalancing last

Performance evaluation establishes whether a manager delivered what was committed. Beyond that accountability, what second function does the evaluation phase serve?

A marketing function: it produces the track record presented to prospective clients, in a form standardised across every portfolio run

A learning function: it locates the sources of out- and underperformance so that later investment decisions can be informed by them

A compliance function: it evidences that each trade was executed on the terms the client's own account agreement required of the firm

A pricing function: it establishes the values at which the portfolio's holdings are carried for the reporting of the period just ended

Mean-variance optimisation has several documented weaknesses. One of them concerns the preferences it assumes rather than the inputs it is fed. What is that weakness?

It assumes investors value each asset class for itself rather than for its contribution to the whole, which conflicts with diversifying

It assumes investors measure risk by the probability of missing a stated goal, which conflicts with its own use of variance

It rests on a quadratic utility function, which may not represent real investors who exhibit loss aversion

It assumes investors revise their expected returns as prices move, which conflicts with holding to a fixed strategic allocation

A firm intends to keep using mean-variance optimisation while limiting the damage its known weaknesses do. Which combination of practical measures does portfolio construction use for that purpose?

Replacing expected returns with realised returns, replacing variance with downside deviation, and leaving the resulting weights unconstrained

Longer estimation windows, monthly rather than daily return data, equal weighting of the asset classes, and judgment applied to the resulting weights

Dropping asset classes whose returns are hardest to estimate, raising the risk-free rate used, and re-running the optimiser on each forecast

Constraints limiting extreme allocations, shrinkage techniques, scenario analysis across possible environments, and qualitative judgment

Setting a strategic asset allocation draws on three ingredients. Which three?

Capital market assumptions on each class's expected return and risk, optimisation techniques seeking the efficient combination, and judgment on what suits the client

Historical returns realised by each asset class, the return target the client has stated, and the fee the manager charges for running each class

The benchmark's current weights, the manager's tracking error budget, and the custodian's valuation schedule for the classes that trade infrequently

A survey of what peer institutions allocate, the consultant's ratings of each manager, and the trustees' judgment about complexity in the reporting

What does managing a portfolio's liquidity risk consist of in practice?

Cash equal to a year of expected withdrawals, refusal of any holding without a daily quoted price, and a custodian guarantee of redemptions

A minimum allocation to liquid instruments, a cap on the illiquid proportion set against the client's liquidity needs, and stress-testing under adverse scenarios

Concentration limits on each issuer and sector, exclusion of securities below a rating threshold, and monitoring of issuers for deterioration

Matching every holding to a dated liability, selling the least liquid position first in a redemption, and avoiding classes with wide spreads

A manager raises the portfolio's exposure to the momentum factor and cuts its exposure to value, on an assessment of when each factor's premium is most likely to be realised. Which source of active return is being pursued, and what does its success turn on?

Asset allocation alpha, whose success turns on the asset classes that were overweighted subsequently outperforming those underweighted

Security selection alpha, whose success turns on identifying individual securities priced below what the analysis says they are worth

Factor timing alpha, whose success turns on the assessment of when each factor's premium will be realised being right

Market timing alpha, whose success turns on the portfolio's total market exposure being raised and cut ahead of the market's turns

Two funds charge the same active management fee. Fund A maintains a composition very similar to its benchmark; fund B deviates substantially from it. What did the research on active share find about funds like A and B?

Both kinds showed persistence, the research finding that the level of the fee rather than the size of the deviation predicted which funds repeated

Neither kind showed persistence, the research finding active share uncorrelated with any measure of subsequent performance at all

Funds like A showed the greater persistence, their smaller deviation producing a steadier margin over the benchmark year after year

Funds like B showed greater persistence of outperformance, while a fund like A charging active fees is a closet index fund

A pension fund and a high-net-worth individual are both clients of the same firm. On which dimensions does the institutional mandate typically differ from the individual one?

A longer time horizon, greater access to illiquid and alternative strategies, and more sophisticated risk management and reporting requirements

A shorter time horizon, a narrower range of permitted asset classes, and lighter reporting because the trustees receive the same statements

A longer time horizon, a smaller pool of capital under any one policy, and reporting confined to an annual valuation of the whole fund

An identical time horizon, access confined to listed securities, and heavier reliance on the manager's discretion between formal reviews

An individual client wants a portfolio that can be shaped around their own tax position and can exclude particular securities and sectors. What does a separately managed account offer that meets this, and what limits who can use one?

Daily liquidity at a published net asset value, limited by a redemption fee charged to discourage short holding periods

Full customisation to the client's tax position and restrictions, limited by a minimum investment that takes sufficient assets to meet

A share of a pooled portfolio run to a single mandate, limited by the need for that pool to stay inside its stated policy

A tax outcome guaranteed in advance by the custodian, limited by the range of securities the custodian is willing to hold

Before an index is an appropriate basis for evaluating a manager's performance, what must it satisfy?

