From relativity we have m = M sqrt(1 - (v/c)^2); where m is invariant mass, M is its inertia, and v its speed relative to a static frame K. We can multiply both sides by c^2 and define e = mc^2 and E = Mc^2.
Then by rearranging terms we find that E^2 = e^2 + k^2; where k = Mvc is the kinetic energy of the mass that has M inertia. Now the fun part.
We set v = c, making the mass m go the speed of light. Impossible you say...you're right, but continue on and see what must happen to make it possible.
We now have k = Mvc = Mcc = Mc^2 = E. In other words the total energy E is now all...entirely...kinetic energy.
Which leaves us with E^2 = e^2 + k^2 = e^2 + E^2 or, ta da, E^2 - E^2 = e^2 = 0. And e = 0 = mc^2. But as c is never zero, that means, drum roll please, m must = 0. And that makes it possible. The object cannot have invariant mass to go the speed of light, like, duh, light does.
But also note, it has energy, E = Mc^2 which is all kinetic. And look here, Mc = P is momentum. And look even further E = Mc^2 = Pc = hF so that P = hF/c where E = hF is the energy of a photon based on its frequency F.
And there you are. The theory of relativity says that photons must have momentum. And in fact, we can measure that momentum in labs; so it's not just a hypothesis, it's an engineering observation.