Before our universe existed, another one may have collapsed in on itself, and some of its black holes may have survived the crunch. That is the central claim of a theoretical paper posted to arXiv in February 2026, which argues that dark matter, the invisible substance making up roughly 27 percent of the cosmos, could be composed of black holes inherited from a prior cosmic cycle. The idea is striking, but it faces a gauntlet of observational constraints that already limit how much of dark matter any population of black holes can account for.
The bounce hypothesis and its predictions
The paper, titled “Cosmological Bounce Relics: Black Holes, Gravitational Waves, and Dark Matter,” builds on a class of models known as bouncing cosmologies. Instead of treating the Big Bang as an absolute beginning, these models propose that a previous universe contracted to an extraordinarily dense state and then rebounded, seeding the expansion we observe today. The authors describe two pathways by which black holes could emerge from this process: some survive intact through the bounce itself, while others form shortly afterward when extreme density fluctuations collapse under their own gravity.
Crucially, the paper predicts that these relic black holes would generate distinctive gravitational wave signatures, ripples in spacetime that future detectors could search for. If those signals were found at the predicted frequencies and amplitudes, it would constitute the first observational evidence linking any object in our universe to a preceding cosmic epoch.
The bounce concept draws on ideas from quantum gravity, a still-developing framework that attempts to merge general relativity with quantum mechanics. A related line of research explores so-called Planck Star Remnants, ultra-compact objects that quantum effects prevent from collapsing into true singularities. A separate arXiv preprint develops formal stability criteria and abundance estimates for these remnants, asking whether enough of them could form to account for the universe’s missing mass. Both papers remain unreviewed preprints as of May 2026.
What observations already tell us
The idea that black holes might explain dark matter is not new. Bernard Carr and Stephen Hawking laid the theoretical groundwork in a landmark 1974 paper published in Nature, and the concept surged back into prominence after LIGO’s first gravitational wave detection in 2015, when Bird and collaborators proposed that the merging black holes could be primordial in origin. NASA has highlighted the hypothesis as well, noting that primordial black holes gobbling up surrounding matter could produce detectable X-ray emissions and patterns in cosmic background light.
But decades of searching have progressively narrowed the window. The OGLE (Optical Gravitational Lensing Experiment) collaboration published results in Nature using microlensing observations of millions of stars in the Milky Way. Microlensing works by watching for the telltale brightening that occurs when a massive, invisible object drifts between a distant star and Earth, bending the star’s light like a cosmic magnifying glass. OGLE’s data set 95 percent upper limits on compact objects across a broad mass range. A Nature News and Views commentary summarized the implication bluntly: primordial black holes appear too scarce to explain dark matter at the masses tested.
Gravitational wave observatories add a second layer of restriction. A recent analysis modeled how Planck-scale relics and similar compact objects would contribute to a diffuse background hum of gravitational waves, then compared those predictions against LIGO and Virgo sensitivity data. The conclusion: if these objects formed through standard statistical processes (physicists call them Gaussian initial conditions, meaning the primordial density fluctuations followed a simple bell-curve distribution), their mergers would already have produced a detectable signal. No such signal has appeared. Only formation pathways involving rarer, more exotic fluctuation patterns remain viable, and even those are tightly bounded.
The gaps that remain
The bounce-relic hypothesis occupies a narrow space that observations have not yet sealed shut. OGLE’s microlensing survey is powerful, but it probes specific mass ranges. If relic black holes cluster at masses just above or below the most constrained windows, or if they are distributed more smoothly than a typical dark matter halo population, they could dodge the strongest limits while still contributing a fraction of the total dark matter budget.
Whether exotic, non-Gaussian formation channels can produce enough relic black holes to matter is an open question. Generating the kind of extreme density spikes required typically demands unusual physics, and those same spikes can leave fingerprints in the cosmic microwave background or the large-scale arrangement of galaxies, both of which are themselves well measured. The surviving parameter space is real but small.
No direct observational evidence currently ties any detected black hole to a pre-bounce universe. The gravitational wave signatures predicted by the bounce-relic paper have not been observed, and the instruments best suited to test them are still under development. The European Space Agency’s LISA mission, a space-based gravitational wave observatory, is targeting a launch in the mid-2030s. Third-generation ground-based detectors like the Einstein Telescope and Cosmic Explorer remain in the planning and early construction phases. Until those facilities come online, the hypothesis stays in the realm of mathematical prediction.
There is also an unresolved relationship between Planck Star Remnants and the broader class of bounce relics. Both invoke quantum gravity effects, but they differ in mass scale, formation timing, and the signatures they would leave behind. Whether they represent two faces of the same physics or genuinely distinct phenomena is something theorists have not yet sorted out.
Where this fits among dark matter candidates
Black holes of any origin are just one entry on a long list of dark matter candidates. The two most widely studied alternatives are WIMPs (weakly interacting massive particles) and axions, both hypothetical particles that would interact with ordinary matter so feebly that they have eluded every detector built to find them so far. Decades of underground experiments and collider searches have failed to produce a confirmed WIMP detection, and axion searches, while increasingly sensitive, have likewise come up empty. That persistent null result is part of what keeps more exotic ideas, including primordial and relic black holes, alive in the theoretical conversation.
The bounce-relic proposal adds a cosmological dimension that particle candidates lack: it ties the dark matter question to the origin and possible cyclical nature of the universe itself. If dark matter turned out to be debris from a previous cosmos, it would not just identify the substance but reshape our understanding of what came before the Big Bang, a question most of modern cosmology treats as unanswerable.
What would settle the question
The path forward is observational. Microlensing surveys will continue to tighten constraints across additional mass ranges. LIGO, Virgo, and the forthcoming KAGRA detector will accumulate more sensitive gravitational wave data with each observing run, progressively squeezing or revealing the stochastic background signal that relic populations would produce. When LISA and the next generation of ground-based observatories begin operating, they will probe frequency bands and sensitivities that current instruments cannot reach, directly testing the spectral predictions laid out in the bounce-relic paper.
For now, the idea that dark matter could be black holes carried over from a universe before ours is a carefully constructed “what if” rather than a confirmed discovery. It generates specific, falsifiable predictions, which is exactly what separates productive speculation from idle conjecture. The observational tools to deliver a verdict are being built. The universe, as usual, will have the final word.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
