Do Ladies Wear Girdles

INTRODUCCION

INTRODUCTION
We have studied which is a stochastic process (a set of random variables indexed
an index t) and is a series (a realization of a stochastic process). We are interested in the stochastic process that generates a given economic variable, but only observe a time series. WHAT CAN BE DONE? At first nothing ....... with unless you are willing to take on two key assumptions:

• Stationarity. This assumption of identically distributed replaces the previous courses and is stronger.
• Ergodicity. This course replaces the independent and vague terms mean asymptotic uncorrelated. These

dos supuestos seremos capaces de describir ciertos aspectos del proceso estocastico habiendo observado solo una serie temporal. Recordad el ejemplo en clase cuando construimos una serie temporal eligiendo de forma aleatoria un alumno por columna y como eramos capaces de poder predecir ("decir algo") sobre la fila en la que estarian sentados los primeros estudiantes de la clase de al lado. Esto lo haciamos calculando la media de nuestra serie temporal, que es un estimador consistente gracias a los dos supuestos mencionados previamente.

DONDE ESTAMOS? Queremos predecir Z(t+1) dado el conjunto de informacion I(t). Si la funcion de perdida es cuadratica, la mejor prediccion es E[Z(t+1)/I(t)]. Esta esperanza condicional puede ser muy complicada and therefore we are forced to further restrict the world where we think we live. In this course, this world is going to be parametric and linear. Within this world are the ARMA models. In the next chapter will witness the birth of these models.

do not you remember the first practice of the course. If it costs you do not lose heart that the important thing is to let you know how to do prior to the next and that teachers are practices and practical classes bi-weekly.