What Is the Treynor Ratio? Formula, Worked Example, and What Counts as Good
The Treynor ratio is your portfolio's return above the risk-free rate divided by its beta. Here's the formula, a worked example on real funds, why a portfolio can rank first on return and last on Treynor, and where it breaks down.
The Treynor ratio is a portfolio's return above the risk-free rate, divided by its beta. It answers one question: how much did you earn for each unit of market risk you carried? A higher number means more reward per unit of that risk. It's the reason a portfolio can beat the index and still score worse than it.
How is the Treynor ratio calculated?
Subtract the risk-free rate from your portfolio's return, then divide by your portfolio's beta.
Treynor ratio = (Rp − Rf) ÷ βp
Rp is the portfolio's return over a stated window, Rf is the risk-free rate over the same window, and βp is the portfolio's beta against a stated benchmark. The numerator is what you earned above cash; the denominator is how much of the market's movement you signed up for.
Jack Treynor introduced the measure in How to Rate Management of Investment Funds (Harvard Business Review, vol. 43, no. 1, 1965, pp. 63–75), where he called it reward to volatility. A CFA Institute review — Measures of Risk-Adjusted Return: Let's Not Forget Treynor and Jensen, Deborah Kidd, CFA, 2011 — describes it and Jensen's alpha as metrics that "isolate the portion of a portfolio's return explained by its sensitivity to market risk."
A ratio without those three labels isn't reproducible. PortLens uses the S&P 500 total-return index as benchmark and the FRED 1-Year Treasury Constant Maturity Rate (DGS1), accessed 11 August 2026, as its risk-free rate over a one-year trailing window by default; see our methodology.
What does the Treynor ratio look like on a real portfolio?
Like this. A growth tilt of 40% VOO, 35% VGT and 25% VUG scored 0.168 over the year to August 10, 2026. The figures use the same modules that power a PortLens scan over the 252 trading days from August 8, 2025 to August 10, 2026, against the S&P 500 total-return index, with daily dividend-adjusted closes. The returns are period returns, not annualized; betas are measured over the same window, not borrowed from elsewhere.
| Fund | Weight (as typed) | Period return | Beta (this window) | Weight × beta |
|---|---|---|---|---|
| VOO | 40% | 22.76% | 0.993 | 0.397 |
| VGT | 35% | 38.16% | 1.682 | 0.588 |
| VUG | 25% | 17.04% | 1.311 | 0.328 |
| Portfolio | 100% | 26.14% | 1.313 |
The shown weights are the allocations as the portfolio stands at the end of the window — what you would type into a scan today. Buy-and-hold measures the dollars that were actually at work, so each weight is rolled back by that holding's own period return to recover its start-of-window share, and the return on those starting amounts is 26.14% — reproducible from the two columns above; see our methodology. DGS1 stood at 4.01% on August 7, 2026. That gives:
(26.14% − 4.01%) ÷ 1.313 = 0.168
Read it as 22.13 percentage points across 1.313 beta, or 16.8 percentage points per 1.0 beta.
Why can two portfolios with similar returns have very different Treynor ratios?
Because the return is only the numerator. Run four more allocations through the identical window, benchmark and rate, and the ranking rearranges itself. The S&P 500 returned 22.80% over this window and DGS1 was 4.01%, so the market's own ratio — beta 1.0 by construction — was 0.188.
| Portfolio | Return | Beta | Treynor | Rank by return | Rank by Treynor |
|---|---|---|---|---|---|
| Growth-tilted (VOO 40 / VGT 35 / VUG 25) | 26.14% | 1.31 | 0.168 | 1 | 5 |
| Dividend and value (VYM 50 / VTV 30 / VOO 20) | 26.11% | 0.66 | 0.333 | 2 | 1 |
| Total market and international (VTI 70 / VXUS 30) | 24.37% | 1.04 | 0.196 | 3 | 2 |
| Core and international (VTI 60 / VXUS 25 / VUG 15) | 23.20% | 1.08 | 0.178 | 4 | 4 |
| S&P 500 index fund only (VOO) | 22.76% | 0.99 | 0.189 | 5 | 3 |
The first two rows are the whole argument. They finished 0.03 percentage points apart on return — 26.14% against 26.11% — yet the growth-tilted book's Treynor is about half the other's, because one carried a beta of 1.31 and the other 0.66.
The growth-tilted book ranks first on return and last on Treynor, below a plain S&P 500 index fund. It beat the market by 3.34 percentage points and needed 31% more market sensitivity to do it. That is the job of the measure: it asks whether the extra return was earned or bought with beta.
Two cautions. The table describes what these funds did over one past window and predicts nothing about any of them — all investing involves risk, including possible loss of principal, and past performance does not guarantee future results. And the dividend-and-value row scores well partly because its measured beta came out low over a window when those holdings moved differently from a tech-led index; a ratio built on beta inherits every weakness of the beta beneath it. What any two of these funds hold in common is a separate question, answered on the VGT vs VUG overlap page.
What is a good Treynor ratio?
There isn't a universal one, and the arithmetic says why: the ratio is denominated in a market return that changes every window, so the same portfolio produces very different numbers without changing a single holding.
The one reference point the arithmetic hands you for free is the market itself. A broad index has a beta of 1.0 against itself by construction, so its Treynor ratio over any window is simply its own return minus the risk-free rate. Over the window above that was 22.80% − 4.01%, or 0.188. It's a comparison point, not a target.
