Capital Levels in the Canadian Property/Casualty Insurance Industry

1
Capital Levels in the Canadian Property/Casualty
Insurance Industry

• Peter Carayannopoulos
• Mary Kelly
• Wilfrid Laurier University

2
Agenda

• Motivation.
• Canadian marketplace.
• Areas of investigation
• Capital holdings and firm risk.
• Regulatory changes in 2003 and the distribution
  of capital in the industry
• Conclusions.

3
Motivation

• Look at holdings of Canadian p/c insurers taking
  regulatory framework as given.
• Do capital holdings reflect firm risk?
• What is initial impact of changes in solvency
  requirements in 2003?
• Little research undertaken on capital holdings of
  Canadian p/c insurers.

4
Areas of Investigation

• What firm characteristics influence capital
  holdings between 1990 2004?
• What is the impact of the new MCT test on
• The level of capital holdings?
• The relationship between capital holdings and
  firm characteristics?
• The relationship between capital holdings and a
  firms portfolio of assets and liabilities?

5
P/C Insurance in the Canadian Economy

• Over 200 private insurance companies in Canada
  organized in approximately 120 groups.
• A small industry
• 35.9 million in premiums in 2003,
• 71 billion in assets in 2003.
• 2.6 of world wide p/c insurance premiums.
• Market share of top 10 firms around 55.

6
Regulation of Insurers

• OSFI regulates solvency via level of capital,
  adequacy of reserves, prudent investment
  strategies.
• Provincial regulators monitor products and
  practices.
• Firms may also be subject to provincial solvency
  requirements.

7
Minimum Asset Test (MAT) vs. Minimum Capital Test
(MCT)

• MAT
• Value asset levels on liquidation basis
• Assets Available total assets held by firm less
  those non-admitted or otherwise not available.
• Assets Required total liabilities required
  margin recoverables.
• MAT statistic is

• MCT
• Value asset levels on on-going basis.
• Capital required based on both asset and
  liability risk.
• Asset risk type of security, maturity and grade.
• Liability risk unearned premium reserve, NPW by
  line.
• Calculation of capital required / capital
  available must exceed 150 to pass test.
• Recommended targets of 170 - 210.

• Firm must have positive ratio to pass test.
• Higher ratio needed to avoid regulatory oversight

8
Summary of Insurer Data1990 - 2004
Number Strictly Cdn Insurers Number Strictly Cdn Insurers Number Strictly Cdn Insurers 64
Number Of Mutual Insurers Number Of Mutual Insurers Number Of Mutual Insurers 50
Number Of Firms Number Of Firms Number Of Firms 268
Number Of Observations Number Of Observations Number Of Observations 2358
Average Median Median
Surplus To NPW Ratio 3.51 0.93 0.93
2 Year U/W Results 108.2 105.5 105.5
Liability And AB To Total NPW 34.78 37.33 37.33
Personal Property Auto PD To Total NPW 47.19 48.21 48.21
Asset Portfolio In Govt Bonds 68.37 68.77 68.77
NPW (1000 Cdn) 113,440 32,753 32,753
9
Distribution of Capital Levels

• Capital level measure by Surplus / NPW
• 5 had capital levels below 0.33, 7 had capital
  levels 10.
• Firms with higher Surplus / NPW more likely to be
  mutual insurers.

Firm characteristics Surplus / NPW lt1 Surplus / NPW gt 1
Statutory Assets 94.2 mil 59.2 mil
2 year avg u/w ratio 106.54 109.92
10
Determinants of Capitalization

• Amount of capital a firm should carry depends on
• Probability of insolvency.
• Agency costs.
• Asymmetric information / growth opportunities.
• Product market interactions.

