Risk-Adjusted Returns: A Complete Guide with Formulas 2026

What Are Risk-Adjusted Returns?

A risk-adjusted return measures how much return an investment generates per unit of risk taken. It's the single most important concept in portfolio evaluation because raw returns, on their own, tell you almost nothing about whether an investment was actually good.

Consider two funds that both returned 20% last year. Fund A achieved this with 5% annualised volatility. Fund B achieved the same return with 30% volatility. On a raw-return basis they look identical, but Fund A delivered four times more return per unit of risk. Any rational investor would prefer Fund A - the same reward for dramatically less uncertainty.

This is what risk-adjusted return metrics formalise. They divide some measure of return by some measure of risk, producing a ratio that allows fair comparison between investments with different risk profiles. The higher the ratio, the more efficiently the investment converts risk into return.

The concept sits at the heart of modern portfolio theory. William Sharpe's foundational work in the 1960s gave us the first widely adopted risk-adjusted metric, and the framework has expanded steadily since. In 2026, institutional investors, hedge fund allocators, and quantitative traders routinely evaluate performance using multiple risk-adjusted metrics - each capturing a different dimension of the risk-return trade-off.

The risk-adjusted return formula in its most general form is:

Risk-Adjusted Return = Excess Return / Risk Measure

What counts as "excess return" and "risk measure" varies by metric, and that variation is what makes different ratios useful in different contexts.

---

Why Raw Returns Are Misleading

Raw returns ignore risk entirely, and that omission leads to consistently poor investment decisions. A strategy returning 15% per year sounds good until you learn it experienced a 60% drawdown along the way. A fund showing 8% annual returns looks modest until you discover its worst month was -1.2%.

A Concrete Example

Imagine comparing two UK equity strategies over five years:

Strategy A beats Strategy B by 0.6 percentage points on raw return. But Strategy B more than doubles Strategy A's Sharpe ratio, experienced a fraction of the drawdown, and never had a month worse than -3.8%. Most investors would sleep far better holding Strategy B - and could lever it up to match Strategy A's return at a fraction of the risk.

This illustrates the core problem: raw returns reward risk-taking regardless of whether that risk was sensible. Risk-adjusted metrics reward efficient risk-taking.

Survivorship Bias Compounds the Problem

The funds you see in performance tables are the ones that survived. Funds that took excessive risk and blew up don't appear in historical databases. This survivorship bias makes raw returns even more misleading because the highest raw returns in any sample often came from strategies that were simply lucky - taking enormous risk that happened not to materialise during the measurement window.

Risk-adjusted metrics partially correct for this. A fund with high returns but extreme volatility will show a mediocre Sharpe ratio, flagging that the returns may have been achieved through excessive risk rather than genuine skill.

---

The Main Risk-Adjusted Metrics: An Overview

Six risk-adjusted metrics dominate professional practice. Each uses a different definition of risk, making it better suited for certain types of investments and certain types of questions.

Each of these deserves a detailed explanation, which follows below.

---

Sharpe Ratio

The Sharpe ratio is the most widely used risk-adjusted performance metric. It divides a portfolio's excess return over the risk-free rate by the standard deviation of returns.

Sharpe Ratio = (Rp - Rf) / σp

Where:

• Rp = the portfolio's average return
• Rf = the risk-free rate (typically the yield on short-dated government bonds)
• σp = the standard deviation of portfolio returns

When to Use It

The Sharpe ratio works well as a general-purpose comparison tool when return distributions are roughly symmetrical. For diversified equity portfolios, bond funds, and balanced strategies, it's the standard starting point. In 2026, it remains the default risk-adjusted metric reported by virtually every fund and platform.

Strengths

• Universally understood and reported
• Simple to calculate and interpret
• Works well for normally distributed returns
• Easy to compare across asset classes

Limitations

• Penalises upside volatility equally with downside volatility
• Assumes returns are normally distributed (many real strategies have fat tails or skew)
• Sensitive to the choice of risk-free rate
• Can be gamed by strategies that sell tail risk (consistent small gains until a rare catastrophic loss)

A Sharpe ratio above 0.5 is acceptable for a long-only equity strategy, above 1.0 is good, and above 2.0 is exceptional. Hedge funds and quantitative strategies that achieve Sharpe ratios consistently above 2.0 over long periods are rare.

---

Sortino Ratio

The Sortino ratio improves on the Sharpe ratio by replacing total volatility with downside deviation. It only penalises returns that fall below a minimum acceptable threshold, treating upside volatility as a good thing rather than a risk.

