メモ用です．

pandasで日付・時間の列を処理（文字列変換、年月日抽出など）

import datetime
from dateutil.relativedelta import relativedelta

# 1期後を作る
df["period1"] = df["決算日"]
df_target = df[["ティッカー","period1","時価総額"]].copy()
df_target.columns = ["ティッカー","period1","period1ago"]
df_target["period1"] = df_target["period1"].dt.to_pydatetime()+ relativedelta(months=3)
df = pd.merge(df,df_target,how="inner",on=("ティッカー","period1"))
df