CS-18: THE
ACTUARIAL ROLE IN
THE AUDIT
Brian E. Johnson, ACAS, MAAA
Senior Underwriter
Antitrust Notice

The Casualty Actuarial Society is committed to adhering strictly to the letter and
spirit of the antitrust laws. Seminars conducted under the auspices of the CAS are
designed solely to provide a forum for the expression of various points of view on
topics described in the programs or agendas for such meetings.

Under no circumstances shall CAS seminars be used as a means for competing
companies or firms to reach any understanding – expressed or implied – that
restricts competition or in any way impairs the ability of members to exercise
independent business judgment regarding matters affecting competition.

It is the responsibility of all seminar participants to be aware of antitrust regulations,
to prevent any written or verbal discussions that appear to violate these laws, and to
adhere in every respect to the CAS antitrust compliance policy.
2
The Deal Team
Traditionally the deal team was lead by the UW or Account Exec
and the Actuary was just called upon to “crunch” the numbers
3
Data/Structure
Loss
Cost
4
The Deal Team

Only looking at the submission could lead to a misinterpretation of
what is going on at the company

By looking at the UW files actuaries can quite often interpolate
information wrt the submission that is not evident

In addition you will quite often be able to extract anecdotal
evidence that may not otherwise be available in the company’s
system (information not captured)
5
The Deal Team

Some examples of anecdotal information that can be obtained
from the UW files include:
– Trends in usage of pricing components (credits/debits, IRPM mods,
adjusted ILFs, a-rates, class codes, etc)
– Effects of minimum premium not otherwise captured in rate change
– UWs consideration of the specificities of a risk which lead to pricing
decisions (or lack thereof)
6
The Deal Team

The actuary should also sit in meetings with a cross section of
insurance professionals i.e. UWs and claims. This will allow the
actuary to hear about important changes to the company and
relate them to things they may see in the data.
7
Case Study – A Tale of Two Actuaries

Two actuaries (at different companies) Alpha and Beta both
received a submission from Pandorum Insurance Company who is
looking to buy a quota share on their umbrella business.

Alpha decides that he is too busy to participate on the UW/Claims
review. Instead he sends a couple of questions back to
Pandorum’s actuary to confirm his understanding of the data in
the submission

Beta attended The Actuarial Role in the Audit at the CARe meeting
and very wisely decides it is well worth his time to attend the
audit.

First we will look at a few aspects of the submission and then see
how each actuary utilizes them in their analysis
8
Case Study – A Tale of Two Actuaries

One exhibit that was of interest to both actuaries was the
historical and projected limits distribution

According to the submission, Pandorum had made a concerted
effort to move up into higher layers (away from the lead position)
to layers that they perceive to be more adequate.

