Skip to content
Menu
Finabase Learn

Measure and manage portfolio risk Lesson 4 of 14

Portfolio and risk · Measuring return and risk

Sharpe ratio: return per unit of risk, calculated on real funds

The Sharpe ratio formula, a worked calculation on SPY, how nine portfolios ranked from 2005 to 2026, and why a ratio from one decade rarely survives the next.

On this page
  1. The idea
  2. How it works
  3. Where it fails or misleads
  4. Compared with its neighbours
  5. How Finabase does it
  6. Frequently asked questions
  7. Summary
  8. Check your understanding
  9. Exercise

The Sharpe ratio is an investment's return above the risk-free rate, divided by the volatility of that return. It answers a question that a return alone cannot: how much did the investor earn for each unit of risk taken? William Sharpe proposed it in 1966 to compare mutual funds, and it has become the most quoted risk-adjusted measure in finance.

A return of 12% sounds better than 8%. If the first came with swings three times as large, it may have been the worse deal, because the calmer investment could have been scaled up, with borrowing, to earn more at the same risk. The Sharpe ratio puts both on one scale. This lesson calculates it step by step on SPY, compares nine portfolios over 2005 to 2026, and shows how unstable the number is from one period to the next.

The idea

Plot each investment as a point: its volatility on the horizontal axis, and its average return above Treasury bills on the vertical axis. Treasury bills themselves sit at the origin, with no excess return and essentially no volatility.

Now draw a straight line from the origin to a fund's point. The steeper the line, the more return the fund delivered for each point of volatility. The slope of that line is the Sharpe ratio.

Return above T-bills against volatility, 2005–2026 −2% 0% 2% 4% 6% 8% 10% 12% 14% 0% 5% 10% 15% 20% 25% Slope of this line = SPY’s Sharpe ratio EFA’s Sharpe ratio SPY 0.65 QQQ 0.75 EFA 0.33 IWM 0.42 AGG 0.27 TLT 0.15 GLD 0.54 VNQ 0.33 60/40 0.66 Volatility (standard deviation of monthly returns, annualized) Average yearly return above 3-month Treasury bills Monthly total returns, dividends reinvested. The number beside each fund is its Sharpe ratio.
Nine portfolios, January 2005 to October 2026, monthly total returns. Each label shows the fund and its Sharpe ratio; the slope of a line from the origin to a point is that ratio.

Why a line? Because mixing a fund with Treasury bills, or borrowing to buy more of it, moves its point along its own line. Half SPY and half bills has half the excess return and half the volatility. Twice SPY, bought with borrowed money at the bill rate, has double both. The slope stays the same. So the question "which fund was better?" becomes "which line is steeper?", and that does not depend on how much of the fund anyone chose to hold.

In the figure the 60/40 portfolio sits almost exactly on SPY's line. It earned less, but a person who wanted SPY's level of risk could, in principle, have held more of it and ended in the same place.

How it works

The risk-free return here is the 3-month US Treasury bill rate for each month, divided by twelve. Over January 2005 to October 2026 it averaged 1.79% a year, with long stretches near zero and periods above 4%.

A worked example on SPY

The data: SPY's 262 monthly total returns from January 2005 to October 2026.

  1. Average monthly return: 0.946%.
  2. Average monthly Treasury bill return: 0.149%.
  3. Average monthly excess return: 0.946% − 0.149% = 0.797%.
  4. Standard deviation of the monthly excess returns: 4.264%.
  5. Monthly ratio: 0.797 ÷ 4.264 = 0.187.
  6. Annualized: 0.187 × √12 = 0.65.

The same thing in yearly terms: 9.6% of excess return (0.797% × 12) divided by 14.8% of volatility (4.264% × √12). Each point of yearly volatility came with about two-thirds of a point of return above bills.

Why does the ratio grow with √12 and not stay fixed? The average return grows in proportion to time (twelve months earn twelve times one month's average), while volatility grows with the square root of time. Their ratio therefore grows with √12. A Sharpe ratio must always be read with its time unit; the annual one is the convention.

Nine portfolios, one period

Portfolio Return a year (CAGR) Excess over bills Volatility Sharpe ratio
QQQ (Nasdaq-100) 14.8% 13.8% 18.4% 0.75
60% SPY, 40% AGG 7.8% 6.2% 9.4% 0.66
SPY (S&P 500) 10.7% 9.6% 14.8% 0.65
GLD (gold) 10.1% 9.3% 17.4% 0.54
IWM (US small companies) 8.4% 8.2% 19.8% 0.42
EFA (foreign developed stocks) 6.0% 5.5% 16.7% 0.33
VNQ (US real estate) 6.7% 7.1% 21.5% 0.33
AGG (US bonds) 2.9% 1.2% 4.5% 0.27
TLT (long Treasury bonds) 3.0% 2.1% 13.6% 0.15

All figures are total returns with dividends reinvested, before taxes and trading costs. The 60/40 portfolio is rebalanced to its weights every month. The "excess" column is the arithmetic average used by the ratio, which is why it does not equal the compound return minus 1.79%; CAGR: compound annual growth rate, with the formula and its limits explains the difference.