It must be drawn from the same jurisdiction as the portfolio, rebalanced on the same schedule, and agreed between the manager and the custodian

It must be published daily by an independent provider, carry at least five years of history, and bear a lower management fee than the portfolio

It must be investable as a realistic alternative the manager could have held instead, fixed in advance, and reflective of the manager's universe and strategy

It must match the portfolio's realised volatility, be settled once the period's results are known, and exclude the classes the manager avoided

A benchmark holds 60 percent equities returning 10 percent and 40 percent fixed income returning 2 percent, giving a benchmark return of 6.8 percent. A portfolio held 70 percent equities. What is the allocation contribution from equities?

0.10 percentage points, the difference between the portfolio's equity weight and the benchmark's equity weight read as points of return

1.00 percentage point, the full 10 point overweight multiplied by the benchmark's own 10 percent return on its equity holdings for the period

3.20 percentage points, the amount by which the benchmark's return on equities exceeded the return on the benchmark taken as a whole

0.32 percentage points, a 10 point overweight multiplied by the 3.2 points by which the benchmark's equity return beat the benchmark overall

A portfolio held 30 percent fixed income, whose sleeve returned 1 percent, against a benchmark holding 40 percent fixed income returning 2 percent. Which weight scales the selection contribution from fixed income, and what is that contribution?

The benchmark's 40 percent weight, giving minus 0.40 percentage points on the 1 point shortfall inside the category

The portfolio's own 30 percent weight, giving minus 0.30 percentage points on the 1 point shortfall inside the category

No weight at all, the 1 point shortfall inside the category carrying straight through as minus 1.00 percentage point

The 10 point weight difference between the portfolio and the benchmark, giving minus 0.10 percentage points on the shortfall

Attribution decomposes a portfolio's return against its benchmark into allocation, selection and interaction contributions. What is the relationship between those three contributions, summed across every category, and the excess return itself?

They approximate the excess return, the gap being a residual that grows with the number of categories the portfolio is split into

They come to the excess return exactly, so the decomposition checks itself rather than approximating the result

They exceed the excess return, because the interaction contribution counts again the weight and return differences that it multiplies together

They match the excess return only where the portfolio holds every category the benchmark holds, a residual arising in every other case

On 1 January an account is worth $100,000, and by 30 June it has risen to $110,000. The client pays in a further $100,000 on 1 July, bringing it to $210,000, and at 31 December it stands at $189,000. What is the time-weighted return for the year?

Zero percent, the 10 percent gain over the first half of the year offsetting the 10 percent loss over the second half of it

Minus 7.3 percent, the result of compounding the two sub-period returns once the client's mid-year contribution is added into them

Minus 1.0 percent, the 10 percent first sub-period compounded with the minus 10 percent second sub-period to give a factor of 0.99

Minus 5.5 percent, the $11,000 by which the account fell short divided by the $200,000 the client had put into it in total

A client's custodian values the account only once a year, at 31 December, but it does record the date and amount of every contribution and withdrawal. Which return can be computed for the year, and why?

Neither of them, because every return figure requires the account to be valued immediately before each contribution and every withdrawal it received

The time-weighted return, because a single annual valuation is itself enough to strip out the distorting effect of the client's cash flow timing

Both of them, because the net contributions recorded over the year are themselves enough to reconstruct what each sub-period of the year returned

The money-weighted return, because it needs no more than the dated cash flows with the account's opening and closing value

Active share halves the total of the absolute gaps between each security's portfolio weight and its benchmark weight. Against a benchmark of four securities at 25 percent each, a portfolio holds those same four at 40, 30, 20 and 10 percent. What is its active share, and what does that show?

20 percent, which shows that owning every benchmark security and nothing besides is compatible with a material figure

Zero, since the portfolio holds exactly the securities the benchmark holds and holds no security the benchmark does not

40 percent, the sum of the absolute weight gaps, which the measure reports in full where no benchmark security is absent

10 percent, half of the largest single deviation, which is what the measure reduces to when no benchmark name is missing

Portfolio weights sum to 100 percent and benchmark weights sum to 100 percent. What does that imply about the largest value the sum of the absolute weight differences can take, and what is the halving in the active share formula doing?

The sum cannot exceed 100 points, and halving it converts that total into the average deviation per security the portfolio actually holds

The sum cannot exceed 200 points, and halving it stops each deviation being counted twice, as an overweight and again as the matching underweight

The sum has no upper limit at all, and halving it scales the result so that a typical active portfolio reads somewhere near 50 percent

The sum cannot exceed 200 points, and halving it allows for the two sides of each trade needed to move from benchmark weights to portfolio weights

Portfolio A puts 25 percentage points of overweight into one security and carries its matching 25 points of underweight thinly across the rest of the benchmark. Portfolio B is 2.5 points overweight in each of ten securities, its matching underweight spread the same way. What does comparing the two active share figures establish?

That B reads the higher figure, ten separate deviations contributing more absolute difference in total than one large deviation can contribute

That A reads the higher figure, the measure weighting each deviation by the size of the single position within which that deviation happens to sit

Nothing about how either deviation is distributed, the two figures coming out identical while A's result turns on one company and B's on ten

That both portfolios are benchmark-like, a figure assembled from individual deviations of 25 points or less being too small to mark an active portfolio

A manager buys a position and later sells it. Each side of the trade pays 10 basis points of commission and causes 20 basis points of market impact, and crossing the bid-ask spread over the round trip costs 30 basis points. What does the round trip cost?

