The Penrose Inequality Beyond Time Symmetry

The Penrose inequality links the mass measured at infinity to the area enclosing a black hole on an initial slice. In time symmetry it is a theorem; for general extrinsic curvature, the correct horizon prescription and the deformation needed for a proof remain unsettled. As of 2026-08-20, the unrestricted non-time-symmetric problem is open, with strong supporting special cases and high confidence about the boundary between proved and conjectural statements.

Helpful background. The Riemannian Penrose Inequality develops the established $K=0$ theorem, inverse-mean-curvature and conformal-flow ideas, and rigidity. This record begins where nonzero extrinsic curvature changes the problem.

Which mass–area inequality remains open?

Let $(M,g,K)$ be complete asymptotically flat initial data satisfying the dominant energy condition. In four dimensions, a schematic target is

$$G M_{\mathrm{ADM}} \geq \sqrt{\frac{A}{16\pi}},$$

where $A$ must be the area of the physically correct outer enclosure and equality should select the corresponding stationary geometry. The formula alone is not a conjecture until one specifies whether the surface is minimal, marginally outer trapped, generalized apparent, or an outer minimizing enclosure; whether components are connected; and whether charge or angular momentum is fixed.

Time symmetry sets $K=0$, turns the constraint into nonnegative scalar curvature, and makes an outermost apparent horizon a minimal surface. General data lack that reduction. The finite question here is whether an unconditional inequality and rigidity theorem holds for $K\ne0$ with a horizon prescription that survives known counterexamples.

Assessment at 2026-08-20	Meaning
Question status	Open
Evidence status	Supported
Confidence	High that the general-data problem remains open and that naive surface choices fail
Domain	Complete asymptotically flat initial data satisfying the dominant energy condition; spacetime and null-infinity variants are excluded
What could change the conclusion	A general proof, rigidity theorem, or valid counterexample outside current special cases

A starting project: test the chosen horizon area

Fit and preparation. Choose this route if you can compute null expansions of a surface in initial data, distinguish a minimal surface from a marginally outer trapped surface, and evaluate ADM mass. It is accessible through explicit spherical geometry before the elliptic-analysis frontier. Symbolic algebra can check derivatives, but a geometric-analysis adviser should check the surface prescription and energy condition.

Reading and baseline. First verify the Schwarzschild equality case using the linked Riemannian treatment. Next reconstruct the spherical apparent-horizon counterexample in Ben-Dov 2004, arXiv v2, §4 (PDF): identify the asymptotic end, matched regions, selected slice, and every surface whose area enters the argument. The purpose is to recover the failure of that particular area prescription. A horizon label alone is insufficient; compute the relevant expansion and state which enclosing surfaces were allowed.

Deliverable. Produce a calculation comparing ADM mass and the candidate areas in the same geometry, with the matching and dominant-energy assumptions visible. Include a check that the time-symmetric Schwarzschild limit returns equality. This is reproduction of an existing counterexample, and remains useful even if no new inequality results.

Possible extension and stopping point. After a literature search covering later counterexamples and rigidity results, ask whether a specifically defined enclosure continues to work under a small, controlled change of the spherical slice. This is a candidate formulation test, not an assertion that the general-data conjecture is new or solved. Stop if the surface is selected using information unavailable on the initial slice, or if a proposed deformation assumes the unresolved Jang-equation existence theorem. A project on that existence problem needs sustained supervision in geometric PDE; it is not the next algebraic step after the spherical example.

What is established in the stated domain

For three-dimensional time-symmetric data, Huisken and Ilmanen prove the connected-horizon case using weak inverse mean curvature flow in the main theorem and flow construction, pp. 353–437. Bray’s independent conformal-flow proof treats the full outermost minimal boundary, including disconnected components in the main theorem and monotonicity argument, pp. 177–267.

Bray and Lee extend the Riemannian theorem through dimensions three to seven Theorem 1.3 and the rigidity discussion. Bi and Zhu now claim the arbitrary-dimensional Riemannian result, allowing the expected singular set of dimension at most $n-8$ Theorem 1.1, printed pp. 2–3, with proof in §§2–5. That May 2026 work was a v1 preprint at the cutoff. Its dimensional advance remains entirely within $K=0$.

For $K\ne0$, Bray and Khuri formulate a generalized Jang-equation route and prove the spherically symmetric cell, while leaving the nonspherical coupled system conditional Bray and Khuri 2010, §§1–3. The unresolved step is not an algebraic estimate: it is existence and regularity for the coupled deformation with the required boundary behavior.

