Portfolio Performance Measurement and Benchmarks

Why This Topic Matters

Treasury professionals are stewards of corporate cash and investment portfolios, and the Certified Treasury Professional (CTP) exam tests whether you can evaluate how well those portfolios are actually doing. It is not enough to earn a positive return; a treasurer must be able to answer the question: Did we earn an appropriate return given the risk we took and the constraints of our investment policy? Portfolio performance measurement and benchmarking provide the discipline to answer that question objectively.

Within the AFP CTP body of knowledge, this concept sits under Capital Structure and Long-Term Capital because the management of investment portfolios (short-term operating cash, strategic reserves, pension assets, and other long-term funds) is directly linked to the firm's overall financing and capital decisions. Boards, audit committees, and investment committees rely on performance reports to judge both internal treasury staff and external investment managers. Poor measurement can hide excessive risk-taking, reward luck instead of skill, or lead to the firing of a manager who was actually doing exactly what the policy asked. On the exam, this topic appears in conceptual questions, calculation questions, and scenario-based questions about selecting the right benchmark.

What It Is

Portfolio performance measurement is the process of calculating the return earned on an investment portfolio over a defined period and assessing that return in light of the risk assumed. Benchmarking is the comparison of that return (and risk) against a reference standard, typically a market index, a peer group, a target rate, or a custom composite, that represents what a passive or comparable investor could have earned.

The core components you must understand are:

1. Return measurement – how to compute holding period return, total return, annualized return, time-weighted return, and money-weighted (dollar-weighted) return.
2. Risk measurement – standard deviation, beta, duration, and tracking error.
3. Risk-adjusted performance – the Sharpe ratio, Treynor ratio, Jensen's alpha, and information ratio.
4. Benchmark selection – the characteristics of a good benchmark and the common indices used for short-term and long-term portfolios.
5. Performance attribution – explaining why a portfolio outperformed or underperformed its benchmark.
6. Governance – linking performance evaluation back to the Investment Policy Statement (IPS).

How It Works: Measuring Return

Holding Period Return (HPR). The most basic measure is the total return over a single period:

HPR = (Ending Value − Beginning Value + Income Received) / Beginning Value

For a bond or money market instrument, income is the interest or coupon received; for equities, it is dividends. Total return always includes both the price change and the income component. A common exam trap is a question that provides both pieces but tempts you to use only one.

Annualizing returns. Short-term treasury portfolios are often measured over periods shorter than a year. To compare them with annual benchmark rates, returns are annualized. The simple (money-market convention) approach multiplies by 365 (or 360) divided by the days held. The compound approach raises (1 + HPR) to the power of 365/days and subtracts 1. Know which convention a question is asking for; the CTP exam frequently uses a 365-day year for yields unless it specifies a 360-day discount basis.

Time-Weighted Return (TWR). TWR breaks the measurement period into sub-periods at each point when cash flows into or out of the portfolio, computes the return for each sub-period, and geometrically links them: (1 + r1)(1 + r2)...(1 + rn) − 1. Because it removes the effect of the timing and size of contributions and withdrawals, TWR is the industry standard for evaluating an investment manager, who typically does not control when the corporation adds or removes cash.

Money-Weighted Return (MWR) or Dollar-Weighted Return. This is the internal rate of return (IRR) of all cash flows into and out of the portfolio. It is heavily influenced by when money was invested. MWR is appropriate for measuring the experience of the investor or the fund as a whole, because the investor does control the timing of cash flows. If a treasurer poured a large deposit into the portfolio right before a market decline, MWR will be lower than TWR.

Exam distinction to memorize: TWR evaluates the manager; MWR evaluates the investor's actual dollar experience. When the question asks which measure is unaffected by the timing of external cash flows, the answer is TWR.

How It Works: Measuring and Adjusting for Risk

A return cannot be judged in isolation. Two portfolios may both earn 5 percent, but if one did so with Treasury bills and the other with speculative bonds, the first manager performed far better. Treasury professionals use several risk-adjusted metrics:

Sharpe Ratio = (Portfolio Return − Risk-Free Rate) / Standard Deviation of Portfolio Returns. It measures excess return per unit of total risk. A higher Sharpe ratio is better. Use it when the portfolio represents the investor's entire holdings or when total volatility is the concern.

Treynor Ratio = (Portfolio Return − Risk-Free Rate) / Portfolio Beta. It measures excess return per unit of systematic (market) risk. Use it when the portfolio is one piece of a well-diversified total portfolio, because unsystematic risk has already been diversified away.

Jensen's Alpha = Actual Portfolio Return − [Risk-Free Rate + Beta × (Market Return − Risk-Free Rate)]. Alpha is the return earned above what the Capital Asset Pricing Model (CAPM) predicts for the portfolio's beta. Positive alpha indicates value added by the manager; negative alpha indicates underperformance relative to risk taken.

Information Ratio = (Portfolio Return − Benchmark Return) / Tracking Error. Tracking error is the standard deviation of the difference between portfolio and benchmark returns. The information ratio tells you how consistently a manager beats the benchmark per unit of active risk. It is the preferred measure for evaluating active managers against a specific index.

