R/TaxBenefit.R

In miraisolutions/SmaRP: SmaRP: Smart Retirement Planning

#' @title getTaxAmount
#'
#' @rdname getTaxAmount
#'
#' @description This function uses 2 main sources for tax data.
#' At Kanton and Gemeinde level, the source is taxburden.list.
#' At federal level, we use the official taxrate table (BundessteueTabelle) and we try to aproximate the taxable income.
#'
#' @details
#' This function assumes the following objects on the global enviornment
#'  * PLZGemeinden (includes Kirchensteuer)
#'  * taxburden.list
#'  * BundessteueTabelle
#'  * BVGcontributionratesPath, BVGcontributionrates
#'  * MaxBVG, MinBVG
#'  * KinderabzugKG
#'  * NBU, maxNBU
#'  * AHL
#'  * ALV, maxALV
#'  * VersicherungsL, VersicherungsV, VersicherungsK
#'  * BerufsauslagenTarif, BerufsauslagenMax, BerufsauslagenMin
#' @seealso swisstax
#'
#' @param Income Annual salary. `Numeric` scalar.
#' @param rate_group A (Single), B (Married), C (Married Double income) `Character`.
#' @param Age Age of the person. `Numeric`
#' @param NChildren Number of children. `Numeric` scalar.
#' @param postalcode Zip code `Character`
#' @param churchtax Y/N `Character` Y/N
#'
#' @import dplyr
#'
#' @return Tax Amount
#'
#' @examples
#' \dontrun{
#' getTaxAmount(Income = 200000, rate_group = "C", Age = 32,
#'              NChildren = 5, postalcode = 8400, churchtax = "Y")
#' }
#' @export
getTaxAmount <- function(Income,
                         rate_group,
                         Age,
                         NChildren,
                         postalcode,
                         churchtax) {

  # Find Kanton and Gemeinde
  Kanton <- subset(PLZGemeinden, PLZ == postalcode, select = "Kanton")[1, 1]
  GDENR <- subset(PLZGemeinden, PLZ == postalcode, "GDENR")[1, 1]
  GDENAME <- subset(PLZGemeinden, PLZ == postalcode, "GDENAME")[1, 1]

  # Get Tarif
  Tarif <- ifelse(rate_group == "C", "DOPMK",
    ifelse(rate_group == "A" & NChildren == 0, "Ledig",
      ifelse(rate_group == "B" & NChildren == 0, "VOK", "VMK")
    )
  )

  DOfactor <- ifelse(Tarif == "DOPMK", 2, 1)

  # Select Tarif, Gemeinde and build Income Cuts
  taxburden <- filter(taxburden.list[[grep(Tarif, names(taxburden.list))]], Gemeindenummer == GDENR)

  # Get taxrate vector associated to one Gemeinde
  idxNumCols <- !grepl("[a-z]", colnames(taxburden))
  IncomeCuts <- gsub("([0-9])\\.([0-9])", "\\1\\2", colnames(taxburden)[idxNumCols]) %>%
    as.numeric()
  taxrate <- taxburden[1, idxNumCols] %>% as.vector()

  # Constrain Income
  Income <- Income %>%
    max(0) %>%
    min(1e+08)

  # Calc adjustIncomeKG
  # 1. Age adjustment because of BVG contributions
  # Tax burden based on the Pensionkassebeitrage from the examples (5%). Therefore, an adjustment factor is applied accordingly.
  AjustBVGContri <- BVGcontributionratesPath %>%
    filter(years == Age) %>%
    transmute(AjustBVGContri = (0.05 - BVGcontributionrates) * (min(Income, MaxBVG) - MinBVG))

  # 2. NChildren ajustment (only for VMK and DOPMK)
  # Tax burden based on 2 kids. Therefore, an adjustment factor is applied accordingly.
  if (Tarif %in% c("DOPMK", "VMK")) {
    OriKinderabzugKG <- sum(KinderabzugKG[row.names(KinderabzugKG) == Kanton, 1:2])
    AjustKinderabzug <- OriKinderabzugKG - sum(KinderabzugKG[row.names(KinderabzugKG) == Kanton, 1:NChildren])
  } else {
    AjustKinderabzug <- 0
  }

  # 3. NBU (not applied on taxburden source)
  NBUanzug <- min(DOfactor * maxNBU, Income * NBU)

  IncomeKG <- Income + AjustKinderabzug + (DOfactor * AjustBVGContri[1, 1]) - NBUanzug

  TaxAmountKGC <- max(0, IncomeKG * (stats::approx(x = IncomeCuts, y = taxrate, IncomeKG)$y) / 100)

  # Church affiliation
  # By default, assumed church affiliation. If not, there's a discount
  if (churchtax != "Y") {
    TaxAmountKGC <- TaxAmountKGC * Kirchensteuer[Kirchensteuer$Kanton == Kanton, "Kirchensteuer"]
  }

  # Get Taxable Federal Income
  TaxableIncomeFederal <- BVGcontributionratesPath %>%
    filter(years == Age) %>%
    mutate(
      DO = ifelse(Tarif == "DOPMK", DOV, 0),
      BVG = DOfactor * (BVGcontributionrates * (min(Income, MaxBVG) - MinBVG)),
      AHL = Income * AHL,
      ALV = min(DOfactor * maxALV, Income * ALV),
      NBU = min(DOfactor * maxNBU, Income * NBU),
      NetSalary = Income - BVG - AHL - ALV - NBU,
      Verheiratet = ifelse(Tarif == "Ledig", 0, Verheiratet),
      Versicherung = ifelse(Tarif == "Ledig", VersicherungsL, VersicherungsV + NChildren * VersicherungsK),
      Beruf = max