CF exam2 Flashcards
why would actual > expected using CC
Random fluctuation
The beginning IBNR that was estimated was too small
three steps to reinsurance loss reserving
Partition the reinsurance portfolio into reasonably homogeneous exposure groups that are relatively consistent over time with respect to mix of business
Analyze the historical development patterns. If possible, consider individual case reserve development and the emergence of IBNR claims separately
Estimate the future development. If possible, estimate the bulk reserves for IBNER and pure IBNR separately
confidence interval for reserves using lognormal
σ2=ln(1+(s.e.(R)/R)2)
u=ln(R)-σ2/2
CI: exp(u +/- zσ)
why can you just add reported loss to get CI for cumulative losses
Based on the chain-ladder assumptions, all of the variability inherent in the ultimate losses is attributed to the reserve. The claims reported to date are considered known and fixed. Thus, we can simply shift the confidence interval for the reserve by the claims reported to date to obtain the confidence interval for the ultimate losses
iterative for BF, calculating f’s
process and parameter variances of the Cape Cod method and the LDF method
Process variance – the Cape Cod method may or may not produce a higher process variance than the LDF method
Parameter variance – the Cape Cod method produces a lower parameter variance than the LDF method since it requires fewer parameters
parsimony
Since this model is creating a parameter for each accident year and development period (excluding the first development period which does not have a parameter), it tends to overfit. Hence, it does not have enough parsimony. To increase the parsimony, we could start with a basic GLM that only has one parameter in each direction. Then, we could continue adding parameters until we achieve randomly scattered residuals around y = 0 as well as a development pattern that looks like a smoothed version of the standard chainladder factors
graph of the Teng/Perkins method
slope of the first line segment (i.e. P DLD1) represents
the relationship between incremental premiums and incremental losses at the first retro adjustment. Since the factor is greater than 1, this means that we expect to earn more than $1 of premium for every $1 of loss. However, this is misleading due to the inclusion of the basic premium.
di↵erence between bootstrapping paid data and bootstrapping incurred data
Bootstrapping paid data provides a distribution of possible outcomes for total unpaid claims
Bootstrapping incurred data provides a distribution of possible outcomes for IBNR
Explain the benefit of bootstrapping the incurred data triangle.
Bootstrapping incurred data leverages the case reserves to better predict the ultimate claims. This improves estimates, while still focusing on the payment stream for measuring risk
abnormal earnings: BV to use
if you are given minimum capital required, use this for BV for each CY
four organizational details that need to be addressed when developing an internal model
Reporting relationship
Resource commitment
Inputs and outputs
Initial scope
reporting relationship
modeling team reporting line, solid line vs. dotted line reporting
recommended course of action: the reporting line for the internal model team is less important than ensuring they report to a leader who is fair
resource commitment
mix of skill set (actuarial, UW, communication, etc.), full time vs. part time
recommended course of action: since an internal model implementation is considered a new competency, it’s best to transfer internal employees or hire external employees for full-time positions
