The potential issues with cfacshr and cfacpr are discussed in #152 and #95. I'm closing those issues because we've checked and fixed all predictors except for the 13F predictors, which we will leave for another time (potentially the 2025 release).
There may not be any lookahead bias here because of how the 13F code rescales everything in the end. But we should check.