B.4. Continuous Changes

Question 1

Reasons for adjusting continuous changes

Answer 1

To ensure historical data reflects mix of business and levels of social and economic inflation expected in the future period

Question 2

Possible data sources for trends

Answer 2

Historical insurer data
Industry data
Economic data

Question 3

Data adjustments before trending

Answer 3

Adjust for large events, one-time changes, seasonality

Question 4

Advantages for using each type of premium to determine premium trends

Answer 4

Earned - We are trending earned premium anyway

Written - more recent data, as changes in average written premium will ultimately show up as changes in average earned premium.

Question 5

Advantages between AY and CY data to determine loss trends

Answer 5

AY - better match between losses and exposures

CY - no need to estimate ultimate losses

Question 6

Why frequency and severity are typically analyzed separately

Answer 6

They can change for different reasons

Question 7

Two other alternative methods for loss trends

Answer 7

1. Using econometric or GL methods

2. Using incremental CY data by AY

Question 8

Advantages between using paid and reported loss data to determine loss trends

Answer 8

Paid – not subject to changes in case reserving practices

Reported – incorporates more recent information (case could ultimately become paid)

Question 9

Two reasons why excess severity trends are greater than basic or total limits trends

Answer 9

1. For losses above basic limits, trend is entirely in excess layer
2. Losses just under basic limit are pushed into excess layer by the trend, creating new excess losses

Question 10

Three ways to calculate trend from historical data

Answer 10

1. Take an average of the percent changes
2. Fit a line to the data (assumes constant change for each time period)
3. Fit exponential curve (assumes constant percent change between time periods)

Question 11

Information needed to determine average earned date of a future policy period

Answer 11

1. Future rate change effective date
2. Length of time rates are expected to be in effect
3. Policy term late

Question 12

Assumptions commonly made in determining trend periods

Answer 12

1. Policies are written uniformly over time

2. Premiums earned uniformly and losses occur uniformly

Question 13

Two ways to perform first step of a two-step trend

Answer 13

1. Adjust historical period level to be equal to latest level:
   Current trend factor = latest average WP at current / historical average EP at current
2. Trend historical period level to the latest time period (more common for loss trends)

Question 14

Overlap fallacy between development and trending

Answer 14

Development brings data from each historical period to its ultimate level, while trending reflects the difference in ultimate levels from one period to the next.
Loss development makes sure future policy is priced to cover ultimate losses. Loss trend makes sure ultimate losses are at cost levels corresponding to future policy period.