IBF
1988
E S T A B L I S H E D
Mutual Funds & ETFs

Understanding Risk-Adjusted Returns: A Practitioner's Guide

The Bottom LineA risk-adjusted ratio answers a specified measurement question. Compare matched data, benchmarks, targets and cost conventions, then test whether the result matters for the client's whole portfolio. A higher ratio does not independently establish skill, safety, justified fees or a superior after-tax investment.

Two funds can have different returns and volatility yet the same Sharpe ratio. That does not make them interchangeable or settle which one a client should own. It also does not establish that the lower-volatility fund is the only way to obtain a particular expected-return and volatility combination: a funded combination with cash can change the comparison.

Begin with the decision and the feasible alternatives. Then choose a measure that describes the relevant return and risk under explicit conventions, rather than treating a familiar ratio as a universal ranking rule.

Match the Inputs Before Comparing the Scores

Specify the observation frequency, dates, currency, distribution treatment, expenses and return convention. For ratios based on an average, distinguish an arithmetic mean of periodic returns from an annualized compound growth rate. They are different statistics.

Match the risk-free series or benchmark to the same periods. Check the provider's sampling and annualization method; a three-year monthly estimate is not automatically comparable to a five-year quarterly one. Do not subtract an expense again when it is already included in the reported return.

Standard deviation and Sharpe ratios can be calculated for non-normal returns. A normal-distribution assumption is needed for particular probability interpretations, not for the existence of those statistics. Under a hypothetical stable normal distribution with mean 10% and standard deviation 15%, the intervals minus 5% to 25% and minus 20% to 40% contain about 68.27% and 95.45% of outcomes. A historical fund mean and standard deviation do not establish that its future returns obey that distribution or those probabilities.

Look separately at drawdowns, extreme losses, liquidity and valuation practices. No ratio can recover a rare event absent from a short sample simply by changing its denominator.

Sharpe: Excess Return Relative to Variability

The historical Sharpe ratio divides the arithmetic mean of a matched differential-return series by its standard deviation. When comparing with a risk-free return, subtract that return each period. If it is constant, the excess-return and total-return standard deviations are equal; if it varies, use the relevant differential series.

For hypothetical arithmetic mean fund return of 13%, a constant 5% risk-free return and a 6% standard deviation of excess returns at the same frequency, the ratio is 8 divided by 6, or about 1.33. If the differential-return standard deviation is zero, the ordinary ratio is undefined. It describes the specified inputs when that denominator is positive. It does not establish a universally strong score or predict the next period.

Negative ratios show why numerical order needs interpretation. With a constant zero risk-free return, an illustrative mean of minus 5% and standard deviation 5% gives minus 1. A mean of minus 10% and standard deviation 20% gives minus 0.5. The higher ratio accompanies both a lower mean return and more volatility. It is not an automatic reason to prefer that investment.

A Sharpe ratio also omits how a holding combines with other assets and liabilities. Comparisons across strategies need matched data and examination of underlying exposures, nonlinear payoffs and liquidity. The ratio is calculable across categories; that does not make it a universal client preference score.

A Funded Portfolio Counterexample

Use a separate, one-period hypothetical. Fund P has assumed arithmetic expected return 7% and standard deviation 12%. Fund Q has assumed expected return 9% and standard deviation 20%. Cash has an assumed constant 4% return and zero standard deviation. Ignore fees, taxes and trading costs; these are invented forward inputs, not a forecast or a historical fund record.

Both funds have Sharpe ratio 0.25: P uses [7% minus 4%] divided by 12%, and Q uses [9% minus 4%] divided by 20%.

Compare two policies starting with the same $100,000 and no outside cash:

Hypothetical one-period policyInitial holdingsExpected returnStandard deviation
All Fund P$100,000 P7%12%
Fund Q plus cash$60,000 Q and $40,000 cash7%12%

The combination's expected return is 0.60 times 9% plus 0.40 times 4%, or 7%. Its standard deviation is 0.60 times 20%, or 12%, because the cash return is assumed constant. The cash is part of the same initial capital, not extra resources added to Fund Q's path. No borrowing is assumed.

