RISK 6

micael14

sect_6_loss_data_1113.pdf

Home >Business & Finance homework help >Marketing homework help >RISK 6

PRI:6:11/13 - 1 -

Loss Data Analysis

Learning Objectives

1. Discuss why loss data must be collected and analyzed. (p. 2)

2. Identify the specific types of loss data that should be collected. (p. 8)

3. Discuss why collecting and reviewing exposure data is as important as collecting and reviewing loss data. (p. 9)

4. Discuss the components and statistical credibility of loss data. (p. 12)

5. Explain the types of analyses that use loss data. (p. 16)

6. Describe when to use benchmarking in risk management. (p. 18)

PRI:6:11/13 - 2 -

Learning Objective #1: Discuss why loss data must be collected and analyzed.

I. Why Loss Data is Collected and Analyzed

A. Identify the causes of loss frequency and severity B. Identify trends in loss experience C. Focus management’s attention on the organization’s total

cost of risk D. Evaluate potential costs and benefits of loss control

alternatives to gain support E. Evaluate potential costs and benefits of alternative methods

for financing losses

1. Decide between full insurance and retention 2. Choose deductibles and limits 3. Select a cash flow plan

PRI:6:11/13 - 3 -

F. Establish a basis for allocating premiums and/or loss costs

1. Create incentives or disincentives for loss control actions

2. Establish an objective basis for sharing the total cost of

risk 3. Determine optimum retention level for each location vs.

amount of loss the organization as a whole is capable of retaining

G. Establish a method for evaluating performance

1. Operating units’ management

2. Vendors – carriers, TPA’s, brokers (external)

3. In-house claims adjusters

4. Benchmarking loss experience

5. Employee safety incentive programs

PRI:6:11/13 - 4 -

H. Address product/service development and pricing

1. Include total cost of risk in pricing of products and services

2. Redesign products or services based on expected losses

I. Gain ability to respond to legal or regulatory actions

1. Litigation 2. OSHA Survey of Occupational Injury 3. Consumer Product Safety Commissions 4. Environmental Protection Agency 5. Federal Drug Administration

J. Satisfy insurance underwriting requirements

1. Premium negotiation 2. Determine coverage restrictions and exclusions 3. Set appropriate reserves for loss-sensitive programs 4. Establish collateral amounts, e.g., letter of credit, surety

bond

PRI:6:11/13 - 5 -

II. Sources of Loss Data

A. Internal sources – the organization’s loss experience

1. Accident or incident reports 2. First aid logs 3. OSHA logs 4. Insurance carrier or TPA loss runs 5. Litigation records 6. Accounting entries on financial statements

PRI:6:11/13 - 6 -

Wheel chocks not in place at Unsafe: rear wheels of Conditions truck trailer. Practices

Lift truck is driven into truck trailer Property Damage/ and trailer moves. Near Misses or “Near Miss.” Close Calls

Truck trailer moves and driver jumps Total OSHA to ground and Lost Workday Case sprains ankle.

EXAMPLES

Truck trailer moves and driver jumps to ground Lost Workday Case and breaks leg.

Accidents

RECOGNIZED HAZARDS 50 50

10 10

2 2

1 1

Incidents

Serious Accident

PRI:6:11/13 - 7 -

B. External sources – other organization’s loss experience

1. Industry associations 2. Insurance company loss runs (company-wide, not

insured-specific) 3. Bureau of Labor Statistics incident rates (NAICS code) 4. Bureau of Transportation Statistics 5. National Safety Council (Accident Facts –

Occupational and Non-Occupational) 6. Risk and Insurance Management Society Cost of Risk

Survey 7. National Council on Compensation Insurance

PRI:6:11/13 - 8 -

Learning Objective #2: Identify the specific types of loss data that should be collected.

III. Specific Types of Loss Data that should be collected

A. Category of loss – property damage, auto accident, industrial injury, injury from product

B. Date and time of loss – year, date, day of week, shift or time

of day C. Claimant – name, date of hire, occupation, shift D. Location – division, plant, department, operation point E. Hazard – noise level, floor surface, lack of protection,

weather F. Cause – fall from height, repetitive motion, inhalation,

lifting, twisting G. Type – sprain/strain, laceration, disease, water damage, auto

physical damage H. Body part – left wrist, lower back, eye I. Management – supervisor, team leader

PRI:6:11/13 - 9 -

Learning Objective #3: Discuss why collecting and reviewing exposure data is as important as collecting and reviewing loss data.