30 basis points, the spread alone, commission and impact being recovered within the price at which the position is later sold

60 basis points, being 10 of commission, 30 of spread and 20 of impact, each component counted once over the round trip

50 basis points, being 20 of commission and 30 of spread, the market impact already being contained within the spread crossed

90 basis points, being 20 of commission, 30 of spread and 40 of market impact

Two managers face the same cost per round trip and reach the same result before costs. One replaces the portfolio once a year, the other twice. How do their annual implementation costs compare, and what follows?

The second manager's annual cost is double the first's, the cost per round trip being multiplied by the number of replacements

The two annual costs are equal, the cost falling on the portfolio once a year rather than on each replacement of its holdings

The second manager's annual cost is higher but short of double, the second replacement of the year costing less than the first did

The second manager's annual cost is half as much again, only the market impact component recurring on the second replacement

A manager with a large order can work it in pieces over several days instead of executing it at once. As that schedule is stretched out, what happens to the two costs involved, and what follows for the cheapest schedule?

Both costs fall as the schedule lengthens, so the cheapest schedule is the slowest one the manager's mandate will permit him to run

Market impact falls while exposure to the price moving away rises, so the cheapest schedule is neither the fastest nor the slowest

Both costs rise as the schedule lengthens, so the cheapest schedule is to execute the whole order in one transaction at the outset

Market impact rises while exposure to the price moving away falls, so the cheapest schedule is the fastest one the venue allows

A $1,000,000 portfolio holds one position worth $400,000 and another worth $50,000. What does a 25 percent fall in the larger position cost the portfolio, and what is the most the smaller position could ever cost it?

2.5 percent of the portfolio, the fall applying to the portfolio rather than to the position, the smaller one being capped at 5 percent

10 percent of the portfolio, while the smaller position could cost it more than that if its own shares are far the more volatile

10 percent of the portfolio, while the smaller position cannot cost it more than 5 percent even if it becomes entirely worthless

4 percent of the portfolio, the fall applying to the position's share of the whole, the smaller one being capped at 1.25 percent

A $1,000,000 portfolio holds one company's shares worth $400,000 and other holdings worth $600,000. The concentrated position halves in value and does not recover. What must the other holdings return to bring the portfolio back to $1,000,000?

50 percent, the $200,000 shortfall measured against the $400,000 the concentrated position was worth before it fell

25 percent, the $200,000 shortfall measured against the $800,000 the portfolio is worth once the position has halved

20 percent, the $200,000 shortfall measured against the $1,000,000 the portfolio has to be restored to in the end

About 33 percent, the $200,000 shortfall having to be earned on the $600,000 of other holdings that remain

Two clients each hold a large single position in one company's shares. One of them works for that company; the other has no connection to it. Why does the first client's exposure to that company exceed what the position's portfolio weight records?

A reverse at that company can cut the share price and the client's salary and bonus together, so the capital loss and the income loss arrive at once

Shares received from an employer are carried in the account at the price fixed when they were granted rather than at their market value, so the weight recorded understates the position

The client's shares were acquired at a discount to the market, so the gain on them is larger and the tax due on any eventual sale correspondingly heavier

The client's inside knowledge of the company makes the position harder to judge objectively, so its weight in the account is the more likely to drift upward

A foundation distributes 5 percent of its portfolio each year and pays 0.5 percent a year in investment costs, and it wants the portfolio's purchasing power to be the same twelve months later. On an assumed inflation rate of 2.5 percent, what nominal return must the portfolio earn?

5.5 percent, inflation being met out of the growth that the distribution and the costs leave behind in the portfolio

8 percent, the distribution rate, the costs and the inflation rate being simply added together

7.5 percent, the investment costs being met out of the distribution itself rather than having to be earned separately

2 percent, the distribution and the costs being paid out of capital so that only the inflation rate has to be earned

A foundation distributes 5 percent a year. Its portfolio stood at $100,000,000, $100,000,000 and most recently $80,000,000 at three successive year ends. By how much does the distribution fall under a rule of 5 percent of the latest value, and by how much under a rule of 5 percent of the average of the last three year-end values?

By about 7 percent under either rule, the three-year average being the basis against which the 5 percent distribution rate is itself defined

By 20 percent under either rule, the averaging changing when the cut is felt rather than how large the cut on the distribution turns out to be

By 20 percent on the latest value and by about 7 percent on the three-year average, the fall entering one of the three years averaged

By 20 percent on the latest value and by about 13 percent on the three-year average, two of the three years averaged carrying the fall

A portfolio distributes 6 percent a year and earns 4 percent a year after inflation and costs. Given that 0.98 raised to the thirty-fifth power is about 0.49, what has become of the portfolio's purchasing power after thirty-five years?

It is unchanged, the 4 percent real return covering the distribution with the shortfall met out of nominal growth instead

It has fallen by about 70 percent, the 2 percent shortfall accumulating against a base that is itself shrinking each year

It has fallen by about 30 percent, the 2 percent shortfall applying only to the part of the portfolio distributed each year

It has halved, the 2 percent annual shortfall compounding down to about 49 percent of the original real value

Portfolio Management | MPT, Risk & Performance | Financial Regulation Courses

Portfolio Management: Process, Theory, and Performance Evaluation

Portfolio management is the professional discipline of constructing, overseeing, and continuously adjusting a collection of investments to achieve specified financial objectives within defined risk parameters, on behalf of individual investors, institutional clients, or pooled investment vehicles.