This shows why the lower-volatility fund is not automatically the only way to reach the stated mean and volatility. Matching those two moments does not establish identical distributions, drawdowns, tax treatment or liquidity, and it does not establish either policy as the client's winner. Actual cash returns, account permissions, fund availability and costs must replace the assumptions. A longer horizon also requires a specified rebalancing and cash-flow policy.

Sortino: Define the Target and Downside Convention

The Sortino ratio compares average return above a specified target with downside deviation relative to that target. The target may reflect an objective rather than the risk-free rate. Changing it changes both the question and the result.

For a complete finite-sample illustration, use four invented annual returns: minus 10%, 0%, 10% and 20%. Their arithmetic mean is 5%. Define downside deviation as the square root of the sum of squared below-target shortfalls divided by all four observations; observations meeting the target contribute zero.

Target per annual periodDownside deviation under this conventionMean return minus targetSortino ratio
0%5.00%5.00%1.00
5%About 7.91%0.00%0.00

For target zero, only the minus 10% return contributes a squared shortfall, giving the square root of [100 divided by 4], or 5 percentage points. For target 5%, the shortfalls are minus 15 and minus 5 percentage points; the square root of [(225 plus 25) divided by 4] is about 7.91 percentage points.

The different ratios describe the same four returns against different targets. Neither proves that the investment became safer or worse. Check the formula, observation denominator, frequency and annualization before comparing provider figures. A zero downside denominator needs an undefined-ratio check; a short sample without losses does not establish an absence of future downside.

The finite-sample calculation is descriptive. It does not estimate all losses that could occur outside those four invented observations. Sortino is not a complete tail-risk measure and does not automatically solve stale pricing, missed crises or asymmetric-payoff problems. Symmetry alone also does not require a Sortino ratio to equal a Sharpe ratio.

Treynor: Return Relative to a Particular Beta

The Treynor ratio divides average excess return by beta to a specified benchmark. For hypothetical mean return 10%, constant risk-free return 2% and beta 1.2, it is 8% divided by 1.2, or about 6.67% per unit of beta. Those units differ from a Sharpe ratio's return-per-return-variability units; do not compare the displayed numbers as though they share one scale.

The measure may inform analysis within an appropriate benchmark model, but it does not prove all other risk disappears in the client's portfolio. A sector fund's 5% portfolio weight and high R-squared do not establish that residual issuer or sector exposures are neutralized. Examine weights and covariances with the actual other holdings.

A beta near zero or below zero needs special interpretation, and zero can make the ratio undefined. There is no universal R-squared threshold that validates the coefficient. Sample size, benchmark fit, uncertainty, stability and changing exposures matter.

Information Ratio: Active Return, Not a Fee Certificate

The information ratio divides the arithmetic mean of periodic fund-minus-benchmark returns by the standard deviation of that active-return series, called tracking error. A matched mean active return of 2% and tracking error of 4% gives 0.50 under that frequency convention.

Zero tracking error makes the ordinary ratio undefined. With a positive denominator, it describes active return relative to measured tracking variability. It does not establish skill, isolate an adviser's fee or compare the investor's outcome with a purchasable alternative after all costs and taxes. No universal 0.50 cutoff establishes that the fee was justified.

Check whether the fund series is already net of operating expenses and whether the benchmark is a costless index or an actual alternative investment. A ratio can change with a benchmark or lookback even when the underlying fund holdings do not. A favorable short-period result can reflect an exposure, luck or an unusual market environment.

Keep Alpha and Model Fit in Their Proper Roles

In a one-benchmark excess-return regression, alpha is the fitted mean unexplained return. For hypothetical matched mean inputs of fund return 14%, risk-free return 5%, benchmark return 8% and beta 1.15, the calculation is 9% minus [1.15 times 3%], or 5.55 percentage points. These inputs illustrate the formula; they do not establish skill or a forecast.

R-squared describes sample fit, not a pass/fail truth test. In a simple regression with an intercept, perfect negative correlation can have R-squared of 100% just as perfect positive correlation can. A low beta does not establish low total volatility, and high fit does not validate a future return estimate. For the full definitions and a complete regression example, see Standard Deviation, Beta, Alpha, and Sharpe Ratio Explained.

