Page 201 - ic92 actuarial

Foundations of Casualty Actuarial Science

Causality implies a closer relation to cost than
correlation. E.g, mileage in automobile insurance can
be considered a causal variable, as the more miles
a driver drives, more is the cost of insurance. Loss
costs can be divided into claim frequency and
average claim cost.

Causal variables then could be considered to be
directly related to claim frequency or average claim
cost. Automobile mileage is inversely proportional
to average claim cost. Unfortunately, however, the
categorization of variables as 'causal' or 'non-causal'
is ambiguous. E.g, for the automobile insurance,
where, when and how one drives might be more
important than the mileage.

From actuarial point of view, correlated variables
provide more accurate premiums and are thus more
desirable in a competitive market place. Eliminating
correlated non-causal variables may produce a less
accurate rating system and cause market
corrections.

Sashi Publications - www.sashipublications.com 201

Copyright@ The Insurance Times. 09883398055 / 09883380339