As a folklorist and scholar of oral narrative, the answer is, unfortunately, only as far back as the written record exists. It's one of the worst parts of being a narrative scholar. We never have access to the source narratives from which our written records are derived, as the narrators and performers are dead. We have living Performers, of course, and if they have a family tradition, we can go back several generations. Then we could link them to other regional similarities, either by another generational line or by written record. But this is difficult work, and you're not going to find a (good) narrative scholar claiming they found the "original" version of a tale. We can say that there are records of, say, a Cinderella story from Egypt in the first century, but the categorization as a "Cinderella story" is a much later application and can be very loosely applied. The odds are also that our record stating the story is from the first century was written in the eighth century, so we can't really be sure.

There was a Guardian article a couple of years ago that claimed something like "fairy tales extend back to the Bronze Age." Maaaaayyyybbbeeee. But you have to play fast and loose with your definition of fairy tale and your criteria to make a claim like that. The nature of oral narrative is that it's fluid, changing, and performative. Even if we have a written record, we're not going to understand the story as it was delivered. So "tracing the origins" of a story is not really something a narrative scholar will do. Instead, we might look for networks of inspiration and influence and track the changes over time and place, but for oral narrative we won't claim we've found the "original." Sometimes we can do this with literary fairy tales, but even those are often clearly derived from oral narrative and then you're in the same position.

Some scholars to look at on this topic are Jack Zipes, Alan Dundes, Holbek Bengt, and Linda Dégh.