IV. Exposure Data

A. Loss data is often more valuable when losses are indexed or compared to appropriate key units of measurement for an organization.

Example

Workers compensation losses (an increase of 25%)

12/31/X1 $ 400,000 12/31/X2 $ 500,000

At first glance, the loss experience appears to be worsening.

Workers compensation payroll (exposure) (an increase of 50%)

12/31/X1 $10,000,000 12/31/X2 $15,000,000

After comparing the loss experience to the exposure, the loss experience appears to be improving. Losses have increased by 25%, and the exposure has increased by 50%. Barring a 50% increase in the pay level, it appears there are simply more workers; thus, the increased exposure.

PRI:6:11/13 - 10 -

B. Common exposure bases for indexing or comparisons

1. Revenue, gross receipts, net income 2. Units of production 3. Payroll, broken down by classification and state 4. Headcount, hours worked 5. Number of vehicles, broken down by type, use,

geographic area or territory 6. Annual mileage 7. Square footage of building, area, frontage 8. Property values 9. Number of surgeries, outpatient visits, births 10. Number of admissions, bleachers, and events 11. Any other appropriate base that appears to have a close

causal relationship to the number or severity of losses

PRI:6:11/13 - 11 -

Skills Application Scenario #11 – Exposures Identify appropriate exposure measures for evaluating the following types of losses. Workers Compensation: Wages for employees vary dramatically between the five resorts in the continental U.S. because of unionization, availability of workers in the labor markets, and cost of living (the Atlantic City location has the highest average wage rates, with the French Lick location having the lowest). Benefit structures are very similar between the five states with operations. Premises and Product Liability: The total area of the resort buildings is 2,130,000 square feet (s.f.). The breakdown is as follows: Hotel and conference center 1,325,000 s.f. Restaurants 200,000 s.f. Pro shops 50,000 s.f. Spas 30,000 s.f. Shopping 60,000 s.f. Ski chalets 10,000 s.f. Entertainment venues 150,000 s.f. Casinos 300,000 s.f. Corporate office 5,000 s.f. The total area of the golf courses is 3,000 acres, and the area of the ski runs is 500 acres. Ten lakes are on the golf courses for a total of 250 acres. Two chair lifts, two t-bar lifts, and one rope lift are on the “bunny” slope. The entertainment venues seat 8,000 in total.

PRI:6:11/13 - 12 -

Learning Objective #4: Discuss the components and statistical credibility of loss data.

V. Evaluating and Ensuring the Credibility of Loss Data

A. Components of credible loss data

1. Completeness

a. All losses are included and reported, including losses not covered by insurance because of deductibles or exclusions or limit

b. Enough loss data (frequency); a rule of thumb is at least 5 years of data, preferably10+ years and at least 30 data points per year

c. Adequate details about each data record – date of loss, cause of loss, person causing loss, person injured, type of loss, dollar value of loss, etc.

d. Basis for paid and open reserve amounts – ALAE, IBNR, defense costs, etc.

2. Consistency

a. Same types of data provided for each data record – type, cause, time, claimant name, length of employment, etc.

b. Same policy year, accident year, calendar year

c. Same recording methodology – differences between carriers, TPA’s

d. Same definitions of types of injuries, perils, hazards, etc.

PRI:6:11/13 - 13 -

3. Integrity

a. Reliability of data and accuracy of input – errors, omissions, duplications

b. Prompt reporting and current data

c. Accuracy of loss reserving

4. Relevance – data that will yield information on matters of concern to the organization

a. Discontinued operations – divestiture, discontinued operations, transferring exposures to a third party

b. Acquired operations – entire versus partial acquisition

c. Commingling of data of diverse operations

d. Data not relevant to the loss – extraneous data

5. Ability to be organized into useful formats

PRI:6:11/13 - 14 -

B. Statistical credibility is enhanced when:

1. A substantial number of losses exist 2. Losses are extended over a sufficient period of time

(one year is not enough, regardless of how many losses)

3. Minimum variability exists in frequency and severity of losses

4. Stable operations exist over the time period

C. Identification of changes in exposures that may impact loss

frequency and/or severity

1. Introduction of new product or service

2. Equipment, materials, or work process

3. Acquisition, divestiture, merger, restructuring

4. Legal and regulatory matters, including statutory benefits

5. Social and economic environment, including inflation

6. Labor and management issues

7. Safety, incentive, or awards programs

PRI:6:11/13 - 15 -

8. Deductible or retention levels

9. Insurance carrier or third-party administrator

10. Insurance coverage (exclusions, extensions)

11. Demographic changes

D. Other considerations

1. Security of confidential data

2. Cost of collecting data

3. Cost of maintaining data

4. Relevance and usefulness of data

PRI:6:11/13 - 16 -

Learning Objective #5: Explain the types of analyses that use loss data.