Choose the Measurement for the Decision

Ask what needs to be compared: the variability of excess returns, shortfalls from a target, sensitivity to a benchmark, active-return consistency or the combination with the client's existing assets. Identify the measure's denominator and the risks it omits.

Then test a plausible alternative. In the P/Q illustration, explicitly retaining cash changes the portfolio comparison. In the Sortino illustration, changing the target changes the score. In the negative Sharpe example, a higher score fails to establish a preferable return-and-volatility combination. These are reasons to examine assumptions, not reject every use of a metric.

Before recommending a trade, compare available alternatives with the same initial resources and a common horizon. Include fees, taxes, distributions, basis, tax-payment funding, cash needs and account restrictions. This article's simplified examples establish no particular client's after-tax winner.

The Client Conversation

Explain what the metric measures: "This ratio summarizes one part of the return history under a particular method. We will check the inputs and the risks it leaves out."

Then make the comparison complete: "A fund with higher volatility can be held at a smaller weight with cash, if that choice is available and fits the plan. A single fund ranking does not show the full portfolio alternatives."

Finally, state the decision limit: "We need your actual costs, tax lots, cash needs and account options before changing investments. The highest score alone does not establish the best choice for you."

Key Takeaways

  1. Match dates, frequency, cost treatment, benchmark and calculation conventions. Arithmetic mean and compound growth are not interchangeable inputs.
  2. A Sharpe ratio is calculable for non-normal returns but does not capture every risk. Interpret negative ratios and probability bands with care.
  3. Compare funded portfolio alternatives, including retained cash. Equal Sharpe ratios do not settle a client preference or guarantee identical outcomes.
  4. State the Sortino target and downside denominator. A changed target can change the score without changing the investment.
  5. Treynor, information ratio and alpha describe particular models or return comparisons. No threshold independently establishes skill or justified fees.
  6. Connect the statistics to feasible, after-tax choices and the client's whole portfolio. A higher ratio is evidence to interpret, not an instruction to trade.

The Advisor's Edge

The useful skill is identifying what a risk-adjusted number answers and constructing the comparison around the client's actual resources. That includes recognizing when cash, a different benchmark or a changed target alters the apparent ranking.

The Certified Fund Specialist curriculum develops risk measurement, expense analysis and portfolio-construction skills across two modules and 21 chapters. Apply those skills to a reviewable choice instead of a universal score threshold. For the steps that connect evidence to an implementable decision, see Building a Fund Selection Framework.

Sources and Notes: Sharpe definition, differential returns and limitations: Sharpe, 1994. Downside targets, calculation conventions and limits: CFA Institute, The Sortino Ratio. Benchmark regressions: Morningstar MPT statistics. All return, probability-band and portfolio examples are hypothetical under the stated assumptions. They do not establish actual fund distributions or client outcomes.

Opening feedback form...

Put It Into Practice

Free tools built from the CFS® curriculum. Take something to work Monday morning.

The Practice Benchmark Series
Fund Selection Framework for Client Portfolios
A systematic approach to evaluating funds across expense ratios, risk-adjusted returns, and manager tenure.
Interactive · 10 min Start the Benchmark →
The Practitioner Brief Series
Expense Ratios: What the Prospectus Doesn't Tell You
The hidden costs beyond the stated expense ratio that experienced fund analysts know to look for.
PDF Guide · Free
The Specialist’s Edge Series
Active vs. Passive: The Data Behind the Debate
Current data on where active management adds value and where it doesn't. For client conversations.
PDF Guide · Free
“The readings and material associated with the CFS was on-point and highly interesting. As I teach an investment course at UM-Dearborn, I will incorporate several of the concepts and applications into the course material. Thanks to the IBF staff for all their help and support.”
Nicholas A. Vlisides, CFS®  ·  Investment Course Instructor  ·  University of Michigan-Dearborn

Every IBF designation is built for the advisor who wants to be better on Monday morning than they were on Friday.