# Treynor Ratio, Jensen Alpha & Modigliani M-Squared Formula

> Master systematic vs total risk attribution: derive Treynor Ratio, Fama-French Jensen Alpha, and Modigliani M-Squared (M2) basis-point scaling.

Canonical: https://quantcalctools.com/guides/treynor-ratio-jensens-alpha-m-squared-modigliani-leverage-math
Last updated: 2026-11-14T19:45:35.617Z

**Publication:** QuantCalc Tools
**Category:** Cagr Irr Drawdown Compounding
**Published:** 2026-11-13T10:40:25.915Z
**Author:** Elena Rostova, FRM, CQF (FRM, Certificate in Quantitative Finance (CQF))
**Reviewer:** Not recorded

## Answer-first summary

When adding a sleeve to a diversified multi-manager book, systematic beta matters more than standalone sigma; master exact Treynor, Jensen Alpha, and Modigliani M-Squared (M2) equations.

## Key takeaways

- Franco & Leah Modigliani’s M-Squared M^2 = r_f + SR_p * sigma_m ranks standalone portfolios 100% identically to the Sharpe Ratio while expressing outperformance in intuitive annualized basis points.
- When evaluating a candidate satellite sleeve inside a diversified multi-manager pension plan, rank by Treynor-Black Appraisal Ratio (alpha_J / sigma_epsilon) rather than standalone Sharpe!
- Raw Jensen’s Alpha scales linearly with leverage (2x leverage doubles alpha_J without doubling skill) and conflates Fama-French Size/Value/Momentum tilts unless stripped via multi-factor regression.
- Never apply the Treynor Ratio to market-neutral (beta ~= 0) or short-biased (beta < 0) funds, where dividing by near-zero or negative beta produces meaningless explosions.

## Related financial entities

- **Continuous Fractional Half-Kelly Portfolio Leverage Criterion** (HALF-KELLY-LEVERAGE) — product: Continuous Fractional Half-Kelly Portfolio Leverage Criterion (HALF-KELLY-LEVERAGE) refers to an institutional financial benchmark and regulatory standard monitored by BIS Quantitative Working Paper (INTL), currently calibrated at 1.635x (Optimal Gross Leverage) as of October 2026.
- **Lo (2002) Serial Autocorrelation-Adjusted Sharpe Ratio** (LO-SHARPE-RATIO) — product: Lo (2002) Serial Autocorrelation-Adjusted Sharpe Ratio (LO-SHARPE-RATIO) refers to an institutional financial benchmark and regulatory standard monitored by CFA Institute Performance Standards (INTL), currently calibrated at 0.3842 (Annualized Ratio) as of October 2026.
- **Secured Overnight Financing Rate (SOFR)** (SOFR) — product: Broad volume-weighted median measure of the cost of borrowing cash overnight collateralized by U.S. Treasury securities on an ACT/360 basis.

## Analysis

Investment committee debates routinely stall over **which classical risk-adjusted metric actually applies to the decision on the table**: if Fund A has a higher **Sharpe Ratio** (`0.85` vs `0.72`) while Fund B has a higher **Treynor Ratio** and higher **Jensen's Alpha**, which fund should the pension plan hire? And how do you explain a `0.85` vs `0.55` Sharpe difference in intuitive **annualized percentage points** to non-quant trustees? Answering both questions requires separating **standalone total-portfolio choice** (Sharpe & Franco/Leah Modigliani's **`M-Squared` / `M2`**) from **marginal sub-portfolio sleeve allocation** (Treynor & Treynor-Black Appraisal Ratio). Run side-by-side ratio checks in our [Sharpe & Sortino Calculator](/calculators/sharpe-sortino-calculator) and anchor in our [CAGR, IRR & Drawdown Hub](/guides/cagr-irr-volatility-drag-sharpe-sortino-drawdown-recovery-guide).