The ranking by return and the ranking by Sharpe ratio differ. Real estate out-earned foreign stocks and tied them on the ratio, because it was more volatile. Long Treasuries earned slightly more than the broad bond fund and scored about half its ratio, because they swung three times as much.

Where it fails or misleads

It is a noisy estimate

Returns are mostly noise around a small average, so the ratio of the two is noisy too. For SPY over its whole history, March 1993 to October 2026, the Sharpe ratio is 0.60. The standard error of that estimate, using the approximation in Andrew Lo's 2002 paper, is about 0.19. A range of two standard errors runs from roughly 0.2 to 1.0, after 33 years of data.

Shorter samples are worse. Splitting 2005 to 2026 in half:

Fund January 2005 – December 2015 January 2016 – October 2026
SPY 0.45 0.84
AGG 0.74 −0.09
GLD 0.43 0.67

In the first half the bond fund had the best ratio of the three. In the second it had a negative one. A three-year Sharpe ratio, the figure printed on many fund fact sheets, is closer to a description of recent weather than a property of the fund.

SPY’s Sharpe ratio over the last 36 months, 1996–2026 -1 0 1 2 Whole period: 0.60 36-month Sharpe ratio 1995 2000 2005 2010 2015 2020 2025
SPY’s Sharpe ratio over the previous 36 months, 1996 to 2026. The whole-period value is 0.60.

Over rolling three-year windows SPY's ratio ranged from −1.11 (the three years to February 2009) to 2.16 (the three years to May 2015), and it was negative in 17% of the windows. The latest reading, 1.43, says the last three years were unusually good, not that the fund has improved.

It counts upside as risk

The denominator is the standard deviation, which rises with large gains as well as large losses. A strategy with occasional big wins is marked down for them. The Sortino ratio: judging a return by its downside only changes the denominator to deal with this.

It can be flattered

Because it sees only average and spread, the ratio is high for any strategy that produces smooth returns most of the time and a rare, large loss. Selling insurance-like options is the standard example: years of small steady gains, a high Sharpe ratio, then one month that takes back several years. Assets that are priced infrequently, such as private funds, also look smoother than they are and show inflated ratios. Fat tails: why extreme market moves are more common than models say covers the general problem.

Negative ratios do not rank sensibly

When the excess return is negative, higher volatility makes the ratio less negative, so the "better" fund is the one that lost the same amount more erratically. Below zero, compare the returns directly.

It ignores what else is in the portfolio

The ratio judges an investment on its own. An asset with a mediocre ratio can still improve a portfolio if it moves differently from everything else in it. Gold's 0.54 understates what it added to a stock portfolio in this period, which is the subject of Correlation in a portfolio: what a correlation matrix tells you.

Compared with its neighbours

Measure Return used Risk used Best suited to
Sharpe ratio Excess over the risk-free rate Standard deviation (all swings) A whole portfolio; comparing funds
Sortino ratio Excess over a target Downside deviation Lopsided return patterns
Treynor ratio Excess over the risk-free rate Beta (market risk only) One holding inside a diversified portfolio
Information ratio Excess over a benchmark Tracking error Judging an active manager
Calmar ratio Compound return Maximum drawdown Strategies judged by their worst loss

Risk-adjusted return compared: Sharpe, Sortino, Treynor, Calmar and M² puts these side by side on the same nine portfolios.

How Finabase does it

Finabase does not calculate a Sharpe ratio for a symbol, a fund or a portfolio, and none of its screens show one. The pieces it does show:

  • Returns. Symbol pages, the screener and the comparison table give price returns over periods from one week to two years. They exclude dividends.
  • The risk-free rate. The Treasury page lists current bill, note and bond yields. The 3-month bill yield is the usual choice.
  • Portfolio results. A portfolio's report lists profit and return by period, with no volatility or risk-adjusted figure.
  • Backtests. The Backtest and Strategy tools report gain, number of trades and win rate. Neither reports a Sharpe ratio or a drawdown.

To compute a ratio for your own portfolio, export its month-end values, calculate the monthly percent changes, subtract one-twelfth of the bill rate from each, and apply the formula in a spreadsheet: AVERAGE(range) / STDEV.S(range) * SQRT(12). With less than three years of months, treat the result as a rough description of the past.

Frequently asked questions

What is a good Sharpe ratio?

For a broad stock index over several decades, 0.4 to 0.6 has been typical; SPY's is 0.60 since 1993. Ratios above 1 over long periods are rare for anything an ordinary investor can buy. A very high ratio over a few years is more often luck, smoothing or hidden tail risk than skill.

Which risk-free rate should I use?

A short Treasury bill rate in the currency of the returns, matched to the period: the 3-month US bill for dollar returns measured monthly. Using zero instead inflates the ratio whenever bill rates are well above zero, as they have been since 2022.

Does leverage improve the Sharpe ratio?

No. Borrowing at the risk-free rate multiplies excess return and volatility by the same factor, so the ratio is unchanged. In practice borrowing costs more than the bill rate and the ratio falls slightly.