It encompasses the full cycle of investment decision-making from the initial assessment of client objectives and constraints through strategic asset allocation, security selection, implementation, ongoing monitoring, and periodic rebalancing, integrating the disciplines of financial analysis, risk management, tax planning, and client communication into a coherent and systematic approach to growing and preserving investment wealth over time.

Portfolio management is both an art and a science, requiring the rigorous application of quantitative analytical frameworks alongside qualitative judgment about business quality, market dynamics, and the behavioural characteristics of clients that cannot be reduced to formulas.

The science of portfolio management provides the mathematical frameworks for measuring risk, estimating expected returns, optimising asset allocation, and evaluating performance. The art of portfolio management involves the judgments about which frameworks are most appropriate in which circumstances, how to weigh competing objectives and constraints, how to communicate effectively with clients during periods of market stress, and how to maintain investment discipline in the face of the emotional pressures that volatile markets inevitably create.

The portfolio management function is performed in many different institutional contexts, from the individual investment adviser managing personalised accounts for retail clients to the team of analysts, portfolio managers, and risk officers managing multi-billion-dollar institutional portfolios for pension funds, endowments, and sovereign wealth funds.

Despite the enormous variation in scale, sophistication, and institutional complexity across these different contexts, the fundamental principles and objectives of portfolio management are universal: understand the client's objectives and constraints, construct a portfolio optimally positioned to achieve those objectives within the constraints, implement and maintain the portfolio efficiently, and communicate transparently with the client about performance, strategy, and the rationale for investment decisions.

The Portfolio Management Process

The portfolio management process follows a structured sequence of activities that together constitute the complete cycle of investment decision-making and implementation.

The client assessment phase is the foundation of the entire process, establishing the investment objectives, risk tolerance, time horizon, liquidity needs, tax circumstances, legal constraints, and unique preferences that will govern all subsequent portfolio decisions.

As described in detail in the Investment Policy Statement article in Section I, this assessment is typically formalised in a written investment policy statement that documents the client's circumstances and the agreed framework for managing their portfolio. The quality and thoroughness of the client assessment determines the appropriateness of all subsequent decisions, making this phase the most consequential in the entire process despite its sometimes underappreciated importance relative to the more analytically intensive tasks that follow.

The strategic asset allocation phase translates the client's investment objectives and risk tolerance into a target allocation across broad asset classes including equities, fixed income, real assets, and alternative investments.

As described in the Asset Allocation article in Section A, the strategic asset allocation is the most important determinant of the portfolio's long-run risk and return characteristics, accounting for the majority of the variation in long-term investment outcomes across different portfolios.

The strategic allocation is determined through a combination of capital market assumptions about the expected returns and risks of different asset classes, optimisation techniques that seek the most efficient combination of expected return and risk, and qualitative judgment about the appropriateness of different asset classes for the specific client's circumstances and objectives.

The security selection phase involves choosing the specific investments within each asset class allocation to implement the strategic asset allocation most effectively.

For passive portfolios, security selection involves the choice of index funds and ETFs that most efficiently track the target benchmark at minimum cost.

For active portfolios, security selection involves the identification of specific securities or managers that the portfolio manager believes are most attractively priced or most skillfully managed given the investment thesis and analytical framework of the strategy.

The security selection process is where the portfolio manager's analytical skill, market knowledge, and investment judgment are most directly expressed, and where the active manager's potential to add value above a passive benchmark is most directly tested.

The implementation phase converts the investment decisions made in the allocation and selection phases into actual portfolio holdings through the execution of trades in the financial markets.

Effective implementation requires minimising the transaction costs and market impact of trades, timing large transactions to avoid adverse price movements, and coordinating the purchase of new positions with the sale of existing positions to manage the portfolio's cash position efficiently. For large institutional portfolios, the implementation phase may involve significant complexity in managing the trading of large positions across multiple brokers and trading venues to achieve best execution.

The monitoring and rebalancing phase maintains the portfolio's alignment with its strategic objectives through ongoing assessment of portfolio performance, risk characteristics, and alignment with the agreed investment policy, and through periodic rebalancing that restores the portfolio to its target allocation when market movements cause drift. Rebalancing enforces the discipline of buying asset classes that have underperformed and trimming those that have outperformed, maintaining the risk profile of the portfolio and preventing the portfolio's asset allocation from drifting away from the client's agreed risk tolerance as different asset classes appreciate and depreciate at different rates.

The performance evaluation phase assesses whether the portfolio is achieving its objectives through the analysis of absolute and relative returns, risk-adjusted performance metrics, attribution of returns to their sources, and comparison against relevant benchmarks. Performance evaluation serves both an accountability function, assessing whether the manager has delivered on their investment commitments, and a learning function, identifying the sources of outperformance or underperformance that can inform future investment decisions.