## Unifying Sharpe, Modigliani $M^2$, Treynor, and Jensen's Alpha on One Balance Sheet

Let $R_p$ be portfolio return, $R_m$ be market benchmark return (with volatility $\sigma_m$), and $r_f$ be the risk-free [SOFR / T-bill rate](/entities/sofr-benchmark-rate). Decomposing total variance into systematic beta risk plus idiosyncratic residual variance $\sigma_p^2 = \beta_p^2 \sigma_m^2 + \sigma_{\varepsilon}^2$ links all four classical performance metrics via exact algebraic identities:

```formula
Exact Unified Equations — Modigliani M^2, Treynor Ratio & Jensen's Alpha:
 CAPM / Factor Decomposition: R_p - r_f = alpha_J + beta_p * (R_m - r_f) + epsilon_p, sigma_p^2 = beta_p^2*sigma_m^2 + sigma_e^2
 1. Modigliani M-Squared (M^2 — Volatility-Matched Return in % / bps!):
 Leverage/Cash Weight w* = sigma_m / sigma_p (De-lever if sigma_p > sigma_m; lever at r_f if sigma_p < sigma_m)
 M^2 = r_f + (sigma_m / sigma_p) * (E[R_p] - r_f) = r_f + SR_p * sigma_m
 M^2 Alpha (in bps/yr) = M^2 - E[R_m] = (SR_p - SR_m) * sigma_m
 2. Treynor Ratio (Excess Return per Unit of Systematic Beta — Valid ONLY when beta_p > 0):
 Treynor_p = (E[R_p] - r_f) / beta_p = (E[R_m] - r_f) + (alpha_J / beta_p)
 3. Exact Link Between Sharpe & Treynor via Correlation rho_{p,m}:
 Since beta_p = rho_{p,m} * (sigma_p / sigma_m) => SR_p = (rho_{p,m} / sigma_m) * Treynor_p

Worked Pension Committee Paradox (r_f = 4.50%, Market Benchmark E[R_m] = 10.50%, sigma_m = 15.0% -> SR_m = 0.40, Treynor_m = 6.00%):
 * Fund A (Diversified Core, rho = 0.98): E[R_A] = 11.70%, sigma_A = 12.0%, beta_A = 0.784
 -> SR_A = (11.70 - 4.50)/12.0 = 0.600 | M^2 = 4.50% + 0.60*(15%) = 13.50% (+300 bps M^2 Alpha!) | Treynor_A = 7.20/0.784 = 9.18%
 * Fund B (Concentrated Biotech Specialist, rho = 0.50): E[R_B] = 14.50%, sigma_B = 20.0%, beta_B = 0.667
 -> SR_B = (14.50 - 4.50)/20.0 = 0.500 (TRAILS Fund A on Sharpe & M^2 = 12.00%!)
 -> Jensen Alpha_B = 10.00% - 0.667*(6.00%) = +6.00%/yr! | Treynor_B = 10.00 / 0.667 = 15.00% (CRUSHES Fund A!)
```

Look at why **Fund A wins for a single-fund investor** while **Fund B wins as a $10\%$ satellite sleeve inside a multi-manager pension plan**!
- If Fund B is your **entire life savings**, you bear all of its massive idiosyncratic biotech volatility ($\sigma_{\varepsilon} = 17.32\%$), so **Sharpe (`0.60` vs `0.50`) and Modigliani $M^2$ (`13.50%` vs `12.00%`)** rightly crown **Fund A**.
- Inside a **15-manager institutional portfolio**, idiosyncratic variance $\sigma_{\varepsilon}^2 / K$ diversifies away across uncorrelated managers! Only systematic $\beta_B$ adds core market variance, and Fund B delivers **`+600 bps` of Jensen's Alpha** (`Appraisal Ratio = 6.00% / 17.32% = 0.346`), lifting the **overall pension plan's Sharpe Ratio higher than Fund A can**!

![Modigliani M-Squared capital market line construction scaling portfolio volatility to match market benchmark sigma_m via SOFR cash or margin](/api/figure?type=flow&title=Decision%20Map%3A%20Standalone%20Total%20Book%20%28M2%29%20vs%20Multi-Manager%20Sleeve&subtitle=Matching%20each%20classical%20risk-adjusted%20ratio%20to%20the%20exact%20portfolio%20governance%20question.&steps=1.%20Total%20Standalone%20Book%3F%3A%3ARank%20strictly%20by%20Sharpe%20%26%20Modigliani%20M%5E2%20%3D%20r_f%20%2B%20SR_p%20*%20sigma_m%7C2.%20Satellite%20Alpha%20Sleeve%3F%3A%3ADiversifiable%20residual%20risk%20cancels%20-%3E%20Rank%20by%20Appraisal%20Ratio%20%28alpha_J%20%2F%20sigma_e%29%7C3.%20Pure%20Beta%20Replacement%3F%3A%3AReplacing%20passive%20equity%20slice%20-%3E%20Rank%20by%20Treynor%20%3D%20%28E%5BR_p%5D%20-%20r_f%29%20%2F%20beta_p%7C4.%20Audit%20Real%20Financing%3A%3AAdjust%20M%5E2%20when%20borrowing%20%3E%20r_f%20%28broker%20SOFR%20%2B%20125%20bps%20spread%20for%20w*%20%3E%201!%29&metrics=FUND%20A%20M%5E2%20ALPHA%3A%3A%2B300%20bps%2Fyr%3A%3ADe-risked%2012%25%20vol%20scaled%20to%2015%25%20mkt%20vol%7CFUND%20B%20JENSEN%20ALPHA%3A%3A%2B600%20bps%2Fyr%3A%3ASuperior%20satellite%20alpha%20generator%7CFINANCING%20SPREAD%20FIX%3A%3Ar_borrow%20%3E%20r_f%3A%3AKinks%20Capital%20Market%20Line%20above%20w*%3D1)

