发帖
雷达卡
现在要构筑一个变量VT，数据里的变量有：公司名，事件日，return，valued-weighted return，window(事件日前3天到后3天)
ST=return-valued-weighted return
描述材料如下：As in the below materials, VT is a Z-transformation of rank scores calculated for each calendar year and each firm based on the “ST”(daily absolute value of returns minus value-weighted returns) (Blom 1958).
Rank scores are based on methodology that ranks standardized absolute errors over the combined estimation and event periods. Ranks are then converted to a standard normal Z-statistic based on Blom [1958].
Easton and Zmijewski [1993] used the following methodology: standardized squared errors over the estimation and event periods were estimated for each firm-event. The errors were ranked over the combined estimation and event periods for each firm-event. These ranks were converted to a standard normal z-statistic and aggregated by relative event dates to form a standard normal z-statistic. We used the same approach for the standardized absolute errors.
我试了把每个公司每个事件日的ST排序，序号为n，再设VT={n-mean(n)}/sd(n)，做出来的结果似乎不正确，请高人帮忙看看我的问题在哪儿呢，想了好几天了都没找出问题所在，
谢谢指点
京公网安备 11010802022788号
