The decline in global inequality over the last decades has spurred a ‘sunshine’ narrative of falling globalinequality that has been rather oversold, in the sense, we argue, it is likely to be temporary. Our work firstformalizes the intuition that the fall in global inequality will eventually reverse. We derive the location ofthe turning point for a specific measure of inequality: the mean log deviation. We make use of a custombuiltdatabase of global income to estimate this turning point. We find there is a potentially startling globalinequality ‘boomerang’, possibly in the mid-to-late 2020s, which would have happened even if there were nopandemic, and that the pandemic is likely to bring forward the global inequality boomerang. The scholarlysignificance of the main finding is that there is a new type of Kuznets curve, where inequality first falls andthen rises as middle-income countries grow fast and approach the income levels of rich countries. The policysignificance is that interventions to counteract the upward movement in global inequality will require evenstronger focus on lowering the within-country inequality component of global inequality.