Results and open questions at 2026-08-20

Claim or evidence	What it supports	What it does not establish
Huisken–Ilmanen weak flow	Three-dimensional Riemannian inequality for its horizon cell	General $K\ne0$ data
Bray conformal flow	Independent Riemannian proof and disconnected outer boundary	A spacetime deformation
Bray–Lee and Bi–Zhu	Higher- and all-dimensional Riemannian reach	Non-time-symmetric inequality
Generalized Jang programme	A coherent route and restricted cases for general data	Global solvability of the coupled system
Ben-Dov and Carrasco–Mars counterexamples	Convenient apparent-horizon prescriptions can be false	Refutation of every enclosing-area formulation

The positive evidence is structural: two independent three-dimensional proofs and higher-dimensional extensions support the mass–area principle in the Riemannian cell. The negative evidence has equal methodological weight because it eliminates formulations, not merely proof attempts.

Competing formulations of the horizon area

One position seeks a quasi-local surface visible directly on the slice, such as an apparent or generalized apparent horizon. It is attractive because it resembles the time-symmetric boundary and can be handled elliptically. Its weakness is decisive: such areas can be too large relative to ADM mass on allowed slices.

The more conservative position uses an outer enclosure related to the future event horizon or a deformation preserving the physically relevant area. This better matches the heuristic derivation from area increase and final-state settling, but it is harder to formulate without importing global evolution into an initial-data inequality. The live mathematical disagreement concerns that enclosure and the solvability of its deformation, not whether the $K=0$ theorem exists.

Counterexamples, null results, and failed routes

Ben-Dov constructs a spherical counterexample to a naive apparent-horizon-area inequality in the construction and conclusion. Carrasco and Mars similarly disprove a generalized-apparent-horizon proposal in the main counterexample theorem. These results force a better enclosure; they do not refute every spacetime Penrose inequality.

The generalized Jang route remains conditional beyond its established cells because blow-up, boundary conditions, and coupling to a monotone flow have not been controlled generally. Treating the desired solution of that PDE system as available would assume the central difficulty. Likewise, calling Bi–Zhu a solution to “the Penrose inequality in all dimensions” without the word Riemannian would incorrectly move across the $K=0$ boundary.

What would change the assessment

The assessment would change after an unconditional existence and regularity theorem for a deformation that handles general $K$, reduces to the established Riemannian result, treats disconnected horizons, and carries a rigidity statement. A carefully constructed counterexample to the best outer-enclosure formulation would instead weaken the physical conjecture itself. Either result must be tested against the Ben-Dov and Carrasco–Mars geometries.

Research checklist

• State dimension, asymptotic end, dominant-energy condition, and regularity of $(g,K)$.
• Define the selected surface and explain outermost or outer-minimizing conditions; do not substitute “apparent horizon” generically.
• Separate the inequality from equality rigidity and from charged or angular-momentum refinements.
• Verify that a proposed general proof reduces to the established $K=0$ theorem with the same mass normalization.
• Stress-test the surface prescription against both published counterexamples.
• For a Jang-based route, expose existence, blow-up, boundary, and singular-set hypotheses rather than assuming the deformation.

Source and data map

Huisken–Ilmanen and Bray provide independent foundational proofs; Bray–Lee and Bi–Zhu control the dimension axis; Bray–Khuri supplies the general-data programme; Ben-Dov and Carrasco–Mars supply adverse formulation tests. Null-infinity and asymptotically hyperbolic inequalities were excluded to keep the proposition finite.

No independent reproduction was performed, and no executable or machine-checked proof archive was verified. Bi–Zhu v1 had no journal reference or independent proof review at the cutoff. Reproduction means checking its singular-hypersurface inputs and rigidity argument, and checking every general-data claim against the exact Jang-equation hypotheses.

References

• Ben-Dov, Ishai. “Penrose Inequality and Apparent Horizons.” Physical Review D 70 (2004), 124031. DOI. Open PDF.
• Bi, Yuchen, and Jintian Zhu. “Riemannian Penrose Inequality in All Dimensions.” arXiv:2605.00680v1 (2026). arXiv.
• Bray, Hubert L. “Proof of the Riemannian Penrose Inequality Using the Positive Mass Theorem.” Journal of Differential Geometry 59 (2001), 177–267. Journal record.
• Bray, Hubert L., and Marcus A. Khuri. “A Jang Equation Approach to the Penrose Inequality.” Discrete and Continuous Dynamical Systems 27 (2010), 741–766. DOI.
• Bray, Hubert L., and Dan A. Lee. “On the Riemannian Penrose Inequality in Dimensions Less Than Eight.” Duke Mathematical Journal 148 (2009), 81–106. DOI.
• Carrasco, Alberto, and Marc Mars. “A Counterexample to a Recent Version of the Penrose Conjecture.” Classical and Quantum Gravity 27 (2010), 062001. DOI.
• Huisken, Gerhard, and Tom Ilmanen. “The Inverse Mean Curvature Flow and the Riemannian Penrose Inequality.” Journal of Differential Geometry 59 (2001), 353–437. Journal record.