## Realistic Institutional $M^2$ Adjustment When Margin Borrowing Exceeds $r_f$

Textbook Modigliani $M^2$ assumes that if a low-volatility fund ($\sigma_p = 8\%$) is scaled up by $w^* = 15\% / 8\% = 1.875\times$ to match equity index volatility ($\sigma_m = 15\%$), the borrowed $0.875\times$ notional finances at the risk-free T-bill rate $r_f$. In reality, institutional prime brokers or futures clearing members charge $r_b = \text{SOFR} + s_{\text{spread}}$ (`+40 bps` in futures, `+125 bps` on prime margin)! Exact **Spread-Adjusted $M^2$** for $w^* = \sigma_m / \sigma_p > 1$ is:
$$M^2_{\text{real}} = w^* \mathbb{E}[R_p] - (w^* - 1)(r_f + s_{\text{spread}})$$
Subtracting $(w^* - 1) s_{\text{spread}}$ prevents low-volatility bond arbitrage funds from claiming unattainable levered equity equivalence!

## Governance Matrix: Modigliani M-Squared, Treynor, Jensen Alpha & Appraisal Ratio

Align risk denominators and failure boundaries across standalone total-portfolio and multi-manager sleeve decisions:

| Classical Metric | Denominator / Risk Unit | Ranking Identical to Sharpe? | Multi-Factor Extension | Fails / Breaks Down When... |
| :--- | :--- | :--- | :--- | :--- |
| **Modigliani $M^2$** | Scaled to $\sigma_m$ (`%` / `bps`) | **100% Identical!** | Downside $M^2$ (Sortino-matched) | $w^* > 1$ ignores $r_{\text{borrow}} > r_f$ spread |
| **Treynor Ratio** | Systematic $\beta_p$ | No (ignores $\sigma_{\varepsilon}$) | Generalized Treynor (arbitrage pricing) | **$\beta_p \le 0$** (market-neutral or short!) |
| **Jensen's Alpha ($\alpha_J$)** | None (raw `%` excess intercept) | No (scales with leverage!) | **Carhart 4-Factor / FF5 Alpha** | Levered $2\times$ fund doubles $\alpha_J$ artificially! |
| **Appraisal Ratio** | Idiosyncratic $\sigma_{\varepsilon}$ | Optimal weight $\propto \alpha_J / \sigma_{\varepsilon}^2$ | Multi-factor residual IR | Stale marks hide true $\beta_p$ (Scholes-Williams) |

## Multi-Factor Fama-French-Carhart Stripping of Apparent CAPM Alpha

Single-factor CAPM regressions routinely mislabel cheap Size (`SMB`) and Value (`HML`) tilts as manager skill:

![Decomposition of apparent +420 bps CAPM Jensen Alpha into Value, Size, Momentum, and Quality factor premia under Fama-French-Carhart regression](/api/figure?type=bars&title=Stripping%20Fake%20CAPM%20Jensen%20Alpha%20via%205-Factor%20Regression%20%28bps%2Fyr%29&subtitle=Auditing%20a%20%22star%22%20small-cap%20value%20fund%20claiming%20%2B420%20bps%2Fyr%20of%20single-factor%20CAPM%20Jensen%20Alpha.&steps=1-Factor%20CAPM%20Alpha%3A%3A%2B420%20bps%2Fyr%20apparent%20alpha%20vs%20S%26P%20500%20beta%20%3D%201.08%7CLess%20Size%20%28SMB%29%20Tilt%3A%3A-135%20bps%2Fyr%20explained%20by%20passive%20small-cap%20exposure%7CLess%20Value%20%28HML%29%20Tilt%3A%3A-160%20bps%2Fyr%20explained%20by%20cheap%20book-to-market%20tilt%7CTrue%20Net%205-Factor%20Alpha%3A%3A%2B45%20bps%2Fyr%20true%20idiosyncratic%20skill%20after%20Quality%2FMomentum%20%26%20fees!&metrics=CAPM%20ILLUSION%3A%3A%2B420%20bps%2Fyr%3A%3AOmitted%20factor%20beta%20masquerading%20as%20skill%7CTRUE%20FF5%20ALPHA%3A%3A%2B45%20bps%2Fyr%3A%3AStatistically%20indistinguishable%20from%200%7CDIMSON%20LAGGED%20BETA%3A%3ASum%28beta_%7B-1..%2B1%7D%29%3A%3AFixes%20illiquid%20small-cap%20beta%20understatement%7CTREYNOR%20SIGN%20GATE%3A%3ARequire%20beta%20%3E%200.2%3A%3ANever%20compute%20Treynor%20near%20beta%20%3D%200!)