Sortino Ratio = (Rp - T) / Downside Deviation

Where:

• Rp = the portfolio's average return
• T = the target or minimum acceptable return (often the risk-free rate or zero)
• Downside Deviation = standard deviation of returns below T, calculated across all periods

When to Use It

The Sortino ratio is most valuable when the return distribution is asymmetric. Options strategies, trend-following systems, and hedge funds frequently produce returns with meaningful skew. A trend-following strategy that generates many small losses and occasional large gains will have a Sortino ratio considerably higher than its Sharpe ratio - correctly reflecting that the large gains aren't risk.

Strengths

• Doesn't penalise upside volatility
• Better reflects investor preferences (losses hurt more than gains help)
• More accurate for skewed return distributions
• Distinguishes between "good" and "bad" volatility

Limitations

• Downside deviation is estimated from fewer data points than standard deviation, making it noisier
• The target return (T) is subjective - different choices produce different results
• Less widely understood than the Sharpe ratio
• Still a volatility-based measure - doesn't capture extreme tail events well

Sortino ratios are typically higher than Sharpe ratios for the same strategy because the denominator excludes upside moves. A Sortino ratio above 1.0 is generally considered good, above 2.0 is strong, and above 3.0 is exceptional.

---

Calmar Ratio

The Calmar ratio divides annualised return by the absolute value of maximum drawdown. It directly addresses the question most investors actually care about: how much return did I earn relative to the worst loss I had to endure?

Calmar Ratio = CAGR / |Maximum Drawdown|

Where:

• CAGR = the compound annual growth rate over the measurement period
• Maximum Drawdown = the largest peak-to-trough decline in portfolio value

When to Use It

The Calmar ratio is particularly relevant for strategies where drawdown control matters - managed futures, systematic macro, and any approach where investors have explicit drawdown limits. Pension funds and endowments that can't tolerate large interim losses often rank managers primarily by Calmar ratio.

Strengths

• Uses maximum drawdown, which is intuitively meaningful to investors
• Captures the worst-case experience directly
• Doesn't rely on distributional assumptions
• Aligns with how allocators actually think about risk

Limitations

• Maximum drawdown is a single data point (the worst episode), making it sensitive to one bad period
• A strategy that had one early drawdown and then performed well will carry that drawdown in its Calmar ratio indefinitely
• Doesn't capture the frequency or duration of drawdowns
• Typically calculated over a rolling 36-month window, which can miss longer-term patterns

A Calmar ratio above 1.0 means the strategy's annualised return exceeds its worst drawdown. Above 2.0 is strong, and above 3.0 indicates exceptional drawdown-adjusted performance.

---

Treynor Ratio

The Treynor ratio replaces total volatility with beta - the portfolio's sensitivity to the market. It measures how much excess return the portfolio generates per unit of systematic (non-diversifiable) risk.

Treynor Ratio = (Rp - Rf) / βp

Where:

• Rp = the portfolio's average return
• Rf = the risk-free rate
• βp = the portfolio's beta relative to the market benchmark

When to Use It

The Treynor ratio is most appropriate when the portfolio is a component of a larger, diversified allocation. In that context, the portfolio's total volatility is less relevant because unsystematic risk gets diversified away at the aggregate level. What matters is how much the portfolio contributes to the overall allocation's risk-return profile - and that's determined by beta.

Strengths

• Appropriate for portfolios held within a broader diversified context
• Focuses on systematic risk, which is the only risk compensated by the market
• Useful for comparing well-diversified portfolios against each other

Limitations

• Requires a valid beta estimate, which assumes a linear relationship with the benchmark
• Not meaningful for undiversified portfolios where unsystematic risk dominates
• Beta can be unstable over time, especially for strategies with dynamic exposures
• Relies on CAPM assumptions, which don't always hold in practice

Higher Treynor ratios indicate better compensation for systematic risk. There's no universal "good" threshold because the scale depends on the market premium and beta estimation method.

---

Information Ratio

The information ratio measures the consistency of a manager's outperformance relative to a benchmark. It divides active return (the difference between portfolio and benchmark returns) by tracking error (the volatility of that active return).

Information Ratio = (Rp - Rb) / Tracking Error

Where:

• Rp = the portfolio's average return
• Rb = the benchmark's average return
• Tracking Error = the standard deviation of (Rp - Rb) over time

When to Use It

The information ratio is the primary metric for evaluating active managers who are benchmarked against a specific index. A UK equity fund benchmarked to the FTSE 100, for example, should be judged on whether it consistently beats that index relative to the risk taken to do so.

