Experience summaries

After experience data has been prepared for analysis, the next step is to summarize results. The actxps package’s workhorse function for summarizing termination experience is exp_stats(). This function returns an exp_df object, which is a type of data frame containing additional attributes about the experience study.

At a minimum, an exp_df includes:

Optionally, an exp_df can also include:

To demonstrate this function, we’re going to use a data frame containing simulated census data for a theoretical deferred annuity product that has an optional guaranteed income rider. Before exp_stats() can be used, we must convert our census data into exposure records using the expose() function1. In addition, let’s assume we’re interested in studying surrender rates, so we’ll pass the argument target_status = 'Surrender' to expose().

library(actxps)
library(dplyr)

exposed_data <- expose(census_dat, end_date = "2019-12-31",
                       target_status = "Surrender")

The exp_stats() function

To use exp_stats(), pass it a data frame of exposure-level records, ideally of type exposed_df (the object class returned by the expose() family of functions).

exp_stats(exposed_data)
#> Experience study results
#> 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 1 × 4
#>   n_claims claims exposure  q_obs
#>      <int>  <int>    <dbl>  <dbl>
#> 1     2869   2869  132634. 0.0216

The results show us that we specified no groups, which is why the output data is a single row. In addition, we can see that we’re looking at surrender rates through the end of 2019, which exp_stats() inferred from exposed_data.

The number of claims (n_claims) is equal to the number of “Surrender” statuses in exposed_data. Since we didn’t specify any weighting variable, the amount of claims (claims) equals the number of claims.

(amount <- sum(exposed_data$status == "Surrender"))
#> [1] 2869

The total exposure (exposure) is equal to the sum of the exposures in exposed_data. Had we specified a weighting variable, this would be equal to the sum of weighted exposures.

(sum_expo <- sum(exposed_data$exposure))
#> [1] 132633.9

Lastly, the observed termination rate (q_obs) equals the amount of claims divided by the exposures.

amount / sum_expo
#> [1] 0.02163097

Grouped data

If the data frame passed into exp_stats() is grouped using dplyr::group_by(), the resulting output will contain one record for each unique group.

In the following, exposed_data is grouped by policy year before being passed to exp_stats(). This results in one row per policy year found in the data.

exposed_data |> 
  group_by(pol_yr) |> 
  exp_stats()
#> Experience study results
#> 
#>  Groups: pol_yr 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 15 × 5
#>    pol_yr n_claims claims exposure   q_obs
#>     <int>    <int>  <int>    <dbl>   <dbl>
#>  1      1      102    102   19252. 0.00530
#>  2      2      160    160   17715. 0.00903
#>  3      3      124    124   16097. 0.00770
#>  4      4      168    168   14536. 0.0116 
#>  5      5      164    164   12916. 0.0127 
#>  6      6      152    152   11376. 0.0134 
#>  7      7      164    164    9917. 0.0165 
#>  8      8      190    190    8448. 0.0225 
#>  9      9      181    181    6960. 0.0260 
#> 10     10      152    152    5604. 0.0271 
#> 11     11      804    804    4390. 0.183  
#> 12     12      330    330    2663. 0.124  
#> 13     13       99     99    1620. 0.0611 
#> 14     14       62     62     872. 0.0711 
#> 15     15       17     17     268. 0.0634

Multiple grouping variables are allowed. Below, the presence of an income guarantee (inc_guar) is added as a second grouping variable.

exposed_data |> 
  group_by(inc_guar, pol_yr) |> 
  exp_stats()
#> Experience study results
#> 
#>  Groups: inc_guar, pol_yr 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 30 × 6
#>    inc_guar pol_yr n_claims claims exposure   q_obs
#>    <lgl>     <int>    <int>  <int>    <dbl>   <dbl>
#>  1 FALSE         1       56     56    7720. 0.00725
#>  2 FALSE         2       92     92    7103. 0.0130 
#>  3 FALSE         3       67     67    6447. 0.0104 
#>  4 FALSE         4      123    123    5799. 0.0212 
#>  5 FALSE         5       97     97    5106. 0.0190 
#>  6 FALSE         6       96     96    4494. 0.0214 
#>  7 FALSE         7       92     92    3899. 0.0236 
#>  8 FALSE         8      103    103    3287. 0.0313 
#>  9 FALSE         9       87     87    2684. 0.0324 
#> 10 FALSE        10       60     60    2156. 0.0278 
#> # ℹ 20 more rows

Target status

The target_status argument of exp_stats() specifies which status levels count as claims in the experience study summary. If the data passed to exp_stats() is an exposed_df object that already has a specified target status (via a prior call to expose()), then this argument is not necessary because the target status is automatically inferred.

