Choose the Least Survivorship-Biased Fund Panel

A researcher wants the 2016-2025 average return of all funds launched in 2016. Which data pull is least exposed to survivorship bias?

A. Keep only funds that are still alive in 2025, then use their full back history. B. Keep funds alive on each evaluation date, but retroactively delete any fund once it later closes. C. Use a point-in-time panel that keeps every 2016 launch in history until its actual closure date and retains closed funds in the archive. D. Use only funds that are current index constituents in 2025.

Give the best choice.