© 2011 Towers Watson. All rights reserved. Intelligent Use of Competitive Analysis 2012 Ratemaking and Product Management Seminar Kelleen Arquette March 20-21, 2012
© 2011 Towers Watson. All rights reserved.
Intelligent Use of Competitive Analysis2012 Ratemaking and Product Management Seminar
Kelleen Arquette
March 20-21, 2012
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 2
Session agenda and objectives
Today, we will discuss Different industry approaches to competitive analysis Key challenges in performing quantitative competitive analysis Analysis of “on-the-street” prices
AGENDA
Competitive Market Analysis
towerswatson.com 3
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 4
Insurers use various approaches to competitive market analysis — we will consider advantages and disadvantages of six specific techniques
High Degree of Sophistication
Low Degree of Sophistication
CMA: Qualitative Analysis
Market Basket
Company Statistics
Competitor Rate
Changes
Agent Feedback
CMA: Quantitative
Analysis
These options are not mutually exclusive — different approaches can be used in combination
COMPETITIVE MARKET ANALYSIS
External Quotes
Addressing Challenges
towerswatson.com 5
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 6
Although generally more effective, advanced competitive market analysis techniques pose certain challenges
Key Challenges Company selection Credit/tier alignment Missing variables Choice of a comparative rater Particularly for homeowners, alignment of product type Validation of results
The next several pages briefly address each of these challenges
CHALLENGES
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 7
Selecting which competitors to include is important…and trickier than one might think
The ideal is a mix of direct competitors and industry leaders The target market segment should be considered
Competitors targeting the preferred market may be different than competitors targeting the non-standard market
Once you choose a competitor group, selecting which particular company to rate can be challenging For example, Allstate writes auto insurance in at least 14 companies across the country
Relative premium volume may not be a good indicator, as one entity may be a new company (where all new business is being written)
Agent feedback and rate filing reviews can help select the “right” company Some companies write only package policies (personal auto and homeowners on the
same policy). This should be considered in the company selection (impact on coverage alignment and underwriting selection criteria)
COMPANY SELECTION
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 8
Creating an accurate alignment between competitor credit groups and tiers is critically important
If a comparative rater provides automatic credit scoring alignment at all, it usually provides a simplified approach, such as this:
Credit Group
A
B
C
D
Credit Group
1
2
Company A Company B Alignment
Company A
Company B
A, B 1
C, D 2
This approach has obvious disadvantages and could lead to incorrect conclusions
CREDIT/TIER ALIGNMENT
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 9
Alternative approaches to aligning credit groups/tiers can increase accuracy (but can be costly and/or time consuming)
Insurance Score Alignment (Distribution Mapping) – Alignment based on company filed distributions by credit score range or tier Relies on publicly available data More accurate than uniform distribution assumption
Insurance Score Assignment – Assignment based on programmed competitor credit scoring algorithms Requires data at the individual credit attribute level Relies on publicly available data Processing current book of business through programmed algorithms results in an optimal credit
score assignment for each competitor Assumptions may still be necessary, depending on the data source and competitor(s)
CREDIT/TIER ALIGNMENT
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 10
Competitor 3 InsScore x Competitor 4 InsScore
60
70
80
90
100
110
120
130
140
150
160
170
180
190
450 500 550 600 650 700 750 800
Competitor 3 InsScore
Co
mp
eti
tor
4 In
sS
co
re
Credit-based insurance score used by different companies assess risk differently
“Company A” and “Company B” are personal auto insurers Both are national writers with
market share in the top 10 in most states
Credit-based insurance scoring models Company A uses a vendor model
— High score is best (lowest risk)
Company B uses a proprietary model— Low score is best (lowest risk)
Models were found in publicly available filings
Models were programmed using actual credit data
Correlation between the insurance scores, but not perfect
Expect diagonal line if models assessed risk in the same way
No hits/no scores are excluded Company A Insurance ScoreC
om
pan
y B In
suran
ce Sco
re
Company A X Company B Insurance Score Contour Plot
CREDIT/TIER ALIGNMENT
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 11
Companies take different approaches to tier
Tier is a combination of the credit-based insurance score and other variables for both companies
Company A and Company B use different variables in the tier determination
The same data set was used to generate both tier graphs
Examples of variables used include Prior liability limits Lapses in coverage Education Occupation Accident and violations Length of time insured with prior carrier
CREDIT/TIER ALIGNMENT
Company A Tier
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Tier
Ve
hic
le C
ou
nt
Company B Tier
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Tier
Ve
hic
le C
ou
nt
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 12
It is possible for a policy considered low risk for Company A to be considered high risk for Company B
Any tier for Company A has a range of tiers for Company B
Can explain pricing differences at the individual vehicle/policy level
Insurance score or tier alignment approaches miss the opportunity to look at the different approaches to risk assessment at the policy level
Company A Tier x Company B Tier
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Company A Tier
Dis
trib
uti
on
Company B Tier 18
Company B Tier 17