Asset Allocation and Portfolio Construction

Asset allocation is the most consequential portfolio management decision, with research consistently demonstrating that the choice of how to allocate capital across broad asset classes explains the vast majority of the variation in long-term portfolio returns. The strategic asset allocation framework, as described in detail in the Asset Allocation article, provides the top-level structure within which all other portfolio decisions are made.

Modern portfolio theory, developed by Harry Markowitz in his landmark 1952 paper on portfolio selection, provides the foundational mathematical framework for understanding how combining assets with different return characteristics and correlations affects portfolio risk and expected return. The efficient frontier, which represents the set of portfolios that maximise expected return for any given level of risk, provides a theoretical benchmark for portfolio optimisation that has informed decades of academic research and practical portfolio construction.

The practical implementation of mean-variance optimisation faces several well-documented challenges that limit its direct applicability as a portfolio construction tool. The sensitivity of optimisation outputs to small changes in input assumptions, particularly expected returns that are notoriously difficult to estimate reliably, can produce extreme and unstable portfolio allocations that change dramatically with modest changes in assumptions. The historical correlations between asset classes that serve as inputs to the optimisation may not reliably predict future correlations, particularly during market stress when correlations tend to rise toward one in ways that reduce the diversification benefits the optimisation assumed. And the quadratic utility function that underlies mean-variance optimisation may not accurately represent the preferences of real investors who exhibit loss aversion and other behavioural characteristics that deviate from the assumptions of expected utility theory.

Practical portfolio construction approaches address these limitations through a combination of constraints that limit extreme allocations, shrinkage techniques that moderate the sensitivity of optimisation to input assumptions, scenario analysis that tests portfolio behaviour across a range of possible future environments, and qualitative judgment that supplements the quantitative framework with considerations that formal optimisation cannot capture.

Risk Management in Portfolio Management

Risk management is an integral and continuous dimension of the portfolio management process, encompassing the identification, measurement, monitoring, and control of all material risks that could adversely affect the portfolio's ability to achieve its investment objectives.

Market risk is the most fundamental and most pervasive portfolio risk, reflecting the sensitivity of portfolio value to changes in market prices including equity prices, interest rates, credit spreads, and currency exchange rates. Market risk is measured through metrics including portfolio beta for equity market sensitivity, duration for interest rate sensitivity, and DV01 for dollar interest rate exposure as described in the Duration article in Section D. Managing market risk involves calibrating the portfolio's overall market exposure to the client's risk tolerance and adjusting specific risk exposures within the portfolio to reflect the manager's current market views and risk assessment.

Credit risk in fixed income portfolios reflects the possibility that bond issuers will fail to make scheduled payments of principal or interest, generating losses that reduce portfolio value. Credit risk management involves diversification across issuers and industries, concentration limits that prevent excessive exposure to any single issuer, credit quality minimums that exclude securities below specified rating thresholds, and ongoing credit monitoring that identifies deteriorating credits before they default.

Liquidity risk reflects the possibility that the portfolio will need to sell positions at unfavourable prices due to the inability to find buyers at normal market prices in the time available. Liquidity risk management involves maintaining a minimum allocation to liquid instruments, limiting the proportion of the portfolio held in illiquid investments relative to the client's liquidity needs, and stress-testing the portfolio's liquidity position under adverse scenarios to assess whether sufficient liquid assets are available to meet potential redemption demands.

Concentration risk reflects the danger of excessive exposure to a single security, sector, geography, or factor that could result in large losses if that specific exposure performs adversely. Concentration limits that cap the maximum holding in any single issuer, sector, or other risk factor are the primary tool for managing concentration risk, ensuring that no single adverse outcome can cause disproportionate damage to the overall portfolio.

Active Portfolio Management

Active portfolio management attempts to generate returns above the benchmark through the application of investment research, analysis, and manager judgment to identify securities that are mispriced, markets that are temporarily inefficient, or allocation opportunities that offer superior risk-adjusted returns relative to what a passive index approach would provide.

The sources of active return, collectively called alpha, can arise from multiple dimensions of the portfolio management process. Security selection alpha arises from identifying individual securities that are more attractive than their benchmark weight would suggest, purchasing undervalued securities before the market recognises their true value and selling overvalued securities before the market corrects their inflated prices. Asset allocation alpha arises from overweighting asset classes, sectors, or factors that subsequently outperform and underweighting those that underperform, reflecting the portfolio manager's skill in anticipating market and economic developments. Factor timing alpha arises from dynamically adjusting exposure to known systematic return factors such as value, momentum, and quality based on assessments of when each factor's premium is most likely to be realised.

The active share of a portfolio measures the degree to which the portfolio's holdings differ from its benchmark, calculated as half the sum of the absolute differences between each security's weight in the portfolio and its weight in the benchmark. A portfolio with a one hundred percent active share holds no securities in common with its benchmark, while a portfolio with zero active share is identical to its benchmark. Research by Cremers and Petajisto found that high active share funds showed greater persistence of outperformance than low active share funds, suggesting that truly active portfolios that deviate substantially from their benchmark have a better chance of generating genuine alpha than closet index funds that charge active management fees while maintaining portfolio compositions very similar to their benchmarks.

Institutional vs Individual Portfolio Management

Portfolio management takes substantially different forms for institutional and individual clients, reflecting the very different scale, governance structures, investment objectives, and time horizons characteristic of each client type.

Institutional portfolio management for clients including pension funds, endowments, foundations, insurance companies, and sovereign wealth funds involves managing very large pools of capital under formal investment policy statements and governance frameworks, with investment committees or boards of trustees responsible for setting policy and overseeing implementation. Institutional portfolios typically have longer time horizons, greater access to illiquid and alternative investment strategies, more sophisticated risk management and reporting requirements, and more complex regulatory and fiduciary frameworks than individual investor portfolios.

Individual portfolio management for retail and high-net-worth clients must integrate investment management with the full range of personal financial planning considerations including tax planning, estate planning, income planning, insurance needs, and the management of concentrated stock positions and other client-specific circumstances. The personalised nature of individual portfolio management requires the portfolio manager to develop a thorough understanding of each client's unique financial situation, life stage, and emotional relationship with money and markets, and to adapt their investment approach and communication style to the specific needs of each individual.

Separately managed accounts provide individual investors with personalised portfolio management in which the specific securities are held directly in the client's own account rather than pooled with other investors in a fund structure. Separately managed accounts allow for full customisation of the portfolio to the client's specific tax situation, investment preferences, and any restrictions on specific securities or sectors, and they provide transparency into the specific holdings that many investors value. The minimum investment requirements for separately managed accounts, typically ranging from one hundred thousand dollars to several million dollars depending on the strategy, limit their availability to investors with sufficient assets to benefit from the personalisation they provide.

Performance Measurement and Evaluation

The rigorous measurement and evaluation of portfolio performance is an essential component of the portfolio management process, providing the feedback that allows managers and clients to assess the effectiveness of the investment strategy and the quality of its implementation.

The time-weighted return is the standard performance measurement methodology for investment managers, calculating the compound growth of one dollar invested in the portfolio over the measurement period by eliminating the distorting effect of the timing of client contributions and withdrawals that are outside the manager's control. Time-weighted return allows fair comparison of the manager's investment skill across different portfolios with different cash flow patterns and is the methodology required by the Global Investment Performance Standards maintained by the CFA Institute for the presentation of investment manager track records.

The money-weighted return, also called the internal rate of return, measures the actual return experienced by the investor in a specific account, reflecting both the manager's investment skill and the timing of the investor's own contributions and withdrawals. The money-weighted return is the most relevant measure of the investor's actual financial experience but is not appropriate for evaluating manager skill because it can be significantly affected by the timing of client-initiated cash flows that the manager cannot control.

Benchmark selection is one of the most important and most frequently contentious aspects of performance evaluation, as the choice of benchmark determines whether the manager's returns appear strong or weak relative to a relevant comparison point. An appropriate benchmark must be investable, meaning it represents a realistic alternative that the manager could have held instead of the active portfolio, specified in advance rather than selected after the fact to flatter the manager's returns, and reflective of the manager's investment universe and strategy. The use of inappropriate benchmarks, whether intentionally or through carelessness, is one of the most common sources of misleading performance presentations.

Risk-adjusted performance metrics supplement raw return comparisons by expressing the manager's returns relative to the risk taken to achieve them. The Sharpe ratio, which divides excess return above the risk-free rate by the standard deviation of returns, measures the return per unit of total risk. The information ratio, which divides the active return above the benchmark by the tracking error or standard deviation of active returns, measures the consistency and efficiency of active return generation. The Treynor ratio, which divides excess return by portfolio beta, measures the return per unit of systematic market risk.

Attributing a Result to Allocation and Selection

Attribution answers a narrower question than whether the portfolio beat its benchmark, decomposing the difference between the two returns into the part produced by holding different weights from the benchmark and the part produced by the returns earned inside each category held. The allocation contribution of a category is the difference between the portfolio's weight and the benchmark's weight in that category, multiplied by the amount by which the benchmark's return in that category differed from the benchmark's return as a whole. The selection contribution of a category is the benchmark's weight in that category, multiplied by the difference between the return the portfolio earned inside the category and the return the benchmark earned inside it. The interaction contribution is the product of those same two differences, the weight difference and the return difference, and the three contributions summed across every category equal the portfolio's excess return exactly, which is what makes the decomposition a check on itself rather than an approximation.

Take a benchmark of 60 percent equities returning 10 percent and 40 percent fixed income returning 2 percent, so the benchmark return is 0.60 x 10 + 0.40 x 2 = 6.8 percent. A portfolio held 70 percent equities, which returned 11 percent in its hands, and 30 percent fixed income, which returned 1 percent, so the portfolio return is 0.70 x 11 + 0.30 x 1 = 8.0 percent and the excess return is 1.2 percentage points. The allocation contribution from equities is (0.70 - 0.60) x (10 - 6.8) = 0.32 points and from fixed income (0.30 - 0.40) x (2 - 6.8) = 0.48 points, giving 0.80 points from the weighting decision alone, the fixed income figure being positive because the portfolio underweighted the category that returned less than the benchmark as a whole. The selection contribution from equities is 0.60 x (11 - 10) = 0.60 points and from fixed income 0.40 x (1 - 2) = -0.40 points, giving 0.20 points. The interaction contribution is 0.10 x 1 = 0.10 points from equities and (-0.10) x (-1) = 0.10 points from fixed income, giving 0.20 points, and the three total 0.80 + 0.20 + 0.20 = 1.20 points, which is the excess return.