11
What Are Determinants of Capitalization?
Explanatory Variable Expected Relationship Coefficient
Regional Diversity 0.405
Product Diversity -0.503
Reinsurance Usage - 0.002
Var. of Past Experience -0.014
Firm Size - -2.561
Canadian Insurer -3.197
2 Year U/W Ratio 0.0063
Investment Risk Ratio 0.010
Claims Settlement Length - 0.0029
Mutual Insurer or - 2.512
Growth Prospects -0.015
Liability and AB 0.738
Personal Property and Auto PD - 0.0025
R2 33.4
12
Capital Holdings Conclusions

• Most of variability explained by size
• Possible interpretations
• Firms determine capital holdings by adding a
  margin to the regulatory requirements rather than
  on the basis of risk characteristics.
• US market is significantly different from
  Canadian market.

13
Introduction of MCT

• Timeline
• Trial basis for 2001 and 2002.
• Implementation in 2003.
• Goals
• Harmonize solvency requirements across provinces.
• Capital neutral across industry.
• Align capital holdings with firm risk.
• Evaluate risk based on both asset and liability
  holdings.

14
Level of Capital Holdings and MCT
Explanatory Variable Expected Relationship Coefficient
Regional Diversity 0.339
Product Diversity -0.551
Reinsurance Usage - 0.0019
Var. of Past Experience -0.0139
Firm Size - -2.606
Canadian Insurer -3.349
2 Year U/W Ratio 0.0082
Investment Risk Ratio 0.010
Claims Settlement Length - 0.0024
Mutual Insurer or - 2.533
Growth Prospects -0.016
Liability and AB 0.716
Personal Property and Auto PD - 0.003725
Test Period Indicator 0 0.794
Implementation Period Indicator 0 1.408
R2 33.7
15
Level of Holdings MCT Conclusions

• Positive coefficient for implementation period
  suggests that capital holdings have increased.
• Cannot reject hypothesis that there is no
  difference between implementation period and test
  period indicator.
• Cannot reject hypothesis that capital holdings
  increases as a response to 9/11 and NOT impending
  MCT test.

16
MCT and Firm Risk

• Do firms hold greater capital since 2003 because
  firm risk has changed?
• Introduce interaction effects for risk
  characteristics and implementation period.
• Results
• Implementation variable becomes insignificant.
• No change in significance of other risk
  characteristics.
• No interaction effects are significant at 5
  level.
• At 10 level, cross effect of Herfindahl index by
  region and implementation is significant and
  negative.

17
MCT, Asset and Liability Risk

• Are capital holdings aligned with asset and
  liability risk?
• Liability risk
• Firms that u/w liability and automobile AB should
  hold more capital.
• Firms that u/w personal property and automobile
  physical damage should hold less capital.
• Asset risk given below

Asset Class Percentage of Book Value held as Reserve
Cash 0
Government Bonds 0
Commercial Bonds 0.5 to 8 depending on maturity and grade.
Mortgage Loans 4 to 8 depending on residential versus commercial
Preferred Shares 4 to 15 depending on grade of shares
Common Shares 15
18
MCT, Asset and Liability Risk
Explanatory Variable Expected Relationship Coefficient Interaction Effect Coefficient
Firm Size - -2.833 0.7263
Proportion of NPW from Liability AB 0.791 -0.04170
Proportion of NPW from Auto Damage and Personal Property - -0.0025 -0.00253
Proportion of Assets as Govt Bonds 0 0.1079 0.1079
Proportion of Assets as Comm. Bonds -0.1036 -0.04687
Proportion of Assets as Mortgage Loans 0.5730 -0.3175
Proportion of Assets as Preferred Shares 0.1363 -0.0297
Proportion of Assets as Common Shares 0.1165 -0.0950
Implementation Period Indicator 0 0.7936 0.7263
R2 33.3
19
Asset and Liability Risk

• Firm size still explains bulk of variability in
  surplus holdings.
• There is some alignment between portfolio risk
  and amount of surplus held.

20
Conclusions

• First long term study into capital holdings of
  Canadian p/c insurers ? more work is needed.
• Risk characteristics do not greatly influence
  capital holdings of Canadian insurers (as opposed
  to U.S. experience).
• Firm size is most relevant indicator of surplus
  holdings.
• Surplus holdings have increased since the
  introduction of MCT (but may be related to 9/11).
• MCT does not appear to do a better job of
  aligning capital holdings with firm risk.

21
Questions?