/

Stable Distributions

4. The Generalised Central Limit Theorem

[this page | pdf | references | back links]

Return to Abstract and Contents

Next page

 

4.1          The two main reasons why stable laws are commonly proposed for modelling return series are:

 

(a)    The Generalised Central Limit Theorem. This states that the only possible non-trivial limit of normalised sums of independent identically distributed terms is stable; and

 

(b)   Empirical. Many large data sets exhibit fat tails (and skewness), and stable distributions form a convenient family of distributions that can cater for such features (with choice of  and  allowing different levels of fat-tailed-ness or skewness to be accommodated).

 

We focus below on the former, since there are other families of distributions that can be parameterised in ways that can fit different levels of fat-tailed-ness or skewness, including ones simpler to handle analytically such as ones with quantile-quantile plots versus the Normal distribution that are polynomials rather than straight lines, see e.g. Kemp (2009).

 

4.2          The classical Central Limit Theorem states that the normalised sum of independent, identically distributed random variables converges to a Normal distribution. The Generalised Central Limit Theorem shows that if the finite variance assumption is dropped then the only possible resulting limiting distribution is a stable one as defined above. Let  be a sequence of independent, identically distributed random variables. Then there exist constants  and  and a non-degenerate random variable  with

 

 

if and only if  is stable (here  means tends as  to the given distributional form).

 

4.3          A random variable  is said to be in the domain of attraction of  if there exist constants  and  such that the equation in Section 4.2 holds when  are independent identically distributed copies of . The Generalised Central Limit Theorem thus shows that the only possible distributions with a domain of attraction are stable distributions as described above. Distributions within a given domain of attraction are characterised in terms of tail probabilities. If  is a random variable with  and  with  for some  as  then  is in the domain of attraction of an -stable law.  must then be of the form

 


NAVIGATION LINKS
Contents | Prev | Next


Desktop view | Switch to Mobile