R/make_price_impact_data.R

In Martin-Summer-1090/syslosseval: Provide data and functions to support the analysis in Systemic Loss Evaluation

#' make_price_impact_data
#'
#' This function prepares the price impact parameters volatility and average
#' daily volume for the bond exposures in DE, ES, IT, FR, JP, GB, US and Rest of
#' the world for a given stress test base year. Daily volatility for a given
#' year is computed on the daily index log returns of the base year.
#'
#' @param data_idx sovereign bond indices
#' @param data_adv average daily volume data
#' @param base_year the year for which the analysis is done
#'
#' @return A dataframe with variables: Country ADV_sov_bonds, sigma_sov_bonds.
#' @export
#'
#' @importFrom rlang .data
#' @importFrom magrittr %>%
#'
#' @examples
#' make_price_impact_data(sovereign_bond_indices, average_daily_volume_sovereign, 2015)
make_price_impact_data <- function(data_idx, data_adv, base_year) {

  # compute the log returns of bond indices for DE, ES, FR, GB, IT, JP, US, Total based on the daily log returns
  # from the base year fo the analysis

  bond_log_returns <- data_idx %>%
    dplyr::filter(stringr::str_detect(.data$Date, as.character(base_year))) %>%
    dplyr::group_by(.data$Country) %>%
    dplyr::mutate(log_returns = c(diff(log(.data$Value)), NA)) %>%
    dplyr::ungroup() %>%
    stats::na.omit()

  # make a matrix of returns

  returns_matrix <- bond_log_returns %>%
    dplyr::select(.data$Country, .data$Date, .data$log_returns) %>%
    tidyr::pivot_wider(names_from = .data$Country, values_from = .data$log_returns) %>%
    stats::na.omit() %>%
    dplyr::select(-(.data$Date))

  # compute country volatilities

  Sig <- stats::cov(as.matrix(returns_matrix)) %>%
    diag()

  Sigma <- sqrt(Sig)

  Names <- names(Sigma)
  Values <- unname(Sigma)

  vola <- tibble::tibble(Country = Names, Volatility = Values)


  # extract average daily volume from the base_year

  ADV <- data_adv %>%
    dplyr::filter(.data$Year == base_year)


  # combine the impact data

  impact_data <- dplyr::left_join(ADV, vola, by = "Country") %>%
    dplyr::arrange(.data$Country)

  impact_data
}