Errata for Malcolm Kemp’s ExtremeEvents:
Robust Portfolio Construction in the Presence of Fat Tails, published by
Wileys
[this page | pdf | references | back links | custom searches]
In paragraph (a) on page 34, after the words “then the
Central Limit Theorem implies that over this short time interval the return
should be approximately Normal” there should be a footnote saying:
“To be more precise, we are here envisaging a situation
where we have, say, ‘similar’ contributory
factors (each with finite variance) and as we
give less and less weight to each individual factor but have more and more of
them. If instead the contributory factors become more and more non-Normal as (i.e.
do not stay ‘similar’ as ) then the CLT may break
down, see Kemp (2010).”
A counter-example, if the distributional form of the
individual factors changes as increases, which is what
“Kemp (2010)” here refers, to is given in Extreme Events Errata
(1).
On page 171 Equation 5.25 should read:
And Equation 5.26 should read:
On page 171 Equation 5.25 should read: