High Order Cumulant and Moment Program

    We have a Stata module that will execute either the high-order cumulant (preferred) or moment (not preferred) estimators. You can read about the new high-order cumulant estimators here:

    Erickson, Jiang, Whited (2014).

    The cumulant estimators are asymptotically equivalent to the moment estimators, but they have closed form solutions, so there are no computational difficulties.

    The Stata module is extremely easy to use. It is just a one-line command. You can find the code here:


    Old High Order Cumulant Program

    For those of you who prefer Matlab to Stata, here is the old Matlab cumulant code. It unfortunately has limited functionality. For example, it only handles one mismeasured regressor. Also, it won't calculate clustered standard errors, so you would have to code that in yourself.


    High Order Moment Programs

    This section of the page contains the original versions of the programs that Tim Erickson and I used in our series of papers on estimating the errors-in-variables model using high-order moments. (JPE, 2000; Econometric Theory, 2002; Review of Financial Studies, 2012.)

    Here is a tutorial that explains how the estimators work.

    Here is the old code. These programs are no longer being updated, but here they are anway.



    One Mismeasured Regressor Program in SAS

    Identification Test Program in SAS

    Coefficient Signing Programs

    This page contains user friendly versions of the program that Tim Erickson and I used in our paper on signing regression coefficients in the errors-in-variables model. (Finance Research Letters, 2005.)

    The program was originally written in GAUSS, but it has been translated into Matlab, Stata's MATA language, and SAS. They are really easy to use. Please download and read the accompanying instructions before you attempt to use the program.


    Signs Program in Gauss

    Signs Program in Matlab

    Signs Program in Stata's MATA Language

    Signs Program in SAS