Strengths

• Directly measures skill at beating a benchmark
• Accounts for the consistency of outperformance, not just magnitude
• Industry standard for active management evaluation
• Helps distinguish genuine alpha from lucky concentrated bets

Limitations

• Requires a meaningful benchmark - irrelevant for absolute-return strategies
• Can be manipulated by choosing a convenient benchmark
• Tracking error penalises both positive and negative deviations from the benchmark
• A high information ratio doesn't guarantee positive absolute returns

An information ratio above 0.5 is considered good for an active manager, above 0.75 is strong, and above 1.0 is exceptional. Very few managers sustain an information ratio above 1.0 over long periods.

---

Jensen's Alpha

Jensen's Alpha measures the portfolio's return in excess of what the Capital Asset Pricing Model (CAPM) predicts for its level of systematic risk. It isolates the "extra" return that can't be explained by the portfolio's beta exposure.

Jensen's Alpha = Rp - [Rf + βp(Rm - Rf)]

Where:

• Rp = the portfolio's actual average return
• Rf = the risk-free rate
• βp = the portfolio's beta
• Rm = the market's average return

When to Use It

Jensen's Alpha is useful when you want to determine whether a manager is generating genuine value beyond what you'd expect from their market exposure. A portfolio with a beta of 1.2 should return more than the market simply because it takes more market risk. Jensen's Alpha asks whether it returned more than the 1.2x market exposure would predict.

Strengths

• Directly measures manager skill after accounting for market exposure
• Expressed in return units (percentage points), which is intuitive
• Can be extended to multi-factor models (Fama-French Alpha)

Limitations

• Relies entirely on CAPM, which is a simplified model of how markets work
• Beta must be estimated, introducing measurement error
• Doesn't account for non-market risk factors (size, value, momentum)
• Can be misleading if the strategy's risk profile changes over time

A positive Jensen's Alpha means the manager outperformed their CAPM-predicted return. In practice, the multi-factor version (using Fama-French or Carhart factors) is more common because it controls for known sources of return beyond the market.

---

Master Comparison Table

This table summarises all six risk-adjusted metrics side by side:

Which Metric Should You Use?

There's no single correct answer. The right metric depends on the question you're asking:

• "Is this a good investment on a standalone basis?" - Sharpe Ratio
• "Is this strategy efficient at avoiding losses?" - Sortino Ratio
• "Can I tolerate the worst-case loss?" - Calmar Ratio
• "Does this add value to my existing diversified portfolio?" - Treynor Ratio
• "Is the active manager beating their benchmark consistently?" - Information Ratio
• "Is the manager skilled, or just riding market exposure?" - Jensen's Alpha

In practice, professional allocators examine several metrics together. A fund might have an excellent Sharpe ratio but a poor Calmar ratio if it suffered one severe drawdown in an otherwise smooth track record. Looking at just one number always misses something.

---

Calculating Risk-Adjusted Returns in Python

Here's a complete Python implementation that computes all six risk-adjusted metrics for a sample portfolio. This uses NumPy and pandas, standard tools for quantitative analysis.

import numpy as np
import pandas as pd


def sharpe_ratio(
    returns: pd.Series,
    risk_free_rate: float = 0.0,
    periods_per_year: int = 252,
) -> float:
    """Annualised Sharpe ratio."""
    excess = returns - risk_free_rate
    if returns.std(ddof=1) == 0:
        return 0.0
    return (excess.mean() / returns.std(ddof=1)) * np.sqrt(periods_per_year)


def sortino_ratio(
    returns: pd.Series,
    target_return: float = 0.0,
    periods_per_year: int = 252,
) -> float:
    """Annualised Sortino ratio."""
    excess = returns.mean() - target_return
    downside = returns[returns < target_return] - target_return
    downside_var = (downside**2).sum() / len(returns)
    downside_dev = np.sqrt(downside_var)
    if downside_dev == 0:
        return float("inf") if excess > 0 else 0.0
    return (excess / downside_dev) * np.sqrt(periods_per_year)


def calmar_ratio(
    returns: pd.Series,
    periods_per_year: int = 252,
) -> float:
    """Calmar ratio: annualised return / |max drawdown|."""
    cumulative = (1 + returns).cumprod()
    running_max = cumulative.cummax()
    drawdowns = (cumulative - running_max) / running_max
    max_dd = drawdowns.min()
    if max_dd == 0:
        return float("inf")
    annualised_return = (1 + returns.mean()) ** periods_per_year - 1
    return annualised_return / abs(max_dd)