The decomposition shows what a comparison of the two total returns cannot. Two thirds of this manager's 1.2 point advantage came from how the portfolio was weighted rather than from what was bought inside each category, and within the selection contribution the fixed income sleeve subtracted 0.40 points while the equity sleeve added 0.60, so a positive total excess return is consistent with selection having destroyed value in a category. Reading the sign of the total alone would credit the manager with security selection skill the numbers do not support.

Time-Weighted and Money-Weighted Return on the Same Account

The two methods differ in how they are computed and in what each needs to be computed from. A time-weighted return requires the portfolio to be valued immediately before every contribution and withdrawal, a return to be computed for each sub-period those valuations create, and the sub-period returns to be compounded together into one figure for the whole measurement period. A money-weighted return requires no valuation on any intermediate date: the gain is the ending value less the opening value and less the net contributions, and the return is that gain divided by the capital on average invested over the period, which is computed by weighting each contribution by the fraction of the period it was invested for. So an account whose custodian values it only annually can still produce a money-weighted return, while a time-weighted return cannot be computed at all without a valuation on each cash flow date.

An account opens the year at $100,000 and is worth $110,000 on 30 June, a first sub-period return of 10 percent. The client contributes $100,000 on 1 July, taking the account to $210,000, and the account is worth $189,000 on 31 December, a second sub-period return of -10 percent. The time-weighted return is 1.10 x 0.90 = 0.99, a return of -1.0 percent for the year. The money-weighted return starts from a gain of 189,000 - 100,000 - 100,000, which is -$11,000, against capital on average invested of 100,000 for the whole year plus 100,000 for half of it, which is $150,000, giving -11,000 / 150,000 = -7.3 percent. The manager's decisions were identical in both halves of the year, and the 6.3 point gap between the two figures is produced entirely by the date the client chose to contribute, which left $210,000 exposed to the losing half-year and only $100,000 exposed to the gaining one.

Because the measure is half the sum of the absolute differences between each security's portfolio weight and its benchmark weight, a portfolio can hold every security the benchmark holds and still register a substantial figure. Against a benchmark of four securities at 25 percent each, a portfolio holding them at 40, 35, 25 and 0 percent has absolute differences of 15, 10, 0 and 25, summing to 50, so its active share is 25 percent. A second portfolio holding all four at 40, 30, 20 and 10 percent has differences of 15, 5, 5 and 15, summing to 40, so its active share is 20 percent even though it owns everything the benchmark owns and owns nothing the benchmark does not. The measure adds up weight differences, not absent names, so a complete overlap of holdings is compatible with a material active share and tells the reader nothing about the figure on its own.

The same construction puts a ceiling on the measure. Portfolio weights sum to 100 percent and benchmark weights sum to 100 percent, so every percentage point of overweight in the portfolio is matched by a percentage point of underweight somewhere else in it, the sum of the absolute differences can therefore never exceed 200 percentage points, and halving that sum means the figure can never exceed 100 percent. The halving in the formula is not a convention adopted for presentation; it is what stops each deviation being counted twice, once as the overweight and once as the matching underweight.

What the figure cannot report is how the deviation is distributed. A portfolio that is 25 percentage points overweight in a single security, with 25 points of matching underweight spread across the rest of the benchmark, has absolute differences summing to 50 and an active share of 25 percent. A portfolio that is 2.5 points overweight in each of ten securities, again with 25 points of matching underweight spread across the rest, also has differences summing to 50 and an active share of 25 percent. One portfolio's benchmark-relative outcome turns on one company and the other's on ten, and the identical figures do not separate them, so an active share reading has to be accompanied by the holdings it was computed from before it supports any statement about risk.

The Cost of Implementation

The cost of putting a decision into the portfolio is the sum of several charges that fall at different points, and a decision is only worth implementing if its expected advantage exceeds that sum. Take a position bought and later sold where commission is 10 basis points on each side, the bid-ask spread crossed over the round trip is 30 basis points, because the position is bought at the offer and sold at the bid, and the market impact of each order is 20 basis points: the round trip costs 20 + 30 + 40 = 90 basis points. A decision expected to add 60 basis points over the benchmark before costs therefore produces a net loss of 30 basis points once implemented, and the expected advantage of a decision has to be compared against the full round trip rather than against the commission alone.

Because the cost falls on each round trip, the annual cost of implementation is the cost per round trip multiplied by the number of times the portfolio is replaced in the year. At 90 basis points a round trip, a portfolio turned over once a year carries 90 basis points of implementation cost and one turned over twice carries 180, so two managers reaching the same result before costs do not reach the same result after them, and the one trading twice as often has to be twice as right to finish level.

Timing a large transaction pulls in two directions at once. Working the order in pieces over several days reduces the market impact of each piece, because each is smaller relative to the volume available to absorb it, but it leaves the unexecuted remainder exposed to the price moving away before the order is complete. The two costs therefore run in opposite directions as the schedule is stretched, impact falling and exposure to adverse price movement rising, so the cheapest schedule is neither the fastest nor the slowest but the one that balances them, and a trading decision that minimises only one of the two has not minimised the cost of the trade.