Even if the target status exists on the input data, it can be overridden. However care should be taken to ensure that exposure values in the data are appropriate for the new status.

Using the example data, a total termination rate can be estimated by including both death and surrender statuses in target_status. To ensure exposures are accurate, an adjustment is made to fully expose deaths prior to calling exp_stats()2.

exposed_data |> 
  mutate(exposure = ifelse(status == "Death", 1, status)) |> 
  group_by(pol_yr) |> 
  exp_stats(target_status = c("Surrender", "Death"))
#> Experience study results
#> 
#>  Groups: pol_yr 
#>  Target status: Surrender, Death 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 15 × 5
#>    pol_yr n_claims claims exposure  q_obs
#>     <int>    <int>  <int>    <dbl>  <dbl>
#>  1      1      290    290    20199 0.0144
#>  2      2      325    325    18754 0.0173
#>  3      3      292    292    17054 0.0171
#>  4      4      329    329    15602 0.0211
#>  5      5      329    329    13946 0.0236
#>  6      6      334    334    12371 0.0270
#>  7      7      297    297    10869 0.0273
#>  8      8      340    340     9510 0.0358
#>  9      9      308    308     7953 0.0387
#> 10     10      260    260     6489 0.0401
#> 11     11      894    894     6505 0.137 
#> 12     12      398    398     3753 0.106 
#> 13     13      131    131     2135 0.0614
#> 14     14       89     89     1306 0.0681
#> 15     15       23     23      544 0.0423

Weighted results

Experience studies often weight output by key policy values. Examples include account values, cash values, face amount, premiums, and more. Weighting can be accomplished by passing the name of a weighting column to the wt argument of exp_stats().

Our sample data contains a column called premium that we can weight by. When weights are supplied, the claims, exposure, and q_obs columns will be weighted. If expected termination rates are supplied (see below), these rates and A/E values will also be weighted.3

exposed_data |> 
  group_by(pol_yr) |> 
  exp_stats(wt = 'premium')
#> Experience study results
#> 
#>  Groups: pol_yr 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Weighted by: premium 
#> 
#> # A tibble: 15 × 8
#>    pol_yr n_claims claims  exposure   q_obs  .weight  .weight_sq .weight_n
#>     <int>    <int>  <dbl>     <dbl>   <dbl>    <dbl>       <dbl>     <dbl>
#>  1      1      102  83223 25312812. 0.00329 26301746 60742993234     19995
#>  2      2      160 170058 23352461. 0.00728 24275265 56232848027     18434
#>  3      3      124 123554 21246765. 0.00582 22201817 51746834383     16806
#>  4      4      168 176751 19270856. 0.00917 20200019 47142441689     15266
#>  5      5      164 173273 17228978. 0.0101  18134795 42887920479     13618
#>  6      6      152 163034 15246504. 0.0107  16192950 38828949428     12067
#>  7      7      164 153238 13328777. 0.0115  14159437 34291451913     10541
#>  8      8      190 174200 11476433. 0.0152  12346124 30121310640      9130
#>  9      9      181 187337  9546247. 0.0196  10420172 25781142118      7591
#> 10     10      152 157603  7707062. 0.0204   8543150 20882643976      6185
#> 11     11      804 856379  6093168. 0.141    6783273 16219955859      4897
#> 12     12      330 383055  3883534. 0.0986   4525027 11191462577      3093
#> 13     13       99 123357  2450266. 0.0503   2891573  7075934189      1937
#> 14     14       62  75534  1339240. 0.0564   1821026  4655661820      1182
#> 15     15       17  19168   401169. 0.0478    783146  1944755748       510

Expected values and A/E ratios

As common metric in experience studies is the actual-to-expected, or A/E ratio.

\[ A/E\ ratio = \frac{observed\ value}{expected\ value} \]

If the data passed to exp_stats() has one or more columns containing expected termination rates, A/E ratios can be calculated by passing the names of these columns to the expected argument.

