Chapter 2 Flashcards
empirical probability distributions provide
a mutually exclusive, collectively exhaustive list of outcomes
What is the coefficient of variation
coef of v = standard deviation / mean
higher c of v means less predictability
used to compare data sets of different mean and s.d.
Probabilities are expressed numerically as
Numerically, as a fraction, a percentage, or a decimal
1st standard deviation above or below mean
34.13
2nd standard deviation above or below mean
13.59
two s. d. in either direction
47.72
two s. d. in both directions
95.44
probability that is developed based on actual experience
empirical
hazard analysis is a method
That identifies conditions that increase the frequency or severity of loss.
To accurately analyze loss exposures using data on past losses, the data should be
relevant, complete, consistent, and organized.
If a fire damaged a building in 2002 and it cost $200,000 to repair the building in 2002, then $200,000 represents the value of the loss in
Nominal Dollars
nominal value / dollar
is an economic value expressed in monetary terms (that is, in units of a currency).
real value / dollar
aka
constant value / dollar
is a value that has been adjusted from a nominal value to remove the effects of general price level price changes over time.
Current value / dollar
is a term describing income in the year in which a person, household, or family receives it. For example, the income someone received in 1989 unadjusted for inflation is in current dollars.
Higher coeff of variation shows
the losses are less predictable.
Under The Prouty Approach of analyzing loss exposures, the four broad categories of loss frequency and the three loss severity categories are
Subjective
Most losses are reduce/prevent, retain, or avoid
TRANSFER: High severity-low frequency;
significant or slight severity-slight frequency
One way of presenting a continuous probability distribution is to divide the distribution
into a countable number of bins.
Liability entries on an organization’s balance sheet are particularly useful to the risk management professional for exploring
Obligations such as mortgage payments
The weighted average of all the possible outcomes of a theoretical probability distribution is the
Expected Value
Probability analysis would be most reliable for projecting losses in an organization
with substantial volume of data on past losses
When a risk management professional knows the mode of a distribution, it allows the professional to focus on the outcomes that are the most
Common
Claim adjuster Phyllis is reviewing fire losses. The dollar amounts of the losses are $2,000, $8,000, $10,000, $15,000, and $20,000. The median loss has an adjusted value of
$10,000
In an array of historical and adjusted auto physical damage losses the adjusted amount column is the historical loss amount adjusted to
Current year dollars using a price index.
with respect to the completeness of past loss data when estimating future loss exposures
What constitutes complete loss data depends largely on the nature of the loss exposure being considered.
a benefit of conducting compliance reviews
They can help minimize or avoid liability loss exposures.
The mean loss value on each account is roughly the same, and Lucy determines that the two accounts exhibit similar underwriting characteristics. All else being equal, Lucy would be best advised to select the account with the
lower standard deviation - IE losses are more predictable since they will be within a lesser range
An empirical probability distribution is based on
historical data
EX: tenant has contractually agreed to indemnify him in the event that the Lucas is found legally liable for a bodily injury claim that occurs on the leased premises.
hold-harmless agreement
Probabilities deduced solely from historical data may change as
new data are discovered or the environment changes.
new data are discovered or the environment changes.
Auto Accidents per mile driven
When developing loss severity distributions, insurance and risk management professionals should organize loss data by
size of loss
arrange values in a data set by size to determine
median
trying to analyze workers compensation loss frequency by cause of loss should use
discrete probability distribution