HistoricalDamages.Rmd

---
title: "Historical hurricane damage data from ICAT"
output: html_document
editor_options: 
  chunk_output_type: console
---

Data from https://www.icatdamageestimator.com/faq#4q
Could update with https://www.reinsurancene.ws/insurance-industry-losses-events-data/

Read in the data only through 2017
```{r}
PerLandfall.df <- readr::read_csv(here::here("data", "stormData.csv")) |>
  dplyr::rename(Name = `STORM NAME`, 
                LandfallDate = `LANDFALL DATE`,
                DamageRank = `DAMAGE RANK`, 
                CurrentDamage2022 = `CURRENT DAMAGE ($ 2022)`,
                BaseDamageDollars = `BASE DAMAGE ($)`,
                LandfallState = `LANDFALL STATE`,
                Category = `CATEGORY`,
                WindsAtLandfall_mph = `WINDS(MPH)`)

PerLandfall.df <- PerLandfall.df |>
  dplyr::mutate(Date = lubridate::parse_date_time(LandfallDate,
                                                  orders = 'mdy'),
                Year = lubridate::year(Date),
                YearName = paste0(Year, Name)) |>
  dplyr::filter(Year >= 1987)
```

Insured damages (losses) are .5 of the total economic damages.
```{r}
PerStorm.df <- PerLandfall.df |>
  dplyr::mutate(InsuredLosses2022 = .5 * CurrentDamage2022) |>
  dplyr::group_by(YearName) |>
  dplyr::summarise(Year = dplyr::last(Year),
                   Name = dplyr::last(Name),
                   PerStormInsuredLosses2022 = sum(InsuredLosses2022))
```

Tranches
```{r}
Z <- PerStorm.df$PerStormInsuredLosses2022

quantile(Z, probs = seq(0, 1, 1/7))
```