Let’s assume we have two sets of expected rates. The first set is a vector that varies by policy year. The second set is either 1.5% or 3.0% depending on whether the policy has a guaranteed income benefit. First, we need to attach these assumptions to our exposure data. We will use the names expected_1 and expected_2. Then we pass these names to the expected argument when we call exp_stats().

In the output, 4 new columns are created for expected rates and A/E ratios.

expected_table <- c(seq(0.005, 0.03, length.out = 10), 0.2, 0.15, rep(0.05, 3))

# using 2 different expected termination assumption sets
exposed_data <- exposed_data |>
  mutate(expected_1 = expected_table[pol_yr],
         expected_2 = ifelse(exposed_data$inc_guar, 0.015, 0.03))

exp_res <- exposed_data |>
  group_by(pol_yr, inc_guar) |>
  exp_stats(expected = c("expected_1", "expected_2"))


exp_res |> 
  select(pol_yr, inc_guar, q_obs, expected_1, expected_2, 
         ae_expected_1, ae_expected_2)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1, expected_2 
#> 
#> # A tibble: 30 × 7
#>    pol_yr inc_guar   q_obs expected_1 expected_2 ae_expected_1 ae_expected_2
#>     <int> <lgl>      <dbl>      <dbl>      <dbl>         <dbl>         <dbl>
#>  1      1 FALSE    0.00725    0.005        0.03          1.45          0.242
#>  2      1 TRUE     0.00399    0.005        0.015         0.798         0.266
#>  3      2 FALSE    0.0130     0.00778      0.03          1.67          0.432
#>  4      2 TRUE     0.00641    0.00778      0.015         0.824         0.427
#>  5      3 FALSE    0.0104     0.0106       0.03          0.985         0.346
#>  6      3 TRUE     0.00591    0.0106       0.015         0.560         0.394
#>  7      4 FALSE    0.0212     0.0133       0.03          1.59          0.707
#>  8      4 TRUE     0.00515    0.0133       0.015         0.386         0.343
#>  9      5 FALSE    0.0190     0.0161       0.03          1.18          0.633
#> 10      5 TRUE     0.00858    0.0161       0.015         0.532         0.572
#> # ℹ 20 more rows

As noted above, if weights are passed to exp_stats() then A/E ratios will also be weighted.

exposed_data |>
  group_by(pol_yr, inc_guar) |>
  exp_stats(expected = c("expected_1", "expected_2"), 
            wt = "premium") |> 
  select(pol_yr, inc_guar, q_obs, expected_1, expected_2, 
         ae_expected_1, ae_expected_2)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1, expected_2 
#>  Weighted by: premium 
#> 
#> # A tibble: 30 × 7
#>    pol_yr inc_guar   q_obs expected_1 expected_2 ae_expected_1 ae_expected_2
#>     <int> <lgl>      <dbl>      <dbl>      <dbl>         <dbl>         <dbl>
#>  1      1 FALSE    0.00471    0.005        0.03          0.942         0.157
#>  2      1 TRUE     0.00235    0.005        0.015         0.470         0.157
#>  3      2 FALSE    0.0105     0.00778      0.03          1.36          0.351
#>  4      2 TRUE     0.00513    0.00778      0.015         0.660         0.342
#>  5      3 FALSE    0.00737    0.0106       0.03          0.698         0.246
#>  6      3 TRUE     0.00479    0.0106       0.015         0.453         0.319
#>  7      4 FALSE    0.0174     0.0133       0.03          1.30          0.579
#>  8      4 TRUE     0.00377    0.0133       0.015         0.283         0.252
#>  9      5 FALSE    0.0146     0.0161       0.03          0.907         0.487
#> 10      5 TRUE     0.00710    0.0161       0.015         0.441         0.473
#> # ℹ 20 more rows

Credibility

If the credibility argument is set to TRUE, exp_stats() will produce an estimate of partial credibility under the Limited Fluctuation credibility method (also known as Classical Credibility) assuming a binomial distribution of claims.4

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(credibility = TRUE) |> 
  select(pol_yr, inc_guar, claims, q_obs, credibility)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 30 × 5
#>    pol_yr inc_guar claims   q_obs credibility
#>     <int> <lgl>     <int>   <dbl>       <dbl>
#>  1      1 FALSE        56 0.00725       0.192
#>  2      1 TRUE         46 0.00399       0.173
#>  3      2 FALSE        92 0.0130        0.246
#>  4      2 TRUE         68 0.00641       0.211
#>  5      3 FALSE        67 0.0104        0.210
#>  6      3 TRUE         57 0.00591       0.193
#>  7      4 FALSE       123 0.0212        0.286
#>  8      4 TRUE         45 0.00515       0.172
#>  9      5 FALSE        97 0.0190        0.254
#> 10      5 TRUE         67 0.00858       0.210
#> # ℹ 20 more rows

Under the default arguments, credibility calculations assume a 95% confidence of being within 5% of the true value. These parameters can be overridden using the conf_level and cred_r arguments, respectively.