## Two Critical Traps: Dimson Lagged Beta & Multi-Factor Alpha Stripping

1. **Illiquid Assets Understate Contemporary $\beta_p$**: Private credit, micro-cap stocks, and high-yield bonds react to S&P 500 moves with a $1\text{--}3$ month lag. Regressing $R_{p,t}$ only on contemporaneous $R_{m,t}$ artificially shrinks $\hat{\beta}_p$ from $0.75$ down to $0.25$—fabricating $+300\text{ bps}$ of fake Jensen's Alpha! Always sum Dimson (1979) lagged betas $\beta_{\text{Dimson}} = \beta_0 + \beta_1 + \beta_2$ alongside [Lo (2002) autocorrelation adjustments](/guides/sharpe-ratio-iid-annualization-lo-2002-autocorrelation-haircut).
2. **Upgrade Single-Index CAPM $\alpha_J$ to Fama-French 5-Factor + Momentum**: Ensure managers are not charging active fees for cheap ETF factor betas; cross-reference [Information Ratio and Tracking Error](/guides/information-ratio-active-return-tracking-error-fundamental-law), [Half-Kelly leverage sizing](/entities/half-kelly-leverage-continuous-fractional-half-kelly-portfolio-leverage-criterion), and institutional broker margin schedules in [broker lookup](/broker-lookup/axi-trading).

## Frequently asked questions

### Can two funds ever rank differently under Modigliani M-Squared (M2) versus the Sharpe Ratio?

No, provided excess returns are evaluated against the same risk-free rate r_f and single-borrowing-rate assumption! Because M^2 = r_f + SR_p * sigma_m where r_f and sigma_m (> 0) are identical positive constants for every fund being compared, M^2 is a strictly increasing linear transformation of SR_p.

### Why shouldn’t I just pick the fund with the highest raw Jensen’s Alpha?

Because raw Jensen’s Alpha has no risk denominator! If Fund X has +1.5% alpha with 2% residual vol (Appraisal Ratio = 0.75), borrowing on margin at 2x leverage turns Fund X into +3.0% alpha with 4% residual vol—doubling Jensen’s Alpha via pure leverage without adding a single basis point of true skill!

### How do I calculate Treynor-Black optimal capital allocation weight for an active satellite fund?

Treynor & Black (1973) proved that the optimal weight allocated to active portfolio A relative to passive market index M is proportional to (alpha_A / sigma_epsilon_A^2) divided by (E[R_M - r_f] / sigma_M^2)—literally the active Appraisal Ratio divided by residual vol relative to market price of risk!

### What is "Downside M-Squared" (or Sortino M2)?

Analogous to standard M^2, Sortino-adjusted M^2 scales portfolio leverage to match the benchmark’s Target Downside Deviation TDD_m rather than total volatility sigma_m: M^2_Sortino = tau + Sortino_p * TDD_m, expressing convex downside-adjusted outperformance in annual percentage points.


## Source URLs

- https://jpm.pm-research.com/content/23/2/45
- https://www.jstor.org/stable/2325486
- https://www.jstor.org/stable/2978634
- https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2025/portfolio-performance-evaluation

## Citation

- APA: Elena Rostova, FRM, CQF. (2026). Treynor Ratio, Jensen’s Alpha & Modigliani M-Squared (M2) Leverage Math. QuantCalc Tools. https://quantcalctools.com/guides/treynor-ratio-jensens-alpha-m-squared-modigliani-leverage-math
- Canonical article: [Treynor Ratio, Jensen’s Alpha & Modigliani M-Squared (M2) Leverage Math](https://quantcalctools.com/guides/treynor-ratio-jensens-alpha-m-squared-modigliani-leverage-math)
- Site research index: [QuantCalc Tools](https://quantcalctools.com/llms-full.txt)