Now move the window. Over the 253 trading days from December 30, 2021 to December 30, 2022 — a market decline, chosen precisely because it was one — the S&P 500 total-return index fell 18.32% and DGS1 stood at 4.73% on December 30, 2022. The market's own ratio over that window was −0.231, and the same five portfolios look like this, calculated under our methodology:
| Portfolio | Return | Beta | Treynor |
|---|---|---|---|
| Dividend and value | −5.03% | 0.77 | −0.127 |
| S&P 500 index fund only | −18.40% | 1.00 | −0.231 |
| Total market and international | −18.74% | 0.95 | −0.248 |
| Core and international | −21.37% | 1.00 | −0.261 |
| Growth-tilted | −26.81% | 1.20 | −0.263 |
The dividend-and-value portfolio scored 0.333 in one window and −0.127 in the other, with nothing about the portfolio changed. A threshold calling 0.333 good and −0.127 bad would be grading the market, not the allocation.
So the useful question isn't "is my Treynor high enough". It's: against what, over which window, and did it beat the market's own ratio over that same window? Two of the four multi-fund portfolios above managed that in the first window; one did in the second.
Is the Treynor ratio a percentage?
No — it's a ratio, and its scale depends entirely on how you wrote the numerator. Express excess return as 22.13 percentage points and the portfolio above gives 16.8; express it as the decimal 0.2213 and you get 0.168. Both describe the identical portfolio, which is why two published Treynor ratios are comparable only when they share a convention, a benchmark, a window and a risk-free rate. PortLens reports the decimal form, computed as described in our methodology.
What does a negative Treynor ratio mean?
It means the portfolio earned less than the risk-free rate over the window — and once the numerator turns negative, the ratio stops ranking portfolios sensibly.
Look again at the 2022 table. The growth-tilted book lost 26.81% while the core-and-international book lost 21.37% — a gap of 5.44 percentage points — yet their ratios are −0.263 and −0.261. Dividing a bigger loss by a bigger beta cancels most of the difference away.
The distortion runs the wrong way, too. Had that portfolio carried a beta of 1.00 instead of 1.20 while losing the same 26.81%, its ratio would have been −0.315 rather than −0.263: more market risk made a losing portfolio look better. No threshold fixes that, so read a negative Treynor ratio as a sign rather than a score.
What's the difference between the Treynor ratio and the Sharpe ratio?
The denominator, and only the denominator. Both divide the same excess return by a measure of risk; Treynor uses beta, Sharpe uses standard deviation. Sharpe's version came second — "in 1966, Sharpe expanded upon Treynor's work to develop the reward-to-variability ratio," as the CFA Institute review puts it.
| Treynor ratio | Sharpe ratio | |
|---|---|---|
| Denominator | Beta against a benchmark | Standard deviation of returns |
| Risk it measures | Market risk only | Total risk, market and stock-specific |
| Needs a benchmark | Yes | No |
Treynor ignores diversifiable risk, so three correlated names can post a respectable ratio while carrying company-specific risk. Sharpe counts that risk but can't separate it from the market's. Neither substitutes for looking at what you actually own.
What doesn't the Treynor ratio tell you?
Almost everything except its one question, and its edges are worth knowing because the output looks so precise.
- Beta is the only risk it sees. The CFA Institute review is blunt about this: "the use of beta as the sole measure of portfolio risk is both the point and the criticism of the Treynor ratio." Idiosyncratic risk is invisible to it by design.
- The denominator is itself a measurement. Beta changes with the benchmark, the window and the return frequency, as portfolio beta shows in detail. Every figure here measures beta over the same window as the return; for these US-listed funds, a PortLens scan uses each fund's published beta instead, so its ratio won't match the tables exactly. That gap is the point, not an exception to it.
- It breaks down at low beta. A near-zero denominator explodes the ratio, and a negative beta inverts its economic sign, making a profitable hedge look terrible. Below a beta of 0.1, PortLens treats the ratio as undefined and scores the component neutrally; the scan card and competition leaderboards show a dash instead of a figure, while Insights surfaces the ratio as a 0.000 sentinel — so 0.000 beside a very low beta there means undefined, not break-even. Low beta isn't low risk either: the same review notes that "a beta of zero does not signify a lack of volatility relative to the market but rather a lack of correlation with market volatility."
- It says nothing about what you hold, or when. Two portfolios with identical Treynor ratios can own entirely different companies, and both tables above describe periods that have already happened.
How do I check my own portfolio's Treynor ratio?
Type your holdings into a free PortLens scan. It obtains a beta for each holding, then computes the weighted portfolio beta, the one-year return and the Treynor ratio against the S&P 500 total-return index, with DGS1 as the risk-free rate. No account is required to start, and entry is manual — PortLens connects to no broker.
The ratio also carries weight inside the product: the overall portfolio score is 40% normalized Treynor, 40% diversification and 20% excess return against the benchmark. Those are transparent design weights rather than weights fitted to historical returns, and the score is not a rating or a recommendation — both are documented in our methodology.
Read beta beside the ratio. A strong-looking figure on a beta of 0.6 may describe a portfolio that barely moved with the market, or a window artifact.
What are the key takeaways?
- The Treynor ratio is excess return over the risk-free rate divided by beta: reward per unit of market risk, not of total risk.
- A portfolio can rank first on return and last on Treynor. The growth-tilted book above returned 26.14% and scored 0.168, below a plain S&P 500 index fund at 0.189.
- There's no universal "good" level. The reference point the arithmetic gives you is the market's own ratio over the same window — 0.188 in one window here, −0.231 in the other.
- Negative Treynor ratios don't rank. A portfolio that lost 5.44 percentage points more than another scored within 0.003 of it, because the bigger loss was divided by a bigger beta.
- Three labels or it's not reproducible: the benchmark, the window and the risk-free rate.
This article is for information and education only and is not investment advice. Analytics referenced are computed as described in our methodology; see our disclosures.