def treynor_ratio(
    returns: pd.Series,
    market_returns: pd.Series,
    risk_free_rate: float = 0.0,
    periods_per_year: int = 252,
) -> float:
    """Annualised Treynor ratio."""
    excess_portfolio = returns - risk_free_rate
    excess_market = market_returns - risk_free_rate
    cov_matrix = np.cov(excess_portfolio, excess_market)
    beta = cov_matrix[0, 1] / cov_matrix[1, 1]
    if beta == 0:
        return 0.0
    return (excess_portfolio.mean() / beta) * periods_per_year


def information_ratio(
    returns: pd.Series,
    benchmark_returns: pd.Series,
    periods_per_year: int = 252,
) -> float:
    """Annualised information ratio."""
    active_return = returns - benchmark_returns
    tracking_error = active_return.std(ddof=1)
    if tracking_error == 0:
        return 0.0
    return (active_return.mean() / tracking_error) * np.sqrt(periods_per_year)


def jensens_alpha(
    returns: pd.Series,
    market_returns: pd.Series,
    risk_free_rate: float = 0.0,
    periods_per_year: int = 252,
) -> float:
    """Annualised Jensen's Alpha."""
    excess_portfolio = returns - risk_free_rate
    excess_market = market_returns - risk_free_rate
    cov_matrix = np.cov(excess_portfolio, excess_market)
    beta = cov_matrix[0, 1] / cov_matrix[1, 1]
    alpha_per_period = excess_portfolio.mean() - beta * excess_market.mean()
    return alpha_per_period * periods_per_year


# --- Generate sample data ---
np.random.seed(42)
n_days = 504  # Two years of trading days
dates = pd.bdate_range("2024-01-02", periods=n_days)

# Portfolio: slight positive skew, moderate volatility
portfolio_returns = pd.Series(
    np.random.normal(0.0004, 0.011, n_days)
    + np.random.exponential(0.0005, n_days),
    index=dates,
)

# Market benchmark: higher volatility, symmetric distribution
market_returns = pd.Series(
    np.random.normal(0.0003, 0.013, n_days),
    index=dates,
)

# Risk-free rate: ~4.5% annualised = ~0.018% per day
rf_daily = 0.045 / 252

# --- Calculate all metrics ---
results = {
    "Sharpe Ratio": sharpe_ratio(portfolio_returns, rf_daily),
    "Sortino Ratio": sortino_ratio(portfolio_returns, target_return=0.0),
    "Calmar Ratio": calmar_ratio(portfolio_returns),
    "Treynor Ratio": treynor_ratio(
        portfolio_returns, market_returns, rf_daily
    ),
    "Information Ratio": information_ratio(
        portfolio_returns, market_returns
    ),
    "Jensen's Alpha (ann.)": jensens_alpha(
        portfolio_returns, market_returns, rf_daily
    ),
}

print("Risk-Adjusted Performance Metrics")
print("=" * 42)
for name, value in results.items():
    print(f"  {name:<26s} {value:>8.3f}")

# --- Summary statistics ---
ann_return = (1 + portfolio_returns.mean()) ** 252 - 1
ann_vol = portfolio_returns.std(ddof=1) * np.sqrt(252)
cumulative = (1 + portfolio_returns).cumprod()
max_dd = ((cumulative - cumulative.cummax()) / cumulative.cummax()).min()

print(f"\nAnnualised Return:   {ann_return:.2%}")
print(f"Annualised Volatility: {ann_vol:.2%}")
print(f"Maximum Drawdown:      {max_dd:.2%}")

Notes on This Implementation

• Annualisation follows standard conventions: multiply Sharpe, Sortino, and Information ratios by sqrt(periods_per_year); multiply Treynor and Jensen's Alpha by periods_per_year directly.
• The risk-free rate is expressed as a per-period figure matching the return frequency. With daily data and a 4.5% annual risk-free rate, the daily figure is 0.045 / 252.
• Beta estimation uses the covariance method: β = Cov(Rp, Rm) / Var(Rm). For production code, you'd want to add rolling windows and potentially use more sophisticated regression techniques.
• The portfolio has a small exponential component to create positive skew, demonstrating how the Sortino ratio exceeds the Sharpe ratio for positively skewed distributions.

