Modeling the error distribution using the Student t distribution

[Kevin Forbes]

I am estimating an ARCH model assuming that the error term has a Student t distribution. The output reports a degree of freedom value of about seven which indicates that the error distribution has more kurtosis than the Gaussian distribution. This makes sense to me. However, it also reports a parameter “/lndfm2” with an estimated value of 1.66. Any sense of what this parameter indicates?

[Jackson Monroe]

What command did you use? It's easier to understand the output if we know what is generating it.

[Kevin Forbes]

The distribution(t) command is one of the options when an ARCH model is estimated Here is some text to clarify matters "distribution(dist [#]) specifies the distribution to assume for the error term. dist may be gaussian, normal, t, or ged. gaussian and normal are synonyms, and # cannot be specified with them.

If distribution(t) is specified, arch assumes that the errors follow Student's t distribution, and the degree-of-freedom parameter is estimated along with the other parameters of the model"


Thanks for your question!