Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
43 Cards in this Set
- Front
- Back
- 3rd side (hint)
Normal TVaR shortcut |
u + σ[(Φ(Ζp)/(1-p] Where Φ(Zp) = (1/sqrt(2pi))•e^-((x^2)/2 |
|
|
Lognormal TVaR |
EX•[Φ(σ-zp)/(1-p) Where Φ is just Z(x) |
|
|
For Bayesian estimate of the expected number of claims and other expected values, what do you have to multiply the posterior distribution by? |
The expected value |
|
|
Formula for indicated relativity? |
1-LER |
|
|
What is the indicated relativity for a deductible of 2 times the basic level deductible b for X~ Exponential(θ)? |
(EX-E[X^2b])/(EX-E[X^b])
(It’s the continuous version of 1-LER) |
|
|
Cov[Xi,Xj] = ? if i is not equal to j |
Cov[Xi,Xj] = a |
|
|
What is another name for the unconditional variance of X |
VarX |
|
|
In terms of Buhlmann credibility , Total Variance = ? |
v+a |
|
|
For finding the written premium for a calendar year, what’s the strategy to solve? |
Only add up the written premiums for the policies that were written in the calendar year (they started in the given calendar year) (Also remember that written premium = earned premium + unearned premium)- so we don’t have to do any multiplication |
|
|
u is what in the ultimate formula for insurance? |
The policy limit (not the maximum covered loss, so be very careful here) |
|
|
What is the formula for the maximum covered loss, m? |
u/α + d
Be very careful not to mix the policy limit and maximum covered loss up |
|
|
PGF formula |
E[z^X] like E[2^n] So anytime you see something with E[2^N] think Poisson PGF
Poisson PGF = e^λ(t-1) |
|
|
Variance of the Payments Per Loss |
E[(X^u]^2] - (E[X^u])^2 |
|
|
First formula for the expected value of a policy limit, E[(X-u)]? |
Integral of x•f(x) from 0 to u + u• S(u) Follows naturally from the Payment per loss with a policy limit Y^L |
|
|
2nd formula for E[(X-u)^k]
(Uses survival function) |
Integral of S(x) from 0 to u |
|
|
Expected value of a bonus formula |
u- E[(X^u)] (If there is a constant, just pull it out) |
|
|
Expected value of a bonus B, EB if B= 500,000(0.7-R)/(3) |
(1/3)• 350,000- E[X^350,000] |
|
|
Which method for selecting loss development factors results in the highest reserve? All-year arithmetic average, 3 year arithmetic average, 2 year arithmetic average, or 1 year arithmetic average? |
All year arithmetic average
(This is always true, so no calculations are needed) |
|
|
Which results in a higher reserve, the All year arithmetic average, or the volume weighted average method? |
You have to calculate them both, but in that one problem the volume weighted average resulted in higher reserves |
|
|
True or False? If X |
False |
|
|
True or False? If X |
True |
|
|
How do you solve for p in this equation? Z((1+p)/2))=1.28 |
Find Z(1.28) from the Z table (=0.8997) Then solve for p in this equation (p+1)/2= 0.8997 |
|
|
Log likelihood for the Weibull distribution with parameters t and θ with n number of losses? (use for likelihood ratio test) |
nln(t) - ntln(θ) + (t-1)• Σln(Xi) - θ^-t • ΣXi^t |
|
|
SFP covers direct loss from fire and lighting, plus at least one of the following coverages? |
• Personal coverage, such as for dwellings • Commercial Coverage, such as general property and multiple locations • Increased covered perils, such as extended coverage for perils of vandalism or malicious mischief • Increased covered loss, such as additional living expenses, rental value, rental income, leasehold interests, demolition expenses, consequential loss or damage, replacement costs, business interruption losses, profits and commissions losses, and extra expenses |
|
|
Formula for ILF |
LAS(u)/LAS(b) u is the increased limit, b is the increased limit |
|
|
When they say penalized log likelihood they mean find what? |
The AIC or SBC/BIC |
|
|
When they say continuously compounded, use what? |
Use e |
|
|
When they say annual compounded rate, use what? |
(1+i) ^ t |
|
|
Formula for losses incurred to date |
Losses Paid to Date + case reserves |
|
|
What do you multiply the incurred losses paid to date by to develop to ultimate losses? |
Use the incurred loss factors (not the paid loss factors) |
A
|
|
The insurance company expects 1000 claims per year. The company reinsured the layer of loss from 2,000 to 10,000 with a reinsurance company. Calculate the expected payment per year for the reinsurance company as a result of the claims from the insurance company |
Solve this equation: 1000[E(X^10,000) - E[(X^2,000)] |
|
|
Quota share treaty |
A reinsurance company will pay for a portion of the total losses |
|
|
Surplus share reinsurance |
A reinsurance company pays for a portion of the losses greater than the retention limit + the retention limit |
|
|
For the alternate method, do you do frequency and severity together or separately? |
Do them separately (so do the loss development factors separately before combining) |
|
|
Closure rate |
# of closed claims/ # of open claims |
|
|
Formula for probability within an interval using Survival functions |
S(a) - S(b) |
|
|
Probability of an interval using CDF’s |
F(b)- F(a) |
|
|
Are rehabilitative and habilitative services and devices an essential benefit under the ACA? |
Yes, they are an essential benefit. |
|
|
Are mental health and substance use disorder services an essential benefit under the ACA? |
Yes, they are an essential benefit under the ACA |
|
|
For aggregate probabilities, when do you multiply by 2? |
When the amounts are different. Ex: 2• P(X=3)•P(X=4)
But when the amounts are the same you don’t multiply by 2 Ex: P(X=3)• P(X=3) |
|
|
True or False: The chain- ladder method is often used on a new line of business? |
False But the expected loss ratio is often used on a new line of business since it does not require claims experience |
|
|
True or False: the expected loss ratio will always produce a larger reserve than the chain- ladder method? |
False |
|
|
For data preparation for ratemaking, we always assume uniformly written policies are issued at the what of the month? |
Midpoint |
|