Finnish has a number ancient loans from Common Germanic and early Norse. Those that correspond to the Greek type with the nominative singular masculine ending in
-ος may sometimes terminate in
-as/-äs (which in Finnish and its sister languages is not an ending, but part of the stem), for example:
kuningas, valas, ansas, parras, mallas, keihäs, rengas, rikas, varas, armas, autuas, hurskas, kernas, sairas, viisas, kallas (more often, however, a pure stem in
a was borrowed: either abstracted from the Germanic declension or reflecting the accusative singular after the nasalization
*-an>*-ã>-a). Also there is
lammas borrowed from an
-os/es-stem corresponding to the Greek
γένος-type.
In Germanic, this ending is directly attested in the earliest (before the 6th century, when massive syncope began) Norse runic inscriptions, for example:
erilaz, raunijaz, Frawaradaz, þewaz, Wiwaz, harabanaz, stainaz, Erafaz, laukaz, Leugaz, leubaz, haitinaz. This
z later became
r: from the first attestations in West Germanic (where it mostly disappears word-finally, surviving in a few cases like the German
wer), and in the middle 1st millennium in North Germanic, where it originally changed into a special kind of
r transliterated as
R (perhaps guttural or palatal), but several centuries later it, too, either disappeared word-finally or merged with the plain
r. In Gothic
-az>-s in most cases, but it survived in
ƕazuh.
In Common Germanic
*o>a, and the final -
s originally split into
-s and
-z, depending on the position of the stress (
Verner's law:
-s remained after a stressed vowel and became voiced after an unstressed one), but later
-z was generalized, so that the inherited
*-os uniformly became
-az. The word
kuningas from your question thus originates from late Common Germanic
*kuningaz of the last centuries BC or the fist centuries CE, borrowed probably from the speech of Scandinavians, not from later German of course.