Elementary particles are considered point-like. Which makes the concept of spin rather problematic. It can neatly be described by group theoretical considerations, but that offers no explanation what it actually is.
But what if we consider particles to be extended particles, their extension reaching into small extra spatial dimensions? For example, in the one-dimensional case, that the particle is a small circle structure around a thin Planck-sized cylinder, which would have the nice property of a Lorenz invariant Planck length, as the extension is perpendicular to the 1d bulk space. This could be extended to 3d if we consider three small curled up dimension (each to a Planck length circle). In 1d, the circle could rotate in two opposite directions, and spin would correspond to actual rotation.