A Concentrated Holding in an Individual Portfolio

The weight of a holding sets the limit of the damage one company's result can do to a portfolio. In a $1,000,000 portfolio, a holding of $400,000 is 40 percent, and a 25 percent fall in that holding costs $100,000, which is 10 percent of the portfolio. A holding of $50,000 in the same portfolio is 5 percent, and even if it becomes entirely worthless it costs $50,000, which is 5 percent of the portfolio, so it cannot cost the portfolio 10 percent of its value under any outcome at all. The company-specific risk the portfolio actually bears is therefore governed by the weight of the position and not only by how volatile the company's shares are.

Repairing a concentrated loss takes a larger return than the loss itself. If the $400,000 holding halves, the portfolio loses $200,000 and stands at $800,000, of which $200,000 is the fallen holding and $600,000 is everything else. If the fallen holding does not recover, restoring the portfolio to $1,000,000 requires the other holdings to produce the whole $200,000 on a base of $600,000, a gain of about 33 percent. The arithmetic is not symmetrical because the loss is taken on the whole position while the recovery has to be earned on the smaller base that is left.

Where the concentrated holding is stock in the client's own employer, the holding and the client's income are exposed to the same events. The business reverse that takes the share price down can also reduce or end the client's salary and bonus, so the capital loss and the income loss arrive together rather than offsetting one another, and the client's total exposure to that one company is larger than the position's portfolio weight records. A portfolio weight computed on the investment account alone therefore understates the concentration for this client in a way it would not for a client holding the same stock in an unrelated company.

A Long-Horizon Portfolio with a Spending Requirement

Where a portfolio exists to fund a continuing distribution, the return it must earn is an addition and nothing more. A foundation that distributes 5 percent of its portfolio each year, pays 0.5 percent a year in investment costs and intends the portfolio's purchasing power to be unchanged at the end of the year must earn the distribution, plus the costs, plus the rate by which prices rose; on an assumed inflation rate of 2.5 percent for the year, the required nominal return is 5 + 0.5 + 2.5 = 8 percent. Because the three terms are simply added, each one bounds the others: on the same assumptions, a distribution set at 4 percent rather than 5 percent lowers the required nominal return to 7 percent, and no change in the allocation is needed to produce that effect.

How the distribution is computed decides how violently it moves with the market. A distribution set at 5 percent of the latest portfolio value falls from $5,000,000 to $4,000,000 when the portfolio falls from $100,000,000 to $80,000,000, a 20 percent cut in the year of the fall. A distribution set instead at 5 percent of the average of the last three year-end values, where those values were $100,000,000, $100,000,000 and $80,000,000, is computed on an average of $93,333,333 and so comes to $4,666,667, about 7 percent below the previous year's $5,000,000. The fall enters only one of the three years averaged, so the spending change is roughly a third of the market change, and the smoothing is produced by the denominator of the distribution rule rather than by any investment decision.

A distribution rate above what the portfolio earns after inflation and costs compounds into a loss of real value that no single year reveals. A portfolio distributing 6 percent a year while earning 4 percent a year after inflation and costs loses 2 percent of its real value each year, which is barely visible in any one year's accounts. Over thirty-five years it is not small: 0.98 raised to the thirty-fifth power is about 0.49, so the portfolio's purchasing power halves, and the distribution, being a percentage of a portfolio worth half as much in real terms, buys half as much as it did at the start.

Examination Relevance and Key Takeaways

Portfolio management is among the most comprehensively tested topics on the Series 65 examination, encompassing the full range of investment management concepts from client assessment through performance evaluation. Candidates must understand the portfolio management process including client assessment, strategic asset allocation, security selection, implementation, monitoring and rebalancing, and performance evaluation, the foundations of modern portfolio theory including the efficient frontier and the role of diversification in reducing portfolio risk, the risk management dimensions of portfolio management including market risk, credit risk, liquidity risk, and concentration risk, the distinction between active and passive portfolio management and the evidence on active management performance, and the performance measurement methodologies including time-weighted return and risk-adjusted metrics.

The core points to retain are these: portfolio management is the professional discipline of constructing and managing investments to achieve specified objectives within defined risk parameters through a structured process encompassing client assessment, strategic asset allocation, security selection, implementation, monitoring, rebalancing, and performance evaluation; the strategic asset allocation is the most important portfolio decision accounting for the majority of long-run return variation; modern portfolio theory demonstrates that combining imperfectly correlated assets reduces portfolio risk below the weighted average of individual asset risks; the efficient frontier represents portfolios that maximise expected return for any given risk level.

Active management seeks to generate alpha through security selection, asset allocation, and factor timing while passive management tracks benchmarks at minimal cost; active share measures the degree to which a portfolio differs from its benchmark with high active share associated with more genuine active management; time-weighted return is the standard manager performance measurement methodology that eliminates the distorting effect of client cash flow timing; and risk-adjusted metrics including the Sharpe ratio, information ratio, and Treynor ratio supplement raw return comparisons by expressing performance relative to the risk taken to achieve it.