Why is my fund's published Sharpe ratio different from my calculation?

Fund sites usually use the last three years of monthly returns, their own choice of risk-free rate, and sometimes returns after fees in a different share class. Different windows alone can move the number by a full point, as the rolling chart shows.

Can the ratio be compared across asset classes?

With care. It compares the reward for total volatility, so it works for whole portfolios. For a single asset that will be one part of a portfolio, its correlation with the rest matters as much as its own ratio.

Summary

The Sharpe ratio divides return above the risk-free rate by the volatility of that return. On a chart of excess return against volatility it is the slope of the line from the origin, and because leverage moves a portfolio along that line it compares investments regardless of how much risk they take. From 2005 to 2026 SPY scored 0.65 and a 60/40 mix 0.66. The number is useful and fragile: it swings widely between periods, it treats large gains as risk, and it rewards strategies that hide rare large losses. It works best over long histories, between similar portfolios, and next to a drawdown figure.

Check your understanding

5 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. Fund A returned 12% a year with 20% volatility. Fund B returned 8% with 10% volatility. Treasury bills paid 2%. Which has the higher Sharpe ratio?

    1. Fund A, because it earned more
    2. Fund B: (8 − 2) ÷ 10 = 0.6 against (12 − 2) ÷ 20 = 0.5
    3. They are equal
    Show the answer

    B. Fund B: (8 − 2) ÷ 10 = 0.6 against (12 − 2) ÷ 20 = 0.5 The ratio divides the return above the risk-free rate by volatility. Fund B earned 0.6 points of excess return per point of volatility and Fund A 0.5.

  2. An investor holds half their money in a fund and half in Treasury bills. Compared with holding only the fund, what happens to the Sharpe ratio?

    1. It halves
    2. It stays the same
    3. It doubles
    Show the answer

    B. It stays the same Mixing with bills halves both the excess return and the volatility, so their ratio is unchanged. The portfolio moves along the same line from the origin, which is why the ratio does not depend on how much risk is taken.

  3. A fund’s three-year Sharpe ratio is 1.4. What is the most reasonable reading?

    1. The fund reliably earns 1.4 points of return per point of risk
    2. The last three years were favourable; the figure may be very different over the next three
    3. The fund is safer than one with a ratio of 0.7
    Show the answer

    B. The last three years were favourable; the figure may be very different over the next three Sharpe ratios from short windows are noisy. SPY’s own three-year ratio has ranged from −1.11 to 2.16, and a bond fund’s went from 0.74 in one decade to −0.09 in the next.

  4. A strategy earns small, steady gains for years by selling options and then loses heavily in one month. Before that month, what would its Sharpe ratio look like?

    1. Low, because the ratio anticipates rare losses
    2. High, because the ratio sees only the average and the spread of the returns so far
    3. Negative
    Show the answer

    B. High, because the ratio sees only the average and the spread of the returns so far Steady gains mean a good average and a small standard deviation. The ratio cannot see a loss that has not yet appeared in the data, which is how it can be flattered.

  5. Why is the monthly Sharpe ratio multiplied by √12 to annualize it?

    1. Average return grows with time and volatility grows with the square root of time
    2. Because there are 12 months and risk is cut in half over a year
    3. It is a convention with no reason behind it
    Show the answer

    A. Average return grows with time and volatility grows with the square root of time Twelve months earn twelve times the average monthly return but only √12 times the monthly standard deviation, so the ratio grows by 12 ÷ √12 = √12.

Exercise

One task to do in Finabase. Nothing is graded or saved.

In the lesson, SPY’s monthly returns from 2005 to 2026 averaged 0.946%, about 11.4% a year, with volatility of 14.8%. Open the Treasury page and read the Last column for the US 3 Month Treasury. Work out (11.4 − that yield) ÷ 14.8. Then work out 11.4 ÷ 14.8, which leaves the risk-free rate at zero. How far apart are the two ratios?

Open the Treasury page
What you should find

With zero the ratio is 0.77. Each point of bill yield takes about 0.07 off, so a 4% bill gives about 0.50, and the period’s own 1.79% average gives the lesson’s 0.65. Leaving out the risk-free rate flatters the ratio most when bill yields are high. This mixes today’s yield with past returns, so it shows the sensitivity and is not a current Sharpe ratio.

Sources and further reading

  1. William F. Sharpe, “Mutual Fund Performance,” Journal of Business, vol. 39, no. 1 (1966)
  2. William F. Sharpe, “The Sharpe Ratio,” Journal of Portfolio Management, vol. 21, no. 1 (1994)
  3. Andrew W. Lo, “The Statistics of Sharpe Ratios,” Financial Analysts Journal, vol. 58, no. 4 (2002)
  4. Board of Governors of the Federal Reserve System — 3-month Treasury bill secondary market rate (TB3MS), via FRED
  5. Finabase chart data — monthly bars and dividends of SPY, QQQ, EFA, IWM, AGG, TLT, GLD and VNQ through October 2026; ratios computed by Finabase