ASSUMPTIONS MADE IN CAT MODELLING REPORT

A CAT Modelling report gives a wealth of information that Reinsurance managers and Underwriters can use for their decision making. Data in the report is given for various return periods like 1000, 500,250,200, 100,50,25, 10 & 5. It gives expected losses for the portfolio for each of the return periods. It gives this data- Line of Business wise- Commercial Property, Industrial Property, Home, Engineering, Motor Private, Motor Commercial, etc. It also gives data on the expected losses to the net account, Quota Share, First Surplus Treaty, Second Surplus Treaty and Facultative portfolio.

You may ask as a reinsured, why should I bother about the expected losses to reinsurer account? The answer is- it is useful data even for the reinsured. I would request you to think how this data can be used by the reinsured. Please write your answer in the comments.

The above data is very useful for reinsured. How this data is useful is not the topic of this blog. What I want to talk is what the assumptions based on which these numbers have been arrived at. If data input or assumptions are wrong, then, the results of the CAT Modelling report are being used so extensively become suspect. Garbage In -Garbage Out.

Difficult to go into each of the aspects but some key mistakes could be:

a) Floater Sum Insured-how is it being distributed to various Cresta Zones?
b) Gross Profit figures -Single value given for all locations-how is it being distributed to various Cresta Zones?
c) Mortgage Home Policies-if annual 10% increase in sum insured is opted, then, is it being accounted for? It needs to be noted for many insurers, expected losses from Home LOB constitute more than 50% of the total exposure (because they are multiyear policies-exposure grows in geometric progression).
d) If these are loss limits for NAT CAT perils – is it being considered? It may be noted that some systems in India are not designed to enter different sum insured/Loss Limit for each peril).
e) Yearbuilt for industrial buildings are taken as 2010 – If we are not capturing year built in the system- how can we be sure about the average age of the buildings?
f) Construction of industrial buildings assumed as 25% light metal frame and 75% steel- If we are not capturing this data in system, how do we know the average construction features of our portfolio?
g) Number of Storeys-Commercial considered as 4 Storeys -if we are not capturing this data-how are we sure about the average number of Storeys of our portfolio?

There are many more such assumptions mentioned in the report. Gives lessons for us on what risk details need to be captured in the system.

Blog by Atmaram Cheruvu