For production use, add input validation, handle edge cases such as empty series or constant returns, and consider using logarithmic returns for better statistical properties over longer horizons.

---

How Fund Managers Use These Metrics

Risk-adjusted metrics aren't academic curiosities - they drive real allocation decisions worth billions of pounds. Here's how professionals apply them in practice.

Portfolio Evaluation

The first application is straightforward: assessing whether an existing portfolio or strategy is performing well. A portfolio manager running a UK equity long-short strategy will track Sharpe, Sortino, and Calmar ratios monthly, comparing them against both absolute thresholds and peer group averages. A declining Sharpe ratio over a rolling 12-month window signals deteriorating risk-adjusted performance, even if raw returns are still positive.

Institutional allocators conducting due diligence on hedge funds typically require a minimum Sharpe ratio (often 1.0 or above) and a minimum Calmar ratio before even considering an allocation. These thresholds filter out strategies that generate returns primarily through excessive risk-taking.

Allocation Decisions

Risk-adjusted metrics directly inform how capital gets distributed. A fund-of-funds allocator choosing between ten potential managers will rank them by Sharpe or Sortino ratio (depending on the strategy type) and concentrate capital in the highest-ranked managers. The insight is simple but powerful: a pound allocated to a high-Sharpe strategy generates more return per unit of risk than a pound allocated to a low-Sharpe strategy.

This extends to asset class allocation as well. If emerging market equities have a Sharpe ratio of 0.3 and investment-grade bonds have a Sharpe ratio of 0.5, the bonds are more efficient on a risk-adjusted basis. An investor could lever up the bond allocation to match the equity return with less total risk - the essence of the risk-parity approach.

Performance Attribution

Information ratios and Jensen's Alpha help decompose where returns actually came from. A UK equity manager who returned 12% when the FTSE All-Share returned 10% has 2% of active return. But if that active return came with 8% tracking error, the information ratio is only 0.25 - suggesting the outperformance was more luck than skill. If the tracking error were 2%, the information ratio would be 1.0, indicating consistent and deliberate alpha generation.

Jensen's Alpha takes this further by stripping out the market-beta component entirely. A manager who simply ran a high-beta portfolio during a bull market would show positive active returns but zero (or negative) Jensen's Alpha, revealing that the outperformance was entirely explained by market exposure rather than stock selection.

Understanding risk management frameworks is essential context for applying these metrics correctly in professional settings.

---

Common Mistakes When Using Risk-Adjusted Metrics

Even experienced practitioners make errors when calculating and interpreting these ratios. Here are the most frequent mistakes and how to avoid them.

Comparing Metrics Across Different Time Periods

A Sharpe ratio calculated over a two-year bull market is not comparable to one calculated over a full market cycle. Short measurement windows can produce extremely misleading figures. A strategy that happened to launch at the start of a strong uptrend will show an inflated Sharpe ratio until it experiences its first significant downturn.

The fix: Always compare metrics over the same time period, and prefer windows that include both rising and falling markets. Three to five years is the minimum for meaningful comparison, and longer is better.

Ignoring Non-Normal Distributions

The Sharpe ratio assumes returns are normally distributed. In reality, many strategies have fat tails (excess kurtosis) and skew. A strategy that sells far out-of-the-money options produces returns with severe negative skew and high kurtosis - frequent small gains punctuated by rare catastrophic losses. Its Sharpe ratio will look attractive right up until the tail event materialises.

The fix: Supplement the Sharpe ratio with the Sortino ratio, Calmar ratio, and direct examination of the return distribution (skewness, kurtosis, histogram). If the distribution is meaningfully non-normal, give more weight to metrics that account for this.

Using the Wrong Risk-Free Rate

The risk-free rate should match the currency and tenor of the investment. A UK-based strategy should use UK gilt yields, not US Treasury yields. Using an outdated or mismatched risk-free rate can meaningfully distort the Sharpe and Treynor ratios.

The fix: Use the yield on 3-month government bills in the portfolio's base currency, updated to the current rate. In 2026, with UK base rates around 3.5-4%, the daily risk-free rate is roughly 0.014-0.016% - not negligible.

Annualising Incorrectly

A common error is annualising the numerator and denominator separately using different methods, or forgetting to annualise at all. The correct approach for the Sharpe and Sortino ratios is to multiply the per-period ratio by the square root of the number of periods per year. For the Calmar ratio, the numerator should already be a CAGR (which is inherently annual), and the denominator is a percentage, so no additional annualisation is needed.

