Econometric Theory/Asymptotic Convergence

Asymptotic Convergence edit

Modes of Convergence edit

Convergence in Probability edit

Convergence in probability is going to be a very useful tool for deriving asymptotic distributions later on in this book. Alongside convergence in distribution it will be the most commonly seen mode of convergence.

Definition edit

A sequence of random variables   converges in probability to   if:

 
 
 

an equivalent statement is:

 
 

This will be written as either   or  .

Example edit

 

We'll make an intelligent guess that this series converges in probability to the degenerate random variable  . So we have that:

 

Therefore our definition for convergence in probability in this case is:

 
 
 

So for any positive values of   we can always find an   large enough so that our definition is satisfied. Therefore we have proved that  .

Convergence Almost Sure edit

Almost-sure convergence has a marked similarity to convergence in probability, however the conditions for this mode of convergence are stronger; as we will see later, convergence almost surely actually implies that the sequence also converges in probability.

Definition edit

A sequence of random variables   converges almost surely to the random variable   if:

 
 

equivalently

 

Under these conditions we use the notation   or  .

Example edit

Let's see if our example from the convergence in probability section also converges almost surely. Defining:

 

we again guess that the convergence is to  . Inspecting the resulting expression we see that:

 

Thereby satisfying our definition of almost-sure convergence.

Convergence in Distribution edit

Convergence in distribution will appear very frequently in our econometric models through the use of the Central Limit Theorem. So let's define this type of convergence.

Definition edit

A sequence of random variables   asymptotically converges in distribution to the random variable   if   for all continuity points.   and   are the cumulative density functions of   and   respectively.

It is the distribution of the random variable that we are concerned with here. Think of a students-T distribution: as the degrees of freedom,  , increases our distribution becomes closer and closer to that of a gaussian distribution. Therefore the random variable   converges in distribution to the random variable   (n.b. we say that the random variable   as a notational crutch, what we really should use is  /

Example edit

Let's consider the distribution Xn whose sample space consists of two points, 1/n and 1, with equal probability (1/2). Let X be the binomial distribution with p = 1/2. Then Xn converges in distribution to X.

The proof is simple: we ignore 0 and 1 (where the distribution of X is discontinuous) and prove that, for all other points a,  . Since for a < 0 all Fs are 0, and for a > 1 all Fs are 1, it remains to prove the convergence for 0 < a < 1. But   (using Iverson brackets), so for any a chose N > 1/a, and for n > N we have:

 

So the sequence   converges to   for all points where FX is continuous.

Convergence in R-mean Square edit

Convergence in R-mean square is not going to be used in this book, however for completeness the definition is provided below.

Definition edit

A sequence of random variables   asymptotically converges in r-th mean (or in the   norm) to the random variable   if, for any real number   and provided that   for all n and  ,

 

Cramer-Wold Device edit

The Cramer-Wold device will allow us to extend our convergence techniques for random variables from scalars to vectors.

Definition edit

A random vector  .


Relationships Between Modes of Convergence edit

Law of Large Numbers edit

Central Limit Theorem edit

Let   be a sequence of random variables which are defined on the same probability space, share the same probability distribution D and are independent. Assume that both the expected value μ and the standard deviation σ of D exist and are finite.

Consider the sum  . Then the expected value of   is nμ and its standard error is σ n1/2. Furthermore, informally speaking, the distribution of Sn approaches the normal distribution N(nμ,σ2n) as n approaches ∞.

Continuous Mapping Theorem edit

Slutsky's Theorem edit