Use this guide to connect investment performance calculations with the decisions they support. Each concept includes an original worked example and a specific error to avoid. Foundations come first, followed by attribution, appraisal, presentation, and manager selection. Work through the calculations, then explain what each result does—and does not—tell an investor.
Ethics and professional judgment
1. Correcting material performance errors
Ethical reporting requires accurate results and a clear response when material errors are discovered. Investigate the cause, establish the affected reports and recipients, and follow the applicable correction policy. A corrected number should remain traceable to its calculation and explanation; quietly replacing a file can leave clients relying on misleading information.
Worked example: A duplicated dividend changes a reported return from 6.4% to 5.9%. The analyst corrects the calculation, documents the duplication, and arranges communication to affected recipients.
Mistake to avoid: Correcting the spreadsheet while leaving the previously distributed report unaddressed.
Reference: CIPM® Program Curriculum | CFA Institute
2. Protecting independent judgment
Benefits from managers or vendors can create pressure to soften analysis. Assess whether a gift, hospitality arrangement, or commercial relationship could compromise independence or appear to do so. Apply the relevant professional and employer rules, disclose the circumstances appropriately, and preserve analytical decisions that can stand without the benefit.
Worked example: A manager offers expensive travel before a review. The analyst declines the offer and evaluates the manager using the same documented criteria as competing firms.
Mistake to avoid: Assuming a benefit is harmless merely because no favorable rating was explicitly requested.
Reference: CIPM® Program Curriculum | CFA Institute
3. Identifying conflicts of interest
A conflict exists when another interest could influence duties owed to a client or employer. Identify financial incentives, personal relationships, and organizational pressures before evaluating results. Disclosure supports informed judgment, but some conflicts also require reassignment or other controls. Explain the relevant interest clearly enough for its potential effect to be understood.
Worked example: An analyst evaluating a manager owns shares in its parent company. The interest is disclosed, and an independent colleague leads the recommendation under the firm’s policy.
Mistake to avoid: Treating disclosure alone as a remedy for every conflict.
Reference: CIPM® Program Curriculum | CFA Institute
4. Maintaining client confidentiality
Performance data can reveal holdings, trading decisions, client identities, and financial circumstances. Share confidential information only with an appropriate authorization or other applicable basis, and restrict access to what the recipient needs. Removing a client’s name may be insufficient when distinctive holdings or account details still identify the client.
Worked example: A training example uses invented holdings and balances rather than an actual pension client’s unusual portfolio, preventing identification through its distinctive positions.
Mistake to avoid: Assuming an unnamed account is anonymous when its holdings make its identity obvious.
Reference: CIPM® Program Curriculum | CFA Institute
5. Dealing fairly when distributing analysis
Fair dealing means avoiding selective treatment that improperly advantages some clients. For performance communications, establish a consistent distribution process and explain legitimate differences in service. Fairness does not require identical portfolios or services, but material information should not be withheld to favor a preferred recipient without a justified basis.
Worked example: A correction affects two clients receiving the same strategy report. Both receive the corrected analysis through the established process, rather than notifying only the larger client.
Mistake to avoid: Confusing fair treatment with giving every client the same investment recommendation.
Reference: CIPM® Program Curriculum | CFA Institute
6. Establishing a reasonable analytical basis
Conclusions about performance need adequate investigation and support. Check data quality, benchmark suitability, methodology, and relevant limitations before recommending action. Distinguish verified observations from estimates and judgments. A vendor calculation can be useful evidence, but its label or reputation does not eliminate the need to understand material assumptions.
Worked example: A dashboard flags exceptional outperformance. Reviewing the benchmark reveals a currency mismatch, so the analyst withholds a skill conclusion until comparable returns are calculated.
Mistake to avoid: Accepting a favorable ranking without examining how the comparison was constructed.
Reference: CIPM® Program Curriculum | CFA Institute
7. Distinguishing actual and hypothetical results
Actual portfolios, model portfolios, and backtests provide different evidence. Identify which produced a result and explain consequential assumptions, including trading costs and how investment decisions were selected. Hypothetical results can illustrate an approach, but they should not be described as an investor’s realized experience or proof that the approach will succeed.
Worked example: A simulated strategy earns 12% before estimated costs. The report labels it hypothetical and shows that a 1% assumed cost reduces the illustrated result to approximately 11%.
Mistake to avoid: Presenting a backtest as a live track record.
Reference: CIPM® Program Curriculum | CFA Institute
8. Preserving reproducible analytical records
A performance conclusion should be reproducible from retained inputs, methods, and decisions. Preserve relevant valuations, cash-flow records, benchmark data, calculation settings, and explanations for adjustments under applicable policies. The record should show what was known when the analysis was performed, rather than relying on a later database that may have changed.