The exhibit is as follows:
9
2007 Distribution
Limit
≤ $1M
>$1M -$2M
>$2M -$3M
>$3M -$4M
>$4M -$5M
>$5M -$10M
>$10M -$15M
>$15M - $20M
>$20M - $25M
Total
% of Total
Attachment
>$1M-$5M
190,825
172,971
238,791
1,346,194
2,010,549
7,876,439
405,381
54,271
127,801
12,423,223
45.73%
>$5M-$10M
0
0
255,740
21,884
382,964
3,161,927
388,650
0
355,040
4,566,205
16.81%
>$10M-$15M
0
0
0
210,958
600,467
1,656,427
1,233,300
1,380,065
250,874
5,332,092
19.63%
>$15M-$20M
0
0
0
30,637
144,345
1,023,931
1,823,161
152,509
440,341
3,614,923
13.31%
>$20M-$25M
0
0
0
0
73,967
258,227
568,760
113,970
47,006
1,061,930
3.91%
>$25M-$50M
0
0
0
0
45,956
124,299
0
0
0
170,255
0.63%
Total
190,825
172,971
494,531
1,609,673
3,258,247
14,101,250
4,419,253
1,700,816
1,221,062
27,168,628
% of Total
0.70%
0.64%
1.82%
5.92%
11.99%
51.90%
16.27%
6.26%
4.49%
100.00%
>$5M-$10M
0
0
286,702
24,533
429,328
3,544,730
435,703
0
398,024
5,119,019
20.19%
>$10M-$15M
0
0
0
214,363
610,157
1,683,156
1,253,202
1,402,335
254,922
5,418,134
21.37%
>$15M-$20M
0
0
0
29,704
139,950
992,760
1,767,659
147,866
426,936
3,504,876
13.82%
>$20M-$25M
0
0
0
0
72,191
252,027
555,103
111,233
45,877
1,036,431
4.09%
>$25M-$50M
0
0
0
0
46,400
125,500
0
0
0
171,899
0.68%
Total
155,248
140,722
480,972
1,363,808
2,933,726
13,006,121
4,341,468
1,705,587
1,229,733
25,357,386
% of Total
0.61%
0.55%
1.90%
5.38%
11.57%
47.87%
17.12%
6.73%
4.85%
100.00%
2008 Distribution
Limit
≤ $1M
>$1M -$2M
>$2M -$3M
>$3M -$4M
>$4M -$5M
>$5M -$10M
>$10M -$15M
>$15M - $20M
>$20M - $25M
Total
% of Total
Attachment
>$1M-$5M
155,248
140,722
194,271
1,095,208
1,635,701
6,407,949
329,802
44,153
103,973
10,107,027
39.86%
10
2009 Distribution
Limit
≤ $1M
>$1M -$2M
>$2M -$3M
>$3M -$4M
>$4M -$5M
>$5M -$10M
>$10M -$15M
>$15M - $20M
>$20M - $25M
Total
% of Total
Attachment
>$1M-$5M
119,670
108,473
149,750
844,223
1,260,852
4,939,459
254,222
34,035
80,146
7,790,830
33.09%
>$5M-$10M
0
0
317,663
27,182
475,692
3,927,533
482,755
0
441,007
5,671,834
24.09%
>$10M-$15M
0
0
0
217,767
619,846
1,709,885
1,273,103
1,424,604
258,971
5,504,175
23.38%
>$15M-$20M
0
0
0
28,772
135,556
961,589
1,712,158
143,223
413,531
3,394,828
14.42%
>$20M-$25M
0
0
0
0
70,415
245,826
541,446
108,497
44,749
1,010,933
4.29%
>$25M-$50M
0
0
0
0
46,843
126,700
0
0
0
173,544
0.74%
Total
119,670
108,473
467,413
1,117,944
2,609,204
11,910,993
4,263,684
1,710,359
1,238,404
23,546,144
% of Total
0.51%
0.46%
1.99%
4.75%
11.08%
43.84%
18.11%
7.26%
5.26%
100.00%
Projected 2010 Distribution
Limit
≤ $1M
>$1M -$2M
>$2M -$3M
>$3M -$4M
>$4M -$5M
>$5M -$10M
>$10M -$15M
>$15M - $20M
>$20M - $25M
Total
% of Total
Attachment
>$1M-$5M
84,092
76,224
105,230
593,237
886,004
3,470,969
178,642
23,916
56,319
5,474,634
25.19%
>$5M-$10M
0
0
348,625
29,832
522,056
4,310,337
529,808
0
483,991
6,224,648
28.64%
>$10M-$15M
0
0
0
221,171
629,536
1,736,614
1,293,004
1,446,874
263,019
5,590,217
25.72%
>$15M-$20M
0
0
0
27,839
131,162
930,417
1,656,656
138,581
400,126
3,284,781
15.11%
>$20M-$25M
0
0
0
0
68,639
239,626
527,789
105,760
43,620
985,434
4.53%
>$25M-$50M
0
0
0
0
47,287
127,901
0
0
0
175,188
0.81%
11
Total
84,092
76,224
453,855
872,079
2,284,683
10,815,864
4,185,899
1,715,131
1,247,074
21,734,902
% of Total
0.39%
0.35%
2.09%
4.01%
10.51%
49.76%
19.26%
7.89%
5.74%
100.00%
Case Study – A Tale of Two Actuaries

In addition the submission has the following rate change summary
TY
2003
2004
2005
2006
2007
2008
2009
2010
Exposure
Adjusted
Rate Chng
12.42%
8.54%
2.00%
-7.53%
-9.24%
2.00%
7.50%
7.00%
12
Case Study – A Tale of Two Actuaries

Both actuaries notice that there is in fact a shifting in the
attachment point to higher layers.

In addition, they also notice that they are also writing higher
limits.

Both actuaries ask if there has been a shifting in the class of
business commensurate with the shifting in the limits/attachment
distribution (the old adage that people who buy higher limits need
them)

The response back is that they are writing the same kinds of risk
and that there has been no shift in the hazard group
13
Case Study – A Tale of Two Actuaries

Both actuaries also notice that the Pandorum seems to have
achieved much bigger rate changes in the soft market than their
competitors.

They both ask to confirm that these are actual exposure adjusted
rate changes and not premium increases.

The response is that these rate change have been adjusted for
increases and decreases in the exposure.