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(credibility = TRUE, conf_level = 0.98, cred_r = 0.03) |> 
  select(pol_yr, inc_guar, claims, q_obs, credibility)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 30 × 5
#>    pol_yr inc_guar claims   q_obs credibility
#>     <int> <lgl>     <int>   <dbl>       <dbl>
#>  1      1 FALSE        56 0.00725      0.0969
#>  2      1 TRUE         46 0.00399      0.0876
#>  3      2 FALSE        92 0.0130       0.125 
#>  4      2 TRUE         68 0.00641      0.107 
#>  5      3 FALSE        67 0.0104       0.106 
#>  6      3 TRUE         57 0.00591      0.0976
#>  7      4 FALSE       123 0.0212       0.145 
#>  8      4 TRUE         45 0.00515      0.0867
#>  9      5 FALSE        97 0.0190       0.128 
#> 10      5 TRUE         67 0.00858      0.106 
#> # ℹ 20 more rows

If expected values are passed to exp_stats() and credibility is set to TRUE, then the output will also contain credibility-weighted expected values:

\[ q^{adj} = Z^{cred} \times q^{obs} + (1-Z^{cred}) \times q^{exp} \] where,

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(credibility = TRUE, expected = "expected_1") |> 
  select(pol_yr, inc_guar, claims, q_obs, credibility, adj_expected_1, 
         expected_1, ae_expected_1)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1 
#> 
#> # A tibble: 30 × 8
#>    pol_yr inc_guar claims   q_obs credibility adj_expected_1 expected_1
#>     <int> <lgl>     <int>   <dbl>       <dbl>          <dbl>      <dbl>
#>  1      1 FALSE        56 0.00725       0.192        0.00543    0.005  
#>  2      1 TRUE         46 0.00399       0.173        0.00482    0.005  
#>  3      2 FALSE        92 0.0130        0.246        0.00905    0.00778
#>  4      2 TRUE         68 0.00641       0.211        0.00749    0.00778
#>  5      3 FALSE        67 0.0104        0.210        0.0105     0.0106 
#>  6      3 TRUE         57 0.00591       0.193        0.00966    0.0106 
#>  7      4 FALSE       123 0.0212        0.286        0.0156     0.0133 
#>  8      4 TRUE         45 0.00515       0.172        0.0119     0.0133 
#>  9      5 FALSE        97 0.0190        0.254        0.0168     0.0161 
#> 10      5 TRUE         67 0.00858       0.210        0.0145     0.0161 
#> # ℹ 20 more rows
#> # ℹ 1 more variable: ae_expected_1 <dbl>

Confidence intervals

If conf_int is set to TRUE, exp_stats() will produce lower and upper confidence interval limits for the observed termination rate.

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(conf_int = TRUE) |> 
  select(pol_yr, inc_guar, q_obs, q_obs_lower, q_obs_upper)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 30 × 5
#>    pol_yr inc_guar   q_obs q_obs_lower q_obs_upper
#>     <int> <lgl>      <dbl>       <dbl>       <dbl>
#>  1      1 FALSE    0.00725     0.00544     0.00920
#>  2      1 TRUE     0.00399     0.00286     0.00520
#>  3      2 FALSE    0.0130      0.0104      0.0156 
#>  4      2 TRUE     0.00641     0.00490     0.00801
#>  5      3 FALSE    0.0104      0.00807     0.0129 
#>  6      3 TRUE     0.00591     0.00446     0.00746
#>  7      4 FALSE    0.0212      0.0176      0.0250 
#>  8      4 TRUE     0.00515     0.00366     0.00675
#>  9      5 FALSE    0.0190      0.0153      0.0229 
#> 10      5 TRUE     0.00858     0.00666     0.0106 
#> # ℹ 20 more rows

If no weighting variable is passed to wt, confidence intervals will be constructed assuming a binomial distribution of claims. However, if a weighting variable is supplied, a normal distribution for aggregate claims will be assumed with a mean equal to observed claims and a variance equal to:

\[ Var(S) = E(N) \times Var(X) + E(X)^2 \times Var(N) \]

Where S is the aggregate claim random variable, X is the weighting variable assumed to follow a normal distribution, and N is a binomial random variable for the number of claims.

