I guess most people don't have access to this and those that do are seeing the same text as you. Apart from suggesting that you contact the authors, my only thought is that the meaning of standardization is itself not rigidly standardized. If something had approximately zero mean, especially long term, then just dividing by the sd would seem to me an acceptable flavour of standardisation.
Also note that "return" rings loud bells with perhaps 30% of Statalist and means "go back" to everyone else.
Nick
[email protected]
Rajesh Tharyan
What do they really mean by standardization?
This is from (page 318, paragraph 3 and 4) the following paper,
Mitchell, M.L., and E. Stafford. (2000). Managerial decisions and long-term stock price performance. Journal of Business, 73, no. 3, pp. 287-329.
The authors calculate the abnormal return (CTAR) on a portfolio of stocks every calendar month as CTAR = Rpt- E(Rpt), Rpt is the actual realised return, E(Rpt) is the expected return.
They then say that, "the monthly CTARs are standardized by estimates of the portfolio standard deviation.." They then go on to say that "each standardized monthly CTAR should have a mean of zero and be independent. Therefore, statistical inference is based on the time series mean of the monthly standardised CTARs and the standard error of the mean."
I am not sure if they mean Z standardization or
They simply divide by the standard deviation, Would it be right to call that standardization?
The CTAR is an abnormal return so the null is that it is zero.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