The fix: Be explicit about annualisation in your code and documentation. State whether reported figures are annualised or per-period, and which convention was used.

Overfitting to a Single Metric

No single ratio tells the complete story. A fund with a Sharpe ratio of 2.0 might have achieved it through a strategy that's highly sensitive to a specific market regime. Its Calmar ratio might be poor if it suffered one deep drawdown. Its information ratio might be mediocre if most of the return came from beta rather than alpha.

The fix: Always examine at least three metrics together - typically Sharpe (overall efficiency), Sortino or Calmar (downside risk), and Information Ratio or Jensen's Alpha (skill vs market exposure). Look at the full return distribution alongside the ratios.

Survivorship and Look-Ahead Bias

When backtesting strategies, accidentally including securities that didn't exist at the time of the simulated trade (look-ahead bias) or excluding securities that were delisted (survivorship bias) will inflate risk-adjusted metrics. These biases produce backtests that look far better than what would have been achievable in practice.

The fix: Use point-in-time databases that include delisted securities and corporate actions. When comparing backtested metrics to live performance, expect degradation of 30-50% in risk-adjusted ratios as a reasonable baseline.

---

Frequently Asked Questions

What is a risk-adjusted return in simple terms?

A risk-adjusted return is a measure of how much profit an investment made relative to the risk it took. Think of it this way: earning 10% by holding a savings account (zero risk) is far more impressive than earning 10% by putting everything on a single speculative stock (enormous risk). Risk-adjusted return ratios formalise this comparison by dividing return by risk, so you can fairly compare investments that take different amounts of risk. The most common version is the Sharpe ratio, which divides excess return by volatility.

Which is the best risk-adjusted return metric?

There's no single best metric - it depends on your situation. The Sharpe ratio is the most widely used and is a good starting point for general comparison. The Sortino ratio is better when returns are skewed (common in options and trend-following strategies) because it only penalises downside volatility. The Calmar ratio is preferred by allocators who care primarily about drawdowns. The Treynor ratio is appropriate when the portfolio sits within a larger diversified allocation. In practice, institutional investors look at multiple metrics together rather than relying on any single one.

How do you calculate the risk-adjusted return formula?

The general risk-adjusted return formula is: Excess Return divided by a Risk Measure. For the Sharpe ratio specifically, this is (Portfolio Return - Risk-Free Rate) / Standard Deviation of Portfolio Returns. To annualise from daily data, multiply by the square root of 252 (trading days per year). For example, if a portfolio has a daily mean excess return of 0.04% and daily standard deviation of 0.8%, the annualised Sharpe ratio is (0.04 / 0.8) x sqrt(252) = 0.05 x 15.87 = 0.79. The Python code in this article shows how to compute all six major metrics programmatically.

What is a good risk-adjusted return?

For the Sharpe ratio, above 0.5 is acceptable, above 1.0 is good, and above 2.0 is excellent. For the Sortino ratio, the thresholds are slightly higher because downside deviation is typically smaller than total standard deviation - above 1.0 is decent, above 2.0 is strong. For the Calmar ratio, above 1.0 means annualised returns exceed the worst drawdown, which is a reasonable minimum standard. These benchmarks vary by strategy type and market conditions. A long-only equity strategy with a Sharpe of 0.8 is performing well; a market-neutral hedge fund with the same figure would be considered mediocre.

Why do risk-adjusted returns matter more than raw returns?

Raw returns tell you nothing about the risk taken to achieve them. A fund that returned 30% by concentrating in a single sector during a boom period took enormous risk that happened to pay off. The same approach could easily have lost 30%. Risk-adjusted metrics reveal whether the return was achieved efficiently or recklessly. They also enable fair comparison between strategies with different risk profiles - a 10% return with 5% volatility (Sharpe 2.0) is objectively better risk-adjusted performance than a 15% return with 20% volatility (Sharpe 0.75), even though the raw return is lower. Professional allocators almost never look at raw returns in isolation.

Can risk-adjusted returns be negative?

Yes. A negative Sharpe ratio means the portfolio underperformed the risk-free rate - you would have been better off holding government bonds. A negative Sortino ratio means returns fell below the minimum acceptable target. A negative information ratio means the active manager underperformed their benchmark. Negative risk-adjusted returns are a clear signal that the strategy isn't compensating investors for the risk being taken. The only exception is Jensen's Alpha, where a slightly negative value might simply mean the manager matched their CAPM-predicted return without adding extra value - not necessarily a disaster, but not evidence of skill either.