The default confidence level is 95%. This can be changed using the conf_level argument. Below, tighter confidence intervals are constructed by decreasing the confidence level to 90%.

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(conf_int = TRUE, conf_level = 0.9) |> 
  select(pol_yr, inc_guar, q_obs, q_obs_lower, q_obs_upper)
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#> 
#> # A tibble: 30 × 5
#>    pol_yr inc_guar   q_obs q_obs_lower q_obs_upper
#>     <int> <lgl>      <dbl>       <dbl>       <dbl>
#>  1      1 FALSE    0.00725     0.00570     0.00894
#>  2      1 TRUE     0.00399     0.00303     0.00494
#>  3      2 FALSE    0.0130      0.0108      0.0152 
#>  4      2 TRUE     0.00641     0.00518     0.00773
#>  5      3 FALSE    0.0104      0.00838     0.0126 
#>  6      3 TRUE     0.00591     0.00466     0.00725
#>  7      4 FALSE    0.0212      0.0181      0.0243 
#>  8      4 TRUE     0.00515     0.00389     0.00641
#>  9      5 FALSE    0.0190      0.0159      0.0221 
#> 10      5 TRUE     0.00858     0.00691     0.0104 
#> # ℹ 20 more rows

If expected values are passed to expected, the output will also contain confidence intervals around any actual-to-expected ratios.

exposed_data |> 
  group_by(pol_yr, inc_guar) |>
  exp_stats(conf_int = TRUE, expected = "expected_1") |> 
  select(pol_yr, inc_guar, starts_with("ae_"))
#> Experience study results
#> 
#>  Groups: pol_yr, inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1 
#> 
#> # A tibble: 30 × 5
#>    pol_yr inc_guar ae_expected_1 ae_expected_1_lower ae_expected_1_upper
#>     <int> <lgl>            <dbl>               <dbl>               <dbl>
#>  1      1 FALSE            1.45                1.09                1.84 
#>  2      1 TRUE             0.798               0.572               1.04 
#>  3      2 FALSE            1.67                1.34                2.01 
#>  4      2 TRUE             0.824               0.630               1.03 
#>  5      3 FALSE            0.985               0.764               1.22 
#>  6      3 TRUE             0.560               0.422               0.707
#>  7      4 FALSE            1.59                1.32                1.88 
#>  8      4 TRUE             0.386               0.275               0.506
#>  9      5 FALSE            1.18                0.948               1.42 
#> 10      5 TRUE             0.532               0.413               0.660
#> # ℹ 20 more rows

Lastly, if credibility is TRUE and expected values are passed to expected, confidence intervals will also be calculated for any credibility-weighted termination rates.

Miscellaneous

Summary method

As noted above, the result of exp_stats() is an exp_df object. If the summary() function is applied to an exp_df object, the data will be summarized again and return a higher level exp_df object.

If no additional arguments are passed, summary() returns a single row of aggregate results.

summary(exp_res)
#> Experience study results
#> 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1, expected_2 
#> 
#> # A tibble: 1 × 8
#>   n_claims claims exposure  q_obs expected_1 expected_2 ae_expected_1
#>      <int>  <int>    <dbl>  <dbl>      <dbl>      <dbl>         <dbl>
#> 1     2869   2869  132634. 0.0216     0.0242     0.0209         0.892
#> # ℹ 1 more variable: ae_expected_2 <dbl>

If additional variable names are passed to the summary() function, then the output will group the data by those variables. In our example, if pol_yr is passed to summary(), the output will contain one row per policy year.