Company B Tier 16
Company B Tier 15
Company B Tier 14
Company B Tier 13
Company B Tier 12
Company B Tier 11
Company B Tier 10
Company B Tier 9
Company B Tier 8
Company B Tier 7
Company B Tier 6
Company B Tier 5
Company B Tier 4
Company B Tier 3
Company B Tier 2
Company B Tier 1
CREDIT/TIER ALIGNMENT
Company A Tier x Company B Tier
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Company A Tier
Ve
hic
le C
ou
nt
Company B Tier 18
Company B Tier 17
Company B Tier 16
Company B Tier 15
Company B Tier 14
Company B Tier 13
Company B Tier 12
Company B Tier 11
Company B Tier 10
Company B Tier 9
Company B Tier 8
Company B Tier 7
Company B Tier 6
Company B Tier 5
Company B Tier 4
Company B Tier 3
Company B Tier 2
Company B Tier 1
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 13
In some cases, a company may simply not collect accurate data on a rating variable that a competitor uses
Depending on the importance of the variable, how missing values are populated can materially affect the results
External data can sometimes be used to fill in missing values using sampling techniques Census and other external data Distributions obtained from competitor filings
Care should be taken in how these variables are populated Suppose a company does not collect data on a 55 & Retired Discount, but driver age is readily
available From census data and other publicly available data, one can obtain a population estimate of
individuals who are retired However, constraints should be placed on the sampling approach to avoid illogical results (e.g.,
a 25-year-old who is “retired”)
MISSING VARIABLES
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 14
There are a number of important due diligence considerations in selecting a comparative rater
Are the rates for your company and the selected competitors already included in the tool (and if not, what are the additional costs to include?) What are the alternatives if additional cost is prohibitive?
Does the software support batch rating hundreds of thousands of policies (or more) in a timely fashion? How much computing power is necessary?
Does the platform accurately perform: Driver assignment for personal auto? Territorial assignment? Tier assignment?
What process does the vendor have in place to ensure accurate results? What types of training and support services does the vendor provide? Does the vendor have a tool to convert your exposure data into a format that the batch
rater can use? Does the vendor have appropriate marketplace knowledge to understand complex rate
filings?
Although companies can decide to perform this work in-house, the effort has significant staffing and cost implications
COMPARATIVE RATER
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 15
Proper alignment of product/coverage is important in order to draw appropriate conclusions
Coverage/FeatureCompetitor A
“Standard” HO-3 PolicyCompetitor B
“Basic” HO-3 Policy
Earthquake Included Excluded
Water Backup Excluded Included
Coverage A Actual cash value, with possible limited replacement cost coverage endorsement
Replacement cost coverage
Coverage C 70% of Coverage A 85% of Coverage A
Identity Theft Included Excluded
PRODUCT ALIGNMENT
State X — Homeowners
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 16
What process will be followed to ensure that the calculated competitor rates are accurate?
Rater accuracy should be considered in selecting a third-party vendor Even the larger comparative rating vendors are often not accurate
Programming errors Credit/tiering alignment Oversimplification/misunderstanding of a competitor’s rating approach Goal may be to get “close enough”
Certain actions can be taken to increase the accuracy of the analysis Hand-rating of a random sampling of policies (which can be time consuming) Verifying calculated average premiums with certain filed materials Conversations with agents (“gut checks”)
VALIDATION OF RESULTS
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 17
Analysis of “On-The-Street” Prices
towerswatson.com 17
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 18
ILLUSTRATIVE
Average Premium All Coverages Combined
Company A Company B Company C
Company D
Company E Company F
Company G
XYZ Company
$0
$200
$400
$600
$800
$1,000
$1,200
$1,400
Average Premium
Beware of potential inherent bias in using current policy mix of business
QUANTITATIVE ANALYSIS
A quantitative analysis can compare new and renewal business pricing against competitors for the entire book…
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 19
$500
$700
$900
$1,100
$1,300
$1,500
$1,700
1 2 3 4 5 6 7 8 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
0
500
1,000
1,500
2,000
2,500
3,000
3,500
Policy Count Company Premium Second Highest Premium Second Lowest Premium
QUANTITATIVE ANALYSIS
Average Premium with Hi/Lo Band by Symbol Policy
Count
Symbol
ILLUSTRATIVE
Average Premium
…and by rating factor/segment…
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 20
The target market position should be identified and then metrics can be developed to monitor competitive position relative to target
…to identify the current competitive position
QUANTITATIVE ANALYSIS
Competitive Metrics $ or % Competiveness % Wins Relative to Market Rank Average premiums
Other metrics may also be considered when determining the target market position $ or % Impacts $ or % Subsidization Expected loss ratio Expected retention Hit/conversion ratio
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 21
Now what?
Identified current competitive position Calculated current premiums for your company and selected competitors, overcoming many hurdles
in calculating the competitor premiums Analyzed the premiums to determine the current competitive position
Identified target competitive position How do you get there?
How should the current rating plan be revised to achieve the target competitive position?
QUANTITATIVE ANALYSIS
© 2011 Towers Watson. All rights reserved. Proprietary and Confidential. For Towers Watson and Towers Watson client use only.
towerswatson.com 22
Contact Information
Kelleen Arquette Consultant, Towers Watson [email protected] 509-258-8876
CONTACT INFORMATION