Duration is the key risk measure for fixed-income treasury portfolios. Comparing the portfolio's duration to the benchmark's duration reveals whether the manager is taking more or less interest rate risk than the policy intends. A portfolio that outperforms its benchmark in a falling-rate environment while running a longer duration has not necessarily demonstrated skill; it has simply taken a bet on rates.

How It Works: Selecting an Appropriate Benchmark

A benchmark must reflect what the portfolio is supposed to do. The widely cited characteristics of a valid benchmark are that it should be:

• Unambiguous – the securities and weights are clearly identified.
• Investable – it is possible to actually hold the benchmark passively.
• Measurable – its return can be calculated frequently.
• Appropriate – it is consistent with the manager's style and the portfolio's objectives, constraints, and risk tolerance.
• Reflective of current investment opinions – the manager has knowledge of the securities in it.
• Specified in advance – it is agreed upon before the evaluation period begins, not chosen after the fact to make results look good.

Typical benchmarks by portfolio type:

Short-term operating cash and liquidity portfolios: 3-month U.S. Treasury bill yields or a T-bill index, the Federal Funds rate, SOFR (Secured Overnight Financing Rate, which has replaced LIBOR), a 1–3 year government or government/credit index, or institutional money market fund peer averages (such as those published by iMoneyNet or Crane Data). These portfolios prioritize safety and liquidity, so the benchmark should have very low duration and high credit quality.

Strategic or core cash with a longer horizon: short-term bond indices such as a 1–3 year or 1–5 year government/credit index.

Long-term portfolios (pension, foundation, long-term reserves): broad bond indices such as the Bloomberg U.S. Aggregate Bond Index, broad equity indices such as the S&P 500, Russell 2000, or MSCI indices for international equity, and blended composite benchmarks weighted according to the policy asset allocation (for example, 60 percent S&P 500 and 40 percent Bloomberg Aggregate).

Peer group comparisons rank the portfolio against similar funds or managers. They are intuitive but suffer from survivorship bias (poor performers drop out), inconsistent risk profiles across the group, and lack of investability. The exam generally treats a market index that matches policy constraints as superior to a peer group as a primary benchmark.

Absolute or target return benchmarks (for example, inflation plus 2 percent, or a specified hurdle rate) are sometimes used for liability-driven or reserve portfolios. They are easy to communicate but do not tell you whether the result was good relative to market conditions.

Benchmark mismatch is a frequent exam theme. If a conservative treasury policy permits only government and high-grade securities with a maximum maturity of one year, benchmarking against a 5-year corporate index is inappropriate because the manager cannot own what is in the index. Conversely, benchmarking a 2-year duration portfolio against overnight rates will make the manager look like a genius in a steep yield curve environment while hiding the interest rate risk being taken.

How It Works: Performance Attribution

Attribution decomposes the difference between portfolio and benchmark return into its sources. For fixed-income portfolios, the main components are:

• Duration or interest rate effect – return from being longer or shorter than the benchmark as rates moved.
• Yield curve effect – return from positioning along the curve (barbell versus bullet).
• Sector allocation effect – return from overweighting or underweighting sectors such as Treasuries, agencies, corporates, or asset-backed securities.
• Security selection effect – return from choosing specific issues within a sector that performed better or worse than their sector average.
• Trading and cost effect – transaction costs, fees, and timing of trades.

For equity or balanced portfolios, attribution typically separates asset allocation effect (deviating from policy weights among asset classes) from selection effect (picking securities within each class). Attribution tells the treasurer whether outperformance came from deliberate, repeatable skill or from a single risk bet that may reverse.

How It Works: Linking to the Investment Policy Statement

The IPS is the foundation. It states the portfolio's objectives (safety, liquidity, yield, in that order for most operating cash), permitted instruments, credit quality minimums, maturity and duration limits, diversification limits, and the benchmark(s) against which performance will be judged. Good governance requires that:

• The benchmark is written into the IPS before the measurement period.
• Performance is reported at regular intervals (monthly or quarterly) to the investment committee.
• Reports include both return and risk measures and a compliance check against policy limits.
• Fees are disclosed so that gross and net returns can both be evaluated.
• External managers are evaluated over a full market cycle (often three to five years) rather than on a single quarter, to separate skill from noise.

Many organizations and managers follow the Global Investment Performance Standards (GIPS) issued by the CFA Institute, which standardize how returns are calculated and presented, requiring time-weighted returns, consistent composite construction, and full disclosure.

Worked Examples

Example 1 – Holding period and annualized return. A treasury portfolio begins the quarter at 50,000,000, receives 375,000 of interest, and ends with a market value of 50,100,000 before considering the interest. HPR = (50,100,000 − 50,000,000 + 375,000) / 50,000,000 = 475,000 / 50,000,000 = 0.95 percent for the quarter. Simple annualized return = 0.95 percent × 4 = 3.80 percent (or 0.95 × 365/91 ≈ 3.81 percent using days). Compound annualized = (1.0095)^4 − 1 ≈ 3.85 percent.

