Pulsar timing arrays have detected the universe humming, and the hum is louder than the standard accounting predicts. In June 2023 NANOGrav's 15-year dataset, with parallel results from EPTA, PPTA, and CPTA, presented evidence for a stochastic nanohertz gravitational-wave background with the Hellings-Downs angular correlation signature, at a strain amplitude of 2.4 (+0.7/-0.6) x 10^-15 at the reference frequency of one inverse year (arXiv:2306.16213). The default source is the cosmic population of supermassive black hole binaries spiraling toward merger. The catch is amplitude: standard binary-population models built from observed galaxy merger rates and black hole scaling relations predict a signal near 1 x 10^-15, a factor of two to several below the measurement, even under optimistic assumptions where every merger promptly forms an efficiently hardening binary (arXiv:2306.16220).
To meet the data, ΛCDM's binary models must push multiple parameters to their edges simultaneously: black hole masses systematically above the local scaling relations, merger rates at the top of allowed ranges, and binaries that traverse the final parsec efficiently, the very step long considered the population's bottleneck. The spectral slope adds its own mild strain against the canonical -2/3 binary prediction. Alternatives, cosmic strings, phase transitions, relic inflationary waves, fit the amplitude but import new physics wholesale.
The standing is young and sharpening fast: the detection is secure, the source attribution is not, and the amplitude excess sits unresolved while datasets lengthen. Individual-binary detections, anisotropy mapping, and the spectral shape over the coming few years will separate an overachieving binary population from something the standard inventory lacks.