A New Test Upends a Long-Held Assumption About Dark Matter

A new study puts very different dark matter theories through the same quantitative test of “naturalness.”

Some of the most familiar assumptions about dark matter may depend less on the type of dark matter being proposed than on the details of how a particular model produces it.

A new analysis by Stefano Profumo, professor of physics at the University of California, Santa Cruz, compares a dozen leading dark matter scenarios using the same mathematical test of fine-tuning. The study finds that particle dark matter and primordial black holes both include models that are relatively natural and others that require parameters to be adjusted with considerable precision. The results appear in Physical Review D.

“There’s a habit of treating primordial black holes as the exotic, fine-tuned alternative, and particle dark matter as the safe, natural default,” said Profumo, deputy director for theory at the Santa Cruz Institute for Particle Physics. “When you actually run the numbers side by side, that story doesn’t hold up. Some black hole scenarios are about as natural as it gets. Some particle scenarios are wildly fine-tuned. And some of each land right in the middle.”

Putting dark matter theories on the same scale

Dark matter accounts for most of the matter in the universe and plays a major role in the formation and motion of galaxies, yet it has never been directly detected. Physicists have proposed many possible explanations, ranging from undiscovered subatomic particles to primordial black holes that could have formed during the universe’s earliest moments.

Without a confirmed detection, researchers often judge competing theories partly by their “naturalness.” In physics, a natural model produces the observed universe without requiring several underlying parameters to take extremely precise values. If a tiny change to one input causes the prediction to change dramatically, the model is considered more fine-tuned.

Naturalness has influenced theoretical physics for decades, but comparisons are often made informally or applied differently across classes of models. Profumo’s study instead uses a single quantitative method to compare very different dark matter proposals on equal terms.

Measuring how sensitive each model is

The analysis uses the Barbieri-Giudice measure, which tracks how strongly a model’s predicted dark matter abundance changes when one of its underlying parameters is altered slightly. A model whose prediction changes only modestly is relatively insensitive to its assumptions. A model whose output shifts sharply requires more precise tuning.

Profumo applied the measure to 12 established dark matter scenarios. These included several particle candidates, such as the long-studied WIMP, or weakly interacting massive particle, as well as multiple mechanisms that could have produced primordial black holes soon after the Big Bang.

The resulting range does not divide neatly into particle models on one side and black hole models on the other. Instead, models from both categories appear across the spectrum of fine-tuning. The comparison suggests that naturalness depends strongly on the specific production mechanism rather than simply on whether the proposed dark matter consists of particles or black holes.

Some familiar models require unexpectedly precise tuning

One primordial black hole scenario performed especially well. In that model, networks of structures known as domain walls collapse to form black holes. It ranked among the most natural scenarios in the analysis and was comparable to some of the least fine-tuned particle models.

A widely studied particle scenario landed near the opposite end of the scale. In that model, dark matter particles annihilate through a resonance associated with the Higgs boson. Producing the required amount of dark matter demands that one parameter be set to within a fraction of a percent, making it one of the most fine-tuned cases examined.

Other particle and primordial black hole models occupied intermediate positions. The result argues against assigning a single naturalness label to an entire category of dark matter candidates.

A common test, not a winner

“Naturalness has real power as a filter for deciding where to look next,” Profumo said. “But it can’t be a shortcut for dismissing an entire category of ideas, like primordial black holes, just because a few individual models within that category happen to be tuned. The tuning lives in the specific model, not in the kind of dark matter you started with.”

The study does not identify dark matter or prove that less fine-tuned models are correct. However, offers a consistent way to compare competing theories and suggests that naturalness is most useful when applied to individual models, not entire categories.

Reference: “Primordial black hole dark matter: A quantitative parameter sensitivity comparison across formation mechanisms and particle candidates” by Stefano Profumo, 10 September 2026, Physical Review D.
DOI: 10.1103/nk1q-5k51

Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.