README.md

In jfkingiii/layers: Reinsurance Layers and Metrics

layers

The layers package provides a simple library for doing common reinsurance calculations outside of Analyze Re.

Installation

You can install the development version of layers from GitHub with:

# install.packages("devtools")
devtools::install_github("jfkingiii/layers")

If using Windows, Rtools should be installed: https://cran.r-project.org/bin/windows/Rtools/

Example

Use the included sample YELT, yelt_test, to try out some of the layers functions.

library(layers)

# Look at the help files
help(package = layers)

# Look at the AAL by line
AAL(yelt_test)
#>                Loss
#> CHC         4872434
#> HOSPITAL   31243978
#> MEDCHOICE  15236345
#> PHYSICIANS 50882467

# Create some layers and get their metrics
options(scipen = 999)
test_layer <- layer(4000000, 1000000, 1, "yelt_test",
                    lobs=c("PHYSICIANS", "CHC", "MEDCHOICE"))
gross_layer <- layer(UNLIMITED, 0, 1, "yelt_test",
                     lobs=c("PHYSICIANS", "CHC", "MEDCHOICE", "HOSPITAL"))
# Aggregate limits and deductibles are optional and default to UNLIMITED and 0.
# Notice the `Sign` value which can be changed using the `minus` function.
# This allows combining layers by adding and subtracting when building portfolios.

test_layer
#>                                   Value
#> Limit:                        4,000,000
#> Attachment:                   1,000,000
#> Participation:                    1.000
#> Loss set:                     yelt_test
#> LOBS:          PHYSICIANS CHC MEDCHOICE
#> Sign:                                 +
gross_layer
#>                                            Value
#> Limit:                                 UNLIMITED
#> Attachment:                                    0
#> Participation:                             1.000
#> Loss set:                              yelt_test
#> LOBS:          PHYSICIANS CHC MEDCHOICE HOSPITAL
#> Sign:                                          +
summary(test_layer)
#>                                   Value
#> Limit:                        4,000,000
#> Attachment:                   1,000,000
#> Participation:                    1.000
#> Loss set:                     yelt_test
#> LOBS:          PHYSICIANS CHC MEDCHOICE
#> Sign:                                 +
#> 
#>                Value
#> Mean:      6,782,625
#> StdDev:    3,381,071
#> VaR 25:   13,889,280
#> VaR 100:  16,049,317
#> VaR 250:  18,392,032
#> tVaR 25:  15,582,691
#> tVaR 100: 17,798,589
#> tVaR 250: 19,018,259
summary(gross_layer)
#>                                            Value
#> Limit:                                 UNLIMITED
#> Attachment:                                    0
#> Participation:                             1.000
#> Loss set:                              yelt_test
#> LOBS:          PHYSICIANS CHC MEDCHOICE HOSPITAL
#> Sign:                                          +
#> 
#>                 Value
#> Mean:     102,235,225
#> StdDev:    16,176,233
#> VaR 25:   134,029,327
#> VaR 100:  148,713,927
#> VaR 250:  152,354,245
#> tVaR 25:  141,894,334
#> tVaR 100: 151,802,890
#> tVaR 250: 154,059,363

# Test some OEPs
VaR(gross_layer, 5, "OEP")
#> [1] 11000000
VaR(gross_layer, 10, "OEP")
#> [1] 15000000
VaR(test_layer, 15, "OEP")
#> [1] 4000000
VaR(test_layer, 20, "OEP")
#> [1] 4000000

# Try a layer with aggregate parameters
agg_layer <- layer(4000000, 1000000, 1, "yelt_test",
                   lobs=c("PHYSICIANS","CHC","MEDCHOICE"),
                   agg_attachment = 4000000, agg_limit = 12000000)
summary(agg_layer)
#>                                    Value
#> Limit:                         4,000,000
#> Attachment:                    1,000,000
#> Participation:                     1.000
#> Loss set:                      yelt_test
#> LOBS:           PHYSICIANS CHC MEDCHOICE
#> Agg Attachment:                4,000,000
#> Agg Limit:                    12,000,000
#> Sign:                                  -
#> 
#>                Value
#> Mean:      3,037,403
#> StdDev:    2,995,537
#> VaR 25:    9,889,280
#> VaR 100:  12,000,000
#> VaR 250:  12,000,000
#> tVaR 25:  11,133,044
#> tVaR 100: 12,000,000
#> tVaR 250: 12,000,000

# Portolio expected losses
layer1 <- layer(100000, 100000, 1, "yelt_test", lobs="PHYSICIANS")
layer2 <- layer(100000, 200000, 1, "yelt_test", lobs="PHYSICIANS")
layer3 <- layer(100000, 300000, 1, "yelt_test", lobs="PHYSICIANS")
P <- portfolio(layer1, layer2, layer3)

expected(layer1) + expected(layer2) + expected(layer3)
#> [1] 19472924
expected(P)
#> [1] 19472924

# Use a portfolio as an argument to portfolio
ceded <- portfolio(layer1, layer2, layer3)
net <- portfolio(gross_layer, minus(ceded))
expected(gross_layer) - expected(ceded)
#> [1] 82762300
expected(net)
#> [1] 82762300
summary(net)
#>                                            Value
#> Limit:                                 UNLIMITED
#> Attachment:                                    0
#> Participation:                             1.000
#> Loss set:                              yelt_test
#> LOBS:          PHYSICIANS CHC MEDCHOICE HOSPITAL
#> Sign:                                          +
#> 
#>                     Value
#> Limit:            100,000
#> Attachment:       100,000
#> Participation:      1.000
#> Loss set:       yelt_test
#> LOBS:          PHYSICIANS
#> Sign:                   -
#> 
#>                     Value
#> Limit:            100,000
#> Attachment:       200,000
#> Participation:      1.000
#> Loss set:       yelt_test
#> LOBS:          PHYSICIANS
#> Sign:                   -
#> 
#>                     Value
#> Limit:            100,000
#> Attachment:       300,000
#> Participation:      1.000
#> Loss set:       yelt_test
#> LOBS:          PHYSICIANS
#> Sign:                   -
#> 
#> 
#>                 Value
#> Mean:      82,762,300
#> StdDev:    14,631,910
#> VaR 25:   112,527,136
#> VaR 100:  123,037,841
#> VaR 250:  128,296,989
#> tVaR 25:  119,841,123
#> tVaR 100: 127,537,772
#> tVaR 250: 131,842,474

jfkingiii/layers documentation built on July 21, 2020, 8:57 p.m.