summary(exp_res, pol_yr)
#> Experience study results
#> 
#>  Groups: pol_yr 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1, expected_2 
#> 
#> # A tibble: 15 × 9
#>    pol_yr n_claims claims exposure   q_obs expected_1 expected_2 ae_expected_1
#>     <int>    <int>  <int>    <dbl>   <dbl>      <dbl>      <dbl>         <dbl>
#>  1      1      102    102   19252. 0.00530    0.005       0.0210         1.06 
#>  2      2      160    160   17715. 0.00903    0.00778     0.0210         1.16 
#>  3      3      124    124   16097. 0.00770    0.0106      0.0210         0.730
#>  4      4      168    168   14536. 0.0116     0.0133      0.0210         0.867
#>  5      5      164    164   12916. 0.0127     0.0161      0.0209         0.788
#>  6      6      152    152   11376. 0.0134     0.0189      0.0209         0.707
#>  7      7      164    164    9917. 0.0165     0.0217      0.0209         0.763
#>  8      8      190    190    8448. 0.0225     0.0244      0.0208         0.920
#>  9      9      181    181    6960. 0.0260     0.0272      0.0208         0.955
#> 10     10      152    152    5604. 0.0271     0.03        0.0208         0.904
#> 11     11      804    804    4390. 0.183      0.2         0.0208         0.916
#> 12     12      330    330    2663. 0.124      0.15        0.0200         0.826
#> 13     13       99     99    1620. 0.0611     0.05        0.0197         1.22 
#> 14     14       62     62     872. 0.0711     0.05        0.0195         1.42 
#> 15     15       17     17     268. 0.0634     0.05        0.0191         1.27 
#> # ℹ 1 more variable: ae_expected_2 <dbl>

Similarly, if inc_guar is passed to summary(), the output will contain a row for each unique value in inc_guar.

summary(exp_res, inc_guar)
#> Experience study results
#> 
#>  Groups: inc_guar 
#>  Target status: Surrender 
#>  Study range: 1900-01-01 to 2019-12-31 
#>  Expected values: expected_1, expected_2 
#> 
#> # A tibble: 2 × 9
#>   inc_guar n_claims claims exposure  q_obs expected_1 expected_2 ae_expected_1
#>   <lgl>       <int>  <int>    <dbl>  <dbl>      <dbl>      <dbl>         <dbl>
#> 1 FALSE        1601   1601   52123. 0.0307     0.0235      0.03          1.31 
#> 2 TRUE         1268   1268   80511. 0.0157     0.0247      0.015         0.637
#> # ℹ 1 more variable: ae_expected_2 <dbl>

Column names

As a default, exp_stats() assumes the input data frame uses the following naming conventions:

These default names can be overridden using the col_exposure and col_status arguments.

For example, if the status column was called curr_stat in our data, we could write:

exposed_data |> 
  exp_stats(col_status = "curr_stat")

Applying exp_stats to a non-exposed_df data frame

exp_stats() can still work when given a non-exposed_df data frame. However, it will be unable to infer certain attributes like the target status and the study dates. For target status, all statuses except the first level are assumed to be terminations. Since this may not be desirable, a warning message will appear informing what statuses were assumed to be terminated.

not_exposed_df <- data.frame(exposed_data)

exp_stats(not_exposed_df)
#> Warning: ✖ No target status was provided.
#> ℹ Death, Surrender was assumed.
#> Experience study results
#> 
#>  Target status: Death, Surrender 
#>  Study range:  to  
#> 
#> # A tibble: 1 × 4
#>   n_claims claims exposure  q_obs
#>      <int>  <int>    <dbl>  <dbl>
#> 1     4639   4639  132634. 0.0350

If target_status is provided, no warning message will appear.

exp_stats(not_exposed_df, target_status = "Surrender")
#> Experience study results
#> 
#>  Target status: Surrender 
#>  Study range:  to  
#> 
#> # A tibble: 1 × 4
#>   n_claims claims exposure  q_obs
#>      <int>  <int>    <dbl>  <dbl>
#> 1     2869   2869  132634. 0.0216

Limitations

The exp_stats() function only supports termination studies. It does not contain support for transaction studies or studies with multiple changes from an active to an inactive status. For information on transaction studies, see vignette("transactions").


  1. See vignette('exposures') for more information on creating exposure records.↩︎

  2. This adjustment is not necessary on surrenders because the expose() function previously did this for us.↩︎

  3. When weights are supplied, additional columns are created containing the sum of weights, the sum of squared weights, and the number of records. These columns are used for re-summarizing the data (see the “Summary method” section on this page).↩︎

  4. See Herzog, Thomas (1999). Introduction to Credibility Theory for more information on Limited Fluctuation Credibility.↩︎