Hidden Action Modelling:

The employer (the principal) hires the employee (the agent) to execute a project (say, developing a software on financial security). Let us  assume that there are two possible outcomes of the project: it succeeds or fails. When the project succeeds, it yields an outcome x,  in money terms. On the other hand, it may also fail with the associated outcome of x_f and  x_s > x_f ≥ 0.  The principal is risk neutral, as he faces many  independent risks and thus can diversify the associated risks  to  his relationship  with the  employees. Through such  an arrangement, the principal would be interested only on the expected revenue. The agent, in our example, can choose among two levels of efforts  e :  she can work  (e =1) or bunk (e = 0). We can normalise the cost of effort for the agent to  e without introducing any distortion to our analysis. Let us start with a contract agreement in which the employer offers the worker a wage w  once the project is finished. We assume the worker's utility to be

u(w)  =  v(w)  - e, where  v'  > 0  and  v"  > 0,  i.e., 

utility  is increasing and concave in wealth. Therefore, the agent is risk-averse.