VI. Types of Analyses that use Loss Data

A. Frequency/severity rankings (examples)

1. Number of losses in each severity range

2. Number and cost of losses at each location, product line, type of vehicle

3. Highest cost hazards, causes of loss, or types of injuries

4. Frequency or severity of injury related to length of service, age, or other demographic data

5. Frequency or severity of injury related to shift, time of day, day of week

B. Evaluation of time intervals (examples)

1. Time between occurrence and reporting

2. Time between occurrence and closing of claim file

3. Average number of days of lost time or restricted duty

4. Likelihood of return-to-work based on number of lost days

PRI:6:11/13 - 17 -

C. Loss projections

1. Projection of losses based on average number of losses and average value of a loss

Average # of losses x average $ of loss = expected annual losses

Commentary

The “average,” or arithmetic mean, may not be the appropriate value to use in projections. In some cases, use of one of the other two measures of central tendency, the median or the mode, may be more appropriate.

2. Triangulation analysis

a. Estimation of frequency or cost of losses from time of initial reserve to final claim settlement

b. Estimation of number and cost of Incurred But

Not Reported losses (IBNR)

c. Estimation of timing of payout on claims

3. Actuarial studies

PRI:6:11/13 - 18 -

Learning Objective #6: Describe when to use benchmarking in risk management.

VII. Benchmarking

A. Benchmarking is a systematic way of continuously comparing an organization’s performance against others at a given time or against itself over a time period.

1. Comparison to “best in industry” or competitors 2. Comparison to “best in class” or those recognized as

performing certain functions at a high level 3. Comparison of “self” over time

B. Benchmarking should be used when

1. A baseline program has been established 2. Internal trending and comparisons are needed 3. Improvement opportunities are sought

C. Benefits of benchmarking

1. Indicates continuous improvement 2. Enhances creativity and thinking 3. Identifies and prioritizes areas for improvement from

trending

PRI:6:11/13 - 19 -

D. Common pitfalls of benchmarking

1. Belief that being lower (or higher) is better

2. Implying more precision or accuracy than what really exists

3. Inappropriate comparison groups – “apples to oranges”

4. Inconsistent comparison data – comparing data that varies from group to group

5. Insufficient comparison group population – not enough data

6. Focus on one causation or factor

7. One-time, point-in-time comparisons in a dynamic environment

8. “Slightly” different comparison data can make the comparison “slightly” invalid

9. Statistically invalid comparisons

10. Statistically massaged data

11. Unknown data – data from a questionable source

PRI:6:11/13 - 20 -

Commentary Examples of Misleading Statistics Average Lottery Winner Takes Home $4,000!! The lottery says that the average winner took home $4,000. Below is how the prizes were paid out: Grand Prize (one winner) $100,000 First Prize (two winners) $ 50,000 each Second Prize (fifty winners) $ 240 each Total paid out $212,000 If you read the lottery promotional material, bought a lottery ticket and saw that you had won a prize, how much do you reasonably think you would have won?

Which Chart Would Your Board Prefer?

Widget Growth

20X1 20X2 20X3 20X4

Widget Growth

0 10

20X1 20X2 20X3 20X4

PRI:6:11/13 - 21 -

Water Causes Crime!! Studies have shown that 99.9% of all crimes are committed within 24 hours of drinking water. This is an example of trying to correlate two unrelated statistics into cause and effect. More World War I Head Injuries Suffered by Soldiers Wearing Helmets than Without!! At the beginning of the war, soldiers wore cloth hats. When helmets were introduced, head injuries skyrocketed. The information missing from earlier statistics was that fatalities from head wounds were not considered injuries; they were fatalities. What was not recognized in the statistics was the apparent increase in the number of head injuries because the soldiers wearing helmets did not die.

PRI:6:11/13 - 22 -

Review of Learning Objectives

1. Discuss why loss data must be collected and analyzed. (p. 2)

2. Identify the specific types of loss data that should be collected. (p. 8)

3. Discuss why collecting and reviewing exposure data is as important as collecting and reviewing loss data. (p. 9)

4. Discuss the components and statistical credibility of loss data. (p. 12)

5. Explain the types of analyses that use loss data. (p. 16)

6. Describe when to use benchmarking in risk management. (p. 18)