FBX exoprt strange values on specularity

hi guys,

Does anybody know why the FBX exporter halves the specular intensity and divides by 5 the hardness ?
Is that how FBX expects it, or is it just an empirical formula ?