The response also goes on to say that the rate change is
calculated as part of the UW process. Details can be found in
each UW file.

Alpha (who is a very busy man) accepts this explanation and
moves forward with his analysis. Beta makes a note to confirm
the process while looking at UW files.
14
Case Study – A Tale of Two Actuaries

While reviewing files Beta notes how the rate change was
calculated in several files. Here is one example what he sees:
Vault101 Construction Company
Rate Change Calculation
Expiring premium:
55,000
Renewal Premium:
62,500
Expiring Sales: 1,500,000
Renewal Sales: 1,650,000
Expiring Limit:
5,000,000
Renewing Limit: 10,000,000
Calculated Rate Change:
3.3%
15
Case Study – A Tale of Two Actuaries

Beta immediately notices that the “rate change” does not account
for the expanding limit.

He does a quick adjustment to reflect the difference in limit and
sees for this particular account the change in rate is not +3.3%
but instead a -1.7%.
16
Case Study – A Tale of Two Actuaries

Beta sets up a meeting to talk to the Chief Pricing Actuary and the
Senior VP of Underwriting.

Beta asks about the rate change calculation in the files compared
to what was in the submission. In addition he also asks about any
changes going forward that may not be reflected in the data.
17
Case Study – A Tale of Two Actuaries

Pandorum’s actuary confirms that the rate increase does not
reflect change in limit. Since their retention ratio is in the high
90’s and historically the limits didn’t change much the system was
not created with that kind of complexity.

He goes on to say it is on ITs list of enhancements for their
pricing/underwriting system.
18
Case Study – A Tale of Two Actuaries

When asked about company changes the SVP of UW comments
that due to Agent feedback they are going reduce their minimum
premium per mill from $1,500 to $1,250 for 75% of their classes.

Given the higher average attachment and propensity to hit
minimum premiums more frequently, Beta asks if they have tried
to estimate the change in rate due to the lowering of the minimum

They both comment that this kind of information is not currently
captured in their pricing system but that it is on IT’s list of future
enhancements.
19
Case Study – A Tale of Two Actuaries

Beta and his UW audit team get through 35 files. Beta is able to
use the data captured in these files to make subjective
adjustments to the rate change to account for the decrease in the
minimum.

In addition, he also uses the change in limits/attachment
distribution and Pandorum’s ILF tables to adjust the rate change
history.

Even though the minimum premiums per million are decreasing
Beta feels as though there is some additional rate lift provided by
the mins.

Using data from the sample of files reviewed Beta estimates the
net rate lift due to minimums.
20
Case Study – A Tale of Two Actuaries

He runs his adjustments by Pandorum’s actuary for ensure they
are somewhat reasonable. The actuary agrees the adjustments
make senses.

Here is a comparison of the unadjusted and adjusted rate
changes.
Original Rate Changes
Exposure
Adjusted
TY
Rate Chng
2003
12.42%
2004
8.54%
2005
2.00%
2006
-7.53%
2007
-9.24%
2008
2.00%
2009
7.50%
2010
7.00%
Adjusted Rate Changes
TY
2003
2004
2005
2006
2007
2008
2009
*2010
Expo Adj. Avg
Rate Chng
12.42%
8.19%
1.86%
-9.71%
-11.03%
-8.75%
4.17%
4.00%
2010 adjusted for min reduction
21
Understanding the Big Picture

In addition to a better understanding of the data, the actuary may
also gain additional insight to other changes at the company that
should be reflected in the loss cost selection.

For example, quite often companies will stop writing certain
classes or lines of business that have been historically
unprofitable for them.

In order to determine how much “credit” can be given for these
kinds of changes the actuary must have a very thorough
understanding of the changes in the UW guidelines and how they
are being implemented by the desk UW’s.
22
Understanding the Big Picture

While understanding rate changes on renewals is an important
step towards estimating loss cost, it is also important to
understand the marginal effect on rate levels of writing new
business.

If company ABC’s average rate change on their renewal book is
say -3%, the business that was non-renewed (written new by
company XYZ) is written at a rate decrease bigger than -3%.

By that same token, when ABC writes a new piece of business,
there may be a difference in the average rate adequacy of the new
business versus the renewal book.
23
Understanding the Big Picture

The actuary can gather statistics from the files which will allow
them to estimate the adequacy of premiums of new business
versus renewed business.

Things to look for and compare include: average rate level by class
of business, average unit rate per vehicle, difference in
discretionary mods, difference in take-up factors in the case of
umbrella business

While estimating the new business rate effect is not exact it
should be considered especially in cases where the company is
growing and there is a material amount of new business.
24