rolling_event_study: Rolling Event Study
In moneyhun/eventful: A Simple Pacakge for Rolling-Window Single-Firm Event Studies

Description Usage Arguments Value Author(s)

View source: R/rolling_event_study.R

Rolling Event Study

rolling_event_study(
  formula,
  data,
  pred_date_range,
  exclude_dates = NULL,
  id_dates = NULL,
  misrep_dates = NULL,
  disc_dates = NULL,
  rolling_window = NULL,
  roll_fixed = NULL,
  orth = NULL,
  p.val_thresh = 0.05,
  simple = FALSE
)

`formula`	A formula object describing the regression to be run. If dummy dates are specified in 'id_dates', they will be added to this formula.
`data`	A data.frame object containing the data for the regression and a 'date' variable.
`pred_date_range`	A vector of the min and max date for the prediction range.
`exclude_dates`	A vector of dates which should be excluded from the fitted regression models.
`id_dates`	A vector of dates which should be 'dummied out' via indicator variables.
`misrep_dates`	A vector of dates which should be flagged as misrepresentation dates in the output. Does not affect the regression model.
`disc_dates`	A vector of dates which should be flagged as disclosure dates in the output. Does not affect the regression models.
`rolling_window`	A number specifying the width of the rolling window. E.g., a window of 50 means that the regression used to predict returns on day i will be fitted on data spanning from day i-50 to day i-1.
`roll_fixed`	Either 'observations' or 'window'. Where excluded dates fall within the rolling window for a given date, this specifies whether the number of observations should remain fixed–meaning the window must be extended–or whether the window should remain fixed–meaning the number of observations will be lower.
`orth`	A formula object describing the orthogonalziation which should be applied (e.g., for y ~ x1 + x2, y in the final regression would be replaced by the residual from this regression.)
`p.val_thresh`	Threshold for flagging residuals as significant. 5% by default.
`simple`	TRUE/FALSE indicating whether full output or just 'table' should be returned. FALSE by default.

If simple is FALSE, an list containing a table of event study results, which the regressions themselves as a column, and many other variables of interest relating to the event study.

Levi Moneyhun

moneyhun/eventful documentation built on Jan. 3, 2022, 12:49 p.m.