Worked example: An analyst saves the benchmark constituent file used in an attribution report. A later index revision therefore does not prevent reconstruction of the original calculation.
Mistake to avoid: Keeping only the final chart and discarding the inputs needed to explain it.
Reference: CIPM® Program Curriculum | CFA Institute
Performance measurement foundations
9. Holding-period return
Holding-period return measures the change in invested value over a defined interval. With no external cash flows or separate income, divide ending value by beginning value and subtract one. State the interval explicitly: a quarterly return and an annual return cannot be compared directly simply because both are percentages.
Worked example: An investment rises from 250 to 270 over one quarter. Its quarterly return is 270 / 250 − 1 = 8%; this is not automatically an 8% annual return.
Mistake to avoid: Dividing the gain by ending value instead of beginning value.
Reference: CIPM® Program Curriculum | CFA Institute
10. Total return includes investment income
Total return includes both price changes and investment income. For a simple holding period without reinvestment or external flows, add income received to ending value before comparing with beginning value. Price return alone can misrepresent income-producing investments and create unfair comparisons between assets with different distribution policies.
Worked example: A share bought for 50 ends at 53 and pays a dividend of 2. Total return is (53 − 50 + 2) / 50 = 10%; price return is 6%.
Mistake to avoid: Omitting distributions because they no longer appear in the security’s market price.
Reference: CIPM® Program Curriculum | CFA Institute
11. External cash flows versus investment proceeds
An external cash flow crosses the boundary of the portfolio being measured. A client contribution or withdrawal is external; a dividend retained inside the portfolio is investment income. Purchases and sales within that boundary generally change holdings without adding investor capital. Correct classification prevents deposits from appearing as investment gains.
Worked example: A portfolio rises from 100 to 130 after a client deposits 20 at period end. The investment gain is 10, giving a 10% return under this timing assumption.
Mistake to avoid: Treating a security sale retained as portfolio cash as a client withdrawal.
Reference: CIPM® Program Curriculum | CFA Institute
12. Time-weighted return
Time-weighted return separates investment performance from the size and timing of external cash flows. When suitable valuations are available, calculate returns between flows and compound the subperiod results. It is particularly useful for evaluating a manager who does not control client contributions or withdrawals. Accurate flow timing and boundary valuations remain essential.
Worked example: A portfolio grows from 100 to 110, receives 50, then grows from 160 to 176. Both subperiod returns are 10%, so linked return is 1.10 × 1.10 − 1 = 21%.
Mistake to avoid: Including the contribution as part of the first subperiod’s investment gain.
Reference: CIPM® Program Curriculum | CFA Institute
13. Money-weighted return
Money-weighted return reflects the amounts and timing of invested capital, commonly through an internal rate of return. Solve for the discount rate that equates contributions with withdrawals and terminal value. Use a consistent timing convention. Unusual cash-flow patterns can produce multiple solutions or no useful solution, so interpretation requires more than finding a numerical root.
Worked example: Invest 100 initially and 50 one year later; receive 176 after two years. A 10% annual rate satisfies 100 × 1.10² + 50 × 1.10 = 176.
Mistake to avoid: Interpreting money-weighted return as independent of investor cash-flow decisions.
Reference: CIPM® Program Curriculum | CFA Institute
14. Modified Dietz estimation
Modified Dietz estimates return by dividing investment gain by beginning value plus time-weighted external flows. Weight each flow by the fraction of the period it remains invested, using the stated timing convention. The method can be less accurate when large flows coincide with volatile returns, so its suitability depends on the circumstances.
Worked example: Beginning value is 100, ending value is 160, and a contribution of 40 occurs halfway through. Return is (160 − 100 − 40) / (100 + 0.5 × 40) = 16.67%.
Mistake to avoid: Giving a period-end contribution the same weight as a beginning-period contribution.
Reference: CIPM® Program Curriculum | CFA Institute
15. Geometric linking of returns
Returns across consecutive periods compound because each period acts on the wealth remaining from the previous period. Multiply the growth factors, then subtract one. Adding percentages ignores this changing base. A gain and an equal-sized percentage loss do not cancel because the loss applies to a different amount of capital.
Worked example: A portfolio gains 20% and then loses 10%. Its cumulative return is 1.20 × 0.90 − 1 = 8%, rather than the 10% obtained by addition.
Mistake to avoid: Adding consecutive returns to calculate the change in investor wealth.
Reference: CIPM® Program Curriculum | CFA Institute
16. Annualized versus cumulative return
Cumulative return describes total growth over the observation period. Annualized return expresses the equivalent compound yearly rate: (1 + cumulative return) raised to 1 divided by years, minus one. Annualization changes the unit of expression rather than the investment outcome. Short-period extrapolations require care and are not forecasts.
Worked example: A portfolio gains 21% over two years. Its annualized return is 1.21^(1/2) − 1 = 10%, while its cumulative return remains 21%.
Mistake to avoid: Reporting cumulative and annualized returns as though they were interchangeable.
Reference: CIPM® Program Curriculum | CFA Institute
17. Arithmetic and geometric average returns
The arithmetic average summarizes the average single-period observation. The geometric average describes compounded growth per period. Variability causes the geometric average to fall below the arithmetic average for ordinary positive wealth factors. Choose the average according to the question, rather than assuming an arithmetic mean represents the investor’s compound experience.
Worked example: Returns of 25% and −20% have an arithmetic average of 2.5%. Since 1.25 × 0.80 = 1, their geometric average is 0%.
Mistake to avoid: Using the arithmetic average as the realized compound growth rate.
Reference: CIPM® Program Curriculum | CFA Institute
18. Gross and net performance
Gross and net return labels need an explicit definition of which expenses are included or deducted. Management fees, transaction costs, and other expenses are different categories. A valid comparison aligns the treatment of these items and the calculation period. For actual portfolios, fee timing can also affect the relationship between reported gross and net results.
Worked example: Under a stated simplified illustration, a 7% gross return less a fee equal to 1% of beginning assets gives 6% net. A differently timed fee calculation need not match exactly.
Mistake to avoid: Assuming a gross return excludes every type of investment cost.
Reference: CIPM® Program Curriculum | CFA Institute
19. Consistent valuation and income accrual
Returns depend on comparable beginning and ending valuations. Recognize earned income consistently and avoid counting the same amount both in asset value and as a separate receipt. For illiquid holdings, valuation uncertainty can affect measured returns even when calculations are mechanically correct. Distinguish a valuation estimate from an observable transaction price.
Worked example: A bond’s reported value already includes 3 of accrued interest. Adding that same 3 again as income would overstate portfolio wealth; the analyst removes the duplicate.
Mistake to avoid: Mixing values that include accrued income with values that exclude it.
Reference: CIPM® Program Curriculum | CFA Institute
20. Returns across currencies
A foreign investment’s base-currency return combines its local return with the change in the base-currency value of the foreign currency. For an unhedged investment, multiply their growth factors. Define the exchange-rate quotation explicitly, because reversing the quotation changes the calculation. Hedging introduces separate gains, losses, and costs.
Worked example: An asset gains 4% locally while its currency appreciates 5% against the investor’s base currency. Base-currency return is 1.04 × 1.05 − 1 = 9.2%.
Mistake to avoid: Adding local and currency returns while ignoring their interaction.
Reference: CIPM® Program Curriculum | CFA Institute
21. Asset-weighted composite returns
A composite combines portfolios representing an investment strategy. Under a simple beginning-asset-weighted calculation with no external flows, each portfolio’s weight equals its beginning assets divided by total beginning assets. This measures the experience of invested capital rather than the average account. More complex flow patterns require an appropriate consistent methodology.
Worked example: Accounts begin with 80 and 20 and earn 5% and 15%. Their asset-weighted return is 0.80 × 5% + 0.20 × 15% = 7%, not 10%.
Mistake to avoid: Using an equal account average when the question specifies asset weighting.
Reference: CIPM® Program Curriculum | CFA Institute
22. Reconciling investment gains
A valuation bridge separates investment gain from external flows before a return is interpreted. Check that beginning assets plus net contributions plus investment gain equal ending assets. This identity detects omissions and sign errors, although it does not by itself determine the appropriate return method because cash-flow timing still matters.
Worked example: Beginning assets are 500, contributions are 60, withdrawals are 20, and ending assets are 565. Investment gain is 565 − 500 − 60 + 20 = 25.
Mistake to avoid: Subtracting a withdrawal twice or treating the reconciled gain as a return percentage.
Reference: CIPM® Program Curriculum | CFA Institute
Explaining investment results
23. Return contribution and active contribution
A holding’s return contribution combines its portfolio weight with its return under the chosen attribution convention. Active contribution compares portfolio and benchmark contributions. A strongly performing holding can contribute little if its weight is small, while a positive contribution can still represent underperformance relative to the benchmark exposure.
Worked example: A sector weighted 20% returns 10%, contributing 2 percentage points. The benchmark weights it at 30% with the same return, so active contribution is −1 percentage point.
Mistake to avoid: Calling a positive portfolio contribution evidence of benchmark outperformance.
Reference: CIPM® Program Curriculum | CFA Institute
24. Allocation effects
Allocation attribution evaluates whether different sector weights helped relative performance. In a Brinson–Fachler formulation, a sector’s allocation effect is the portfolio-minus-benchmark weight multiplied by the sector benchmark return minus the overall benchmark return. An overweight helps when that sector outperforms the overall benchmark, even if its absolute return is negative.
Worked example: A sector has a 10-percentage-point overweight, a 9% benchmark return, and an overall benchmark return of 5%. Its allocation effect is 0.10 × 0.04 = 0.40 percentage points.
Mistake to avoid: Multiplying the overweight by the sector return without checking the attribution model.
Reference: CIPM® Program Curriculum | CFA Institute
25. Selection effects
Selection attribution compares the portfolio’s return within a segment with the corresponding benchmark segment return. In a three-effect Brinson formulation, selection uses the benchmark segment weight. This asks whether the selected securities performed better within the segment, while separating the effect of choosing a different segment weight.
Worked example: A sector’s benchmark weight is 30%; portfolio securities return 8% against a sector benchmark return of 6%. Selection effect is 0.30 × 0.02 = 0.60 percentage points.
Mistake to avoid: Using portfolio weights in a formula that separately reports interaction.
Reference: CIPM® Program Curriculum | CFA Institute
26. Interaction between allocation and selection
In a three-effect Brinson model, interaction captures the joint effect of active weights and within-segment excess returns. Multiply the active weight by the segment return difference. Interaction can be positive when a manager underweights a segment in which its security selection underperforms. Some methodologies combine interaction with selection, so definitions matter.
Worked example: A sector is overweight by 10 percentage points and its selected securities outperform the sector benchmark by 3 percentage points. Interaction contributes 0.10 × 0.03 = 0.30 percentage points.
Mistake to avoid: Adding interaction separately when the reported selection effect already includes it.
Reference: CIPM® Program Curriculum | CFA Institute
27. Attribution reconciliation and residuals
Attribution effects should reconcile to the return difference defined by the model. An unexplained residual may signal missing cash, fees, inconsistent classifications, valuation timing, or approximation. Investigate its source before presenting the explanation as complete. A small residual can arise from rounding, but size alone does not establish its cause.
Worked example: A portfolio returns 8.1% against 7%. Effects of 0.5, 0.4, and 0.2 percentage points sum to the 1.1-point excess return, leaving no residual at the reported precision.
Mistake to avoid: Labeling an unexplained residual as manager skill.
Reference: CIPM® Program Curriculum | CFA Institute
28. Arithmetic excess and geometric relative return
Arithmetic excess return subtracts benchmark return from portfolio return. Geometric relative return divides their growth factors and subtracts one. The first expresses a percentage-point difference; the second expresses relative wealth growth. Both can be useful, but attribution effects must match the chosen definition rather than mixing incompatible totals.
Worked example: Portfolio return is 12% and benchmark return is 10%. Arithmetic excess is 2 percentage points; geometric relative return is 1.12 / 1.10 − 1 = approximately 1.82%.
Mistake to avoid: Calling a percentage-point difference a compounded relative wealth return.
Reference: CIPM® Program Curriculum | CFA Institute
29. Linking attribution across periods
Single-period attribution effects usually cannot be added unchanged to explain cumulative performance. Portfolio and benchmark returns compound separately, introducing cross-period effects. Use a defined linking method that reconciles contributions to the selected cumulative excess-return measure. Check both the total and the interpretation of the linked effects.
Worked example: Portfolio returns of 5% and 6% compound to 11.30%; benchmark returns of 4% and 4% compound to 8.16%. Cumulative arithmetic excess is 3.14 points, not 1 + 2 = 3 points.
Mistake to avoid: Summing monthly active effects without addressing compounding.
Reference: CIPM® Program Curriculum | CFA Institute
30. Currency exposure and hedging attribution
Currency attribution separates the consequences of foreign-currency exposure from local investment results and hedging decisions. Define the currency benchmark and treatment of hedge gains, costs, and interaction. A hedge can reduce currency losses without making the entire portfolio profitable; evaluate its contribution against the exposure and policy it was intended to manage.
Worked example: A portfolio starts at 200, including foreign assets of 100. With unchanged local prices, currency depreciation causes a loss of 5; a hedge earns 3. Net currency contribution is −2 / 200 = −1%.
Mistake to avoid: Reporting the hedge gain while omitting the loss on the hedged exposure.
Reference: CIPM® Program Curriculum | CFA Institute
31. Duration-based interest-rate attribution
For a small parallel yield change, a bond portfolio’s approximate price return is negative modified duration multiplied by the yield change. This identifies interest-rate exposure as a return source, but excludes carry, spread changes, and higher-order effects. Compare portfolio and benchmark sensitivities when explaining active performance.
Worked example: Modified duration is 4 and yields rise by 0.25 percentage points. Approximate price return is −4 × 0.0025 = −1%, before income and other effects.
Mistake to avoid: Entering a 0.25-percentage-point yield change as 0.25 instead of 0.0025.
Reference: CIPM® Program Curriculum | CFA Institute
32. Yield-curve and spread effects
Fixed-income returns can reflect changes in different parts of the government yield curve and changes in credit spreads. Overall duration is insufficient when short and long rates move differently. Key-rate sensitivities help identify curve exposure, while spread sensitivity helps isolate credit repricing. State which movement each approximation measures.
Worked example: Government yields are unchanged, but credit spreads widen 0.20 percentage points. With spread duration of 3, the approximate spread-related price effect is −3 × 0.002 = −0.6%.
Mistake to avoid: Attributing a credit-spread loss to a government-rate move that did not occur.
Reference: CIPM® Program Curriculum | CFA Institute
33. Factor-based return attribution
Factor attribution explains returns through exposures to common drivers and their realized returns, with an unexplained component. Active exposures can reveal that apparent security-selection success came from systematic tilts. Factor results depend on model design and estimation; an unexplained return is not automatically proof of skill or a forecasting opportunity.
Worked example: An active factor exposure of 0.4 meets a factor return of 2%, contributing 0.8 percentage points. Against total active return of 1.1 points, 0.3 points remain unexplained by this simplified model.
Mistake to avoid: Calling every model residual alpha earned through repeatable manager skill.
Reference: CIPM® Program Curriculum | CFA Institute
Risk and performance appraisal
34. Choosing an appropriate benchmark
A benchmark should represent the opportunity set and investment approach being evaluated. Examine asset coverage, currency, return treatment, and consistency with the mandate. A benchmark that is easy to outperform but unrelated to the strategy weakens both attribution and appraisal. Benchmark suitability is a substantive judgment, not merely a data-availability decision.
Worked example: A global equity portfolio measured in euros is compared with a euro-denominated global equity benchmark rather than a domestic bond index. The comparison now addresses its intended investment opportunity set.
Mistake to avoid: Selecting the benchmark after observing which one produces the strongest excess return.
Reference: CIPM® Program Curriculum | CFA Institute
35. Volatility measures total return variability
Standard deviation measures dispersion around average return. It treats positive and negative deviations alike and does not directly measure permanent loss or liquidity risk. Use returns with consistent frequency and state whether the estimate is a population or sample statistic. Annualization through square-root-of-time scaling relies on assumptions about return dependence.
Worked example: For population observations of 0% and 4%, mean return is 2%. Population standard deviation is sqrt((2² + 2²) / 2) = 2 percentage points.
Mistake to avoid: Interpreting low measured volatility as evidence that all important risks are low.
Reference: CIPM® Program Curriculum | CFA Institute
36. Tracking error measures active variability
Tracking error is the standard deviation of portfolio-minus-benchmark returns. It measures variability relative to the benchmark, rather than total portfolio volatility. Use aligned observations and distinguish forecast tracking error from realized tracking error. A portfolio can have substantial market risk while staying very close to its benchmark.
Worked example: Active returns are −1%, 0%, and 1%. Their mean is zero, and sample tracking error is sqrt((1 + 0 + 1) / 2) = 1 percentage point per observation period.
Mistake to avoid: Using portfolio return standard deviation when the question asks for tracking error.
Reference: CIPM® Program Curriculum | CFA Institute
37. Sharpe ratio
The Sharpe ratio divides average return above the risk-free rate by total return volatility. Match the periods and units of all inputs. It supports comparisons involving total risk, but its interpretation depends on return characteristics and the reliability of estimated moments. Negative excess returns and non-normal outcomes can complicate simple rankings.
Worked example: Annual average return is 8%, the risk-free rate is 2%, and annual volatility is 12%. Sharpe ratio is (8 − 2) / 12 = 0.50.
Mistake to avoid: Combining an annual excess return with monthly volatility without adjusting the units.
Reference: CIPM® Program Curriculum | CFA Institute
38. Information ratio
The information ratio divides average active return by tracking error. It evaluates return relative to a benchmark per unit of active variability. Benchmark suitability remains essential: an attractive ratio against an inappropriate benchmark says little about mandate execution. Use consistent observation periods and avoid treating a short favorable record as a stable expectation.
Worked example: Average annual active return is 2.4% and annual tracking error is 4%. The information ratio is 2.4 / 4 = 0.60.
Mistake to avoid: Dividing active return by total portfolio volatility and calling the result an information ratio.
Reference: CIPM® Program Curriculum | CFA Institute
39. Beta and systematic exposure
Beta measures a portfolio’s sensitivity to a specified market return in a linear model. It equals covariance with that market divided by market variance. Beta is not total risk: portfolios with the same beta can have different residual volatility. Estimates depend on the benchmark, sample, and stability of the underlying relationship.
Worked example: Covariance with the market is 0.018 and market variance is 0.015, using decimal returns. Beta is 0.018 / 0.015 = 1.2.
Mistake to avoid: Concluding that beta of zero means the portfolio cannot lose money.
Reference: CIPM® Program Curriculum | CFA Institute
40. Risk-adjusted alpha
Alpha compares performance with the return implied by a specified risk model. In a simple market model, subtract the risk-free rate plus beta times the market risk premium from portfolio return. A positive estimate may reflect skill, omitted factors, sampling variation, or model error. It must be interpreted alongside the model and evidence.
Worked example: Portfolio return is 9%, the risk-free rate is 2%, market return is 7%, and beta is 1.2. Model-implied return is 8%, leaving alpha of 1 percentage point.
Mistake to avoid: Treating positive estimated alpha as conclusive evidence of repeatable skill.
Reference: CIPM® Program Curriculum | CFA Institute
41. Downside deviation and the Sortino ratio
Downside measures focus on outcomes below a defined target rather than all deviations from the mean. The Sortino ratio divides average return above that target by downside deviation. Specify the target and calculation convention, since different treatments of observations can produce different denominators. A higher result does not remove other risks.
Worked example: Average annual return is 7%, the target is 3%, and annual downside deviation is 5%. Sortino ratio is (7 − 3) / 5 = 0.80.
Mistake to avoid: Comparing Sortino ratios calculated with different targets as though their definitions match.
Reference: CIPM® Program Curriculum | CFA Institute
42. Drawdown and recovery
Drawdown measures the decline from a previous wealth peak to a subsequent value. Maximum drawdown is the largest such peak-to-trough decline in the observation window. Recovery requires a larger percentage gain than the preceding percentage loss because the gain starts from a smaller base. Cash flows require appropriate adjustment when constructing the wealth series.
Worked example: A wealth index falls from 125 to 100, a 20% drawdown. Returning to 125 requires a 25% gain from 100.
Mistake to avoid: Assuming a 20% gain restores wealth after a 20% loss.
Reference: CIPM® Program Curriculum | CFA Institute
43. Statistical uncertainty in active performance
Observed outperformance is an estimate rather than a direct measurement of enduring skill. Under suitable assumptions, the standard error of mean active return is sample active volatility divided by the square root of the observation count. Serial dependence, changing exposures, and multiple comparisons can undermine simple significance calculations.
Worked example: Mean monthly active return is 0.2%, sample active volatility is 1%, and there are 25 independent observations. Standard error is 1% / sqrt(25) = 0.2%, giving a t-statistic of 1.
Mistake to avoid: Declaring skill from a positive average without assessing estimation uncertainty.
Reference: CIPM® Program Curriculum | CFA Institute
44. Value at risk and tail limitations
Value at risk describes a loss quantile for a specified horizon and confidence level under a model or data method. It does not state the largest possible loss or the average loss beyond that threshold. Examine tail severity, liquidity, and changing market conditions separately; estimated quantiles depend on the assumptions and observations used.
Worked example: A model reports one-day 95% value at risk of 2 million. It estimates that losses exceed 2 million on 5% of days; it does not cap those losses at 2 million.
Mistake to avoid: Reading value at risk as a worst-case loss guarantee.
Reference: CIPM® Program Curriculum | CFA Institute
Performance presentation and GIPS standards
45. Firm-wide GIPS compliance
The Global Investment Performance Standards (GIPS) provide a framework for fair representation and full disclosure of investment performance. Compliance is a firm-wide claim under the applicable standards, rather than a label attached selectively to an attractive account. Accurate calculations alone do not establish compliance; relevant policies, presentation, and disclosure requirements also matter.
Worked example: A firm has carefully calculated one account’s return. It cannot infer firm-wide GIPS compliance from that account alone; it must evaluate the applicable requirements across the defined firm.
Mistake to avoid: Describing an individual portfolio as independently GIPS compliant.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
46. Defining the reporting firm
The firm definition establishes the organizational boundary for a GIPS compliance claim. It should reflect how the investment business is held out and operated under the applicable standards. A clear boundary supports consistent asset totals, strategy coverage, and accountability. Do not redraw the boundary opportunistically to omit an inconvenient part of the performance record.
Worked example: Two investment teams operate under one defined firm. Reporting only the successful team’s assets as total firm assets would misrepresent that stated boundary.
Mistake to avoid: Changing the firm definition simply to exclude poorly performing business units.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
47. Composite definitions and comparability
A composite groups portfolios managed according to a similar mandate, objective, or strategy. Its definition should identify the investment approach and relevant distinctions sufficiently to support meaningful comparisons. Portfolios with different risk constraints may require different treatment even when they share an asset class. Apply the relevant GIPS requirements when determining membership.
Worked example: Unrestricted global equity accounts and accounts prohibited from holding several major sectors have materially different opportunity sets. The firm assesses those constraints before grouping them together.
Mistake to avoid: Combining portfolios solely because both are labeled equity.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
48. Avoiding selective composite membership
Composite policies should prevent performance-driven selection of portfolios. Inclusion and exclusion should follow documented eligibility rules and the applicable standards, preserving relevant history rather than retaining only current successes. Differences in inception, termination, discretion, or mandate need appropriate treatment; they are not invitations to choose whichever accounts improve the reported return.
Worked example: An eligible account later closes after weak performance. The firm preserves its historical contribution for the periods in which it belonged to the composite.
Mistake to avoid: Deleting a closed account’s past results to improve the surviving composite record.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
49. Dispersion versus time-series risk
Internal dispersion describes differences among portfolio returns within a composite for a period. Time-series risk describes variability in a return series across periods. They answer different questions and should be labeled accordingly. Low dispersion can indicate similar account experiences even when all accounts experience substantial market volatility.
Worked example: Three accounts earn 4%, 5%, and 6% in one year. Their population dispersion is approximately 0.82 percentage points; this does not measure the composite’s volatility over multiple years.
Mistake to avoid: Using cross-account dispersion as a substitute for historical composite volatility.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
50. Disclosures that make returns interpretable
A return needs context to be meaningful. Explain the strategy, benchmark, reporting currency, fee treatment, material methodology, and relevant limitations as appropriate to the presentation and applicable standards. Disclosures should enable interpretation rather than obscure important qualifications. A polished chart does not compensate for missing information about how its numbers were produced.
Worked example: Two charts both show 8%. One is gross of management fees in dollars; the other is net in euros. Clear labels prevent readers from treating them as equivalent outcomes.
Mistake to avoid: Placing a material fee or currency distinction where readers are unlikely to notice it.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
51. Understanding verification and examinations
GIPS verification evaluates specified firm-wide matters through an independent process. It should not be interpreted as a guarantee of investment results or assurance that every individual performance figure is correct. A performance examination addresses a different, more specific scope. Read the relevant report and applicable standards before describing what assurance was provided.
Worked example: A prospect asks whether verification guarantees a composite’s reported 9% return. The firm explains the verification scope and distinguishes it from any separate performance examination.
Mistake to avoid: Advertising verification as a guarantee of returns or investment quality.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
52. Matching presentation periods and return bases
Portfolio and benchmark comparisons should use matching dates and clearly identified return bases. Distinguish calendar-period returns, cumulative growth, and annualized results. Include income consistently and explain material gaps or changes. A comparison can mislead even when each individual number is correct if the periods or definitions differ.
Worked example: A portfolio’s 21% cumulative two-year return is compared with a benchmark’s 9% annualized return. Annualizing the portfolio to 10% creates a valid like-for-like rate comparison.
Mistake to avoid: Comparing a partial-year portfolio result with a full-year benchmark result.
Reference: CIPM® Program Curriculum | CFA Institute; CIPM® Program | Certificate in Investment Performance Measurement
Manager selection and monitoring
53. Translating client objectives into a mandate
Manager selection begins with the role the investment must serve. Define return objectives, risk tolerance, liquidity needs, horizon, and relevant constraints before comparing managers. A manager can be strong in its own strategy yet unsuitable for the client. Evaluate fit against the mandate rather than ranking managers solely by historical return.
Worked example: An investor needs predictable access to capital for near-term payments. A manager requiring a long capital lock-up is excluded despite attractive past returns.
Mistake to avoid: Selecting the highest-return manager before checking whether its strategy fits the client.
Reference: CIPM® Program Curriculum | CFA Institute
54. Combining quantitative and qualitative evidence
Quantitative analysis describes the track record, while qualitative investigation examines how it was produced and whether the organization can sustain its approach. Reconcile the two rather than treating either as sufficient. Useful evidence includes exposures, attribution, decision processes, personnel, controls, and consistency between stated strategy and observed results.
Worked example: A manager reports strong returns but describes a low-market-exposure process. Estimated beta is consistently high, prompting investigation before the results are accepted as evidence of that process.
Mistake to avoid: Allowing a persuasive interview to override contradictory portfolio evidence.
Reference: CIPM® Program Curriculum | CFA Institute
55. Assessing whether a process is repeatable
A repeatable process links an investment rationale to research, portfolio construction, risk control, and implementation. Investigate whether reported results arose from intended decisions or incidental exposures. Success in one favorable environment does not establish durability. Look for coherent explanations across different conditions, while recognizing that evidence cannot guarantee future skill.
Worked example: A purported stock-selection manager outperforms mainly through a persistent sector overweight. The evaluator separates that allocation exposure from evidence supporting the claimed stock-selection process.
Mistake to avoid: Attributing every favorable outcome to the process described in marketing materials.
Reference: CIPM® Program Curriculum | CFA Institute
56. Team continuity and key-person dependence
A track record belongs to a particular combination of people, responsibilities, resources, and decisions. Examine who actually generated it and whether those conditions remain. Team turnover need not invalidate every historical observation, but it can change how relevant that history is. Assess succession, decision authority, incentives, and dependence on individual contributors.
Worked example: The lead decision-maker leaves while the research team remains. The allocator investigates who controlled position sizing and sell decisions before assigning the old record to the successor.
Mistake to avoid: Assuming an unchanged firm name means the investment capability is unchanged.
Reference: CIPM® Program Curriculum | CFA Institute
57. Capacity and implementation costs
Strategy capacity depends on whether additional assets can be deployed without materially weakening implementation or the investment opportunity. Consider trading volume, market impact, position limits, liquidity, and the breadth of available ideas. Historical results from a small asset base may not scale. Capacity is strategy-specific rather than a universal assets-under-management threshold.
Worked example: A trade represents 1% of a security’s daily volume. If the same proportional position grows tenfold with assets, it becomes 10%, raising implementation concerns even though the strategy is unchanged.
Mistake to avoid: Assuming performance scales proportionally as managed assets grow.
Reference: CIPM® Program Curriculum | CFA Institute
58. Operational due diligence
Investment analysis and operational due diligence address different failure sources. Examine valuation governance, asset custody arrangements, reconciliations, service providers, business continuity, and control responsibilities as relevant. Strong returns do not compensate for weak controls. Evaluate whether processes can detect errors and conflicts, including where duties or oversight need meaningful separation.
Worked example: One employee proposes illiquid asset values and approves them without independent challenge. The evaluator flags the control weakness and investigates oversight before recommending an allocation.
Mistake to avoid: Treating attractive investment results as evidence that operational controls are sound.
Reference: CIPM® Program Curriculum | CFA Institute
59. Diversifying a portfolio of managers
Combining managers requires analysis of their underlying exposures and return relationships, rather than counting different firm names. Low correlation can reduce aggregate volatility, but estimates may change during stress. Evaluate common holdings, factors, liquidity, and active risks alongside correlation so that apparently diverse managers do not create concentrated exposure.
Worked example: Two equally weighted managers each have 10% volatility. With zero correlation, combined volatility is sqrt(0.5² × 10² + 0.5² × 10²) = approximately 7.07%; with correlation of one, it is 10%.
Mistake to avoid: Assuming different manager brands automatically provide independent risks.
Reference: CIPM® Program Curriculum | CFA Institute
60. Monitoring and evidence-based termination
Monitoring tests whether the manager continues to fulfill the intended role. Compare performance explanations with changes in people, process, exposures, costs, and controls. Underperformance can be consistent with a sound strategy, while good returns can conceal mandate drift. Escalation or termination should reflect documented evidence and client needs rather than an arbitrary short-term return cutoff.
Worked example: A manager beats its benchmark by expanding beyond agreed risk constraints. The allocator escalates the mandate breach despite favorable returns and assesses whether corrective action is credible.
Mistake to avoid: Retaining a manager solely because recent returns are strong.
Reference: CIPM® Program Curriculum | CFA Institute
Sources
Source verification:
- CIPM® Program Curriculum | CFA Institute
- CIPM® Program | Certificate in Investment Performance Measurement