endogeneity omitted variable bias question

[Prathvajeeth Rajmohan]

Hi there this is stats based but more a theoretrical question:

What exactly does endogenous variables mean, I know there are different forms but I mean endogenity in the sense of omitted variable bias.

I am seeing the effect of derivative usage on firm value, and so regression firm value with my independant variable derivative usage and set of control variables.

I know there is an endogeneity issue in the sense that there a characteristics both unobservable (eg managerial quality) and observable that have a postive effect on firm value and are postively correlated with derivative use.


I understand the enodogenity in this sense mean that these characteristics that are captured by the error term are linked to the the explantoryy variable deriavtive usage.

What my main question is is which variables are ones reffered to as endogenous?

a)is it the firm value that is enodogenous or deriavtive usage that is enodogenous

b) or are firm value and derivative usage both "endogenous variables"

c) or is the observed/unobserved characterisitcs that are referred to as endogenous.

Would really love a very straightforward answer in simple terms as just starting to learn stats. Thanks so much.

[Carlo Lazzaro]

Prathvajeeth:
you probably refer to the case when one omitted variable (embodied in residuals) is, at the same time, correlated to the dependent variable and one independent variable. In that case, your independent variable is endogenous (and you need to switch to instrumental variables regression).
If you're a Stata newbie and you're taking your first steps in econometrics,I would recommend the following textbook: https://www.stata.com/bookstore/micr...ta/index.html; chapter 6 covers what you're interested in.