Example 2 – Sharpe ratio comparison. Portfolio A returned 6 percent with a standard deviation of 8 percent; Portfolio B returned 8 percent with a standard deviation of 14 percent; the risk-free rate is 2 percent. Sharpe A = (6 − 2)/8 = 0.50; Sharpe B = (8 − 2)/14 = 0.43. Although B earned more, A delivered more return per unit of risk and is the better risk-adjusted performer.

Example 3 – Jensen's alpha. A portfolio with a beta of 0.8 returned 7 percent. The market returned 9 percent and the risk-free rate was 3 percent. Expected return = 3 + 0.8 × (9 − 3) = 7.8 percent. Alpha = 7 − 7.8 = −0.8 percent. Despite a positive absolute return, the manager underperformed on a risk-adjusted basis.

Example 4 – Time-weighted vs. money-weighted. A portfolio earns 10 percent in year 1 and −5 percent in year 2. TWR = (1.10)(0.95) − 1 = 4.5 percent over two years. If the company tripled the portfolio size at the start of year 2, the MWR would be well below the TWR because most of the dollars experienced only the negative year. The manager should be judged on the 4.5 percent TWR, not the lower MWR.

Exam Tips: Answering Questions on Portfolio Performance Measurement and Benchmarks

1. Identify what is being evaluated: the manager or the investor. If the question involves judging an external manager or isolating investment skill, the answer involves time-weighted return. If it asks about the actual return earned on the company's dollars or the impact of deposit timing, the answer involves money-weighted (IRR) return.

2. Always include income in total return. When a question gives you beginning value, ending value, and interest or dividends received, add the income. Watch for distractor answers that reflect price change only.

3. Match the risk-adjusted ratio to the denominator. Sharpe uses standard deviation (total risk). Treynor uses beta (systematic risk). Information ratio uses tracking error (active risk relative to the benchmark). Jensen's alpha is a difference in returns, not a ratio. If a question states that the portfolio is one of many well-diversified holdings, lean toward Treynor; if it is the investor's only portfolio, lean toward Sharpe; if it concerns consistency of outperformance against an index, lean toward the information ratio.

4. A higher return is not automatically better. The CTP exam loves scenarios in which the higher-returning portfolio has far more risk. Compute the risk-adjusted measure before choosing. The phrase 'on a risk-adjusted basis' is a signal to calculate Sharpe, Treynor, or alpha.

5. Benchmark selection questions hinge on the word appropriate. The correct benchmark matches the portfolio's permitted securities, credit quality, duration, and objectives as stated in the IPS. Eliminate any answer where the manager could not legally or practically hold the benchmark's constituents. For short-term operating cash, think T-bills, money market fund averages, or short government indices, not the S&P 500 or long bond indices.

6. Remember the benchmark must be specified in advance. Any answer suggesting a benchmark is selected after reviewing results is wrong from a governance standpoint.

7. Watch the duration comparison. If a portfolio beat its benchmark during a period of falling rates and the question notes the portfolio had a longer duration than the benchmark, the outperformance is attributable to interest rate risk, not security selection. The reverse holds for rising rates.

8. Know the day-count convention asked for. Money market yields and bond equivalent yields use 365 days; discount yields (such as the T-bill discount rate) use 360 days. Annualization questions will specify or imply which to use; read carefully.

9. Understand peer group weaknesses. If a question asks for the disadvantage of peer group comparisons, think survivorship bias, inconsistent risk and style across peers, and non-investability. If it asks for an advantage, think ease of understanding and relevance to how similar funds performed.

10. Attribution questions test cause, not amount. When asked where outperformance came from, map the clue to the category: overweight corporates that rallied is sector allocation; chose better bonds within corporates is security selection; ran longer duration as rates fell is duration or interest rate effect.

11. Link everything back to the IPS and governance. The best answer in a treasury context often includes reporting to the investment committee, checking compliance with policy limits, evaluating net of fees, and assessing over a full market cycle rather than one quarter.

12. Eliminate extreme or absolute language. Answers that say a single metric 'always' identifies the best manager or that absolute return 'alone' is sufficient are usually incorrect. Performance evaluation on the CTP exam is about return and risk and policy consistency together.

Quick Formula Summary

HPR = (End − Begin + Income) / Begin
Simple annualized = HPR × (365 / days)
Compound annualized = (1 + HPR)^(365/days) − 1
TWR = (1 + r1)(1 + r2)...(1 + rn) − 1
MWR = IRR of all cash flows
Sharpe = (Rp − Rf) / σp
Treynor = (Rp − Rf) / βp
Jensen's Alpha = Rp − [Rf + βp(Rm − Rf)]
Information Ratio = (Rp − Rb) / Tracking Error

Final Takeaway

Portfolio performance measurement and benchmarking give treasury professionals a disciplined, defensible way to judge investment results. For the CTP exam, master the distinction between time-weighted and money-weighted returns, be fluent with the four main risk-adjusted measures and when each applies, know the characteristics of a valid benchmark and the common indices used for short-term and long-term portfolios, and always evaluate results against the risk taken and the constraints of the Investment Policy Statement. If you can explain not just what a portfolio earned but why and at what risk, you are thinking exactly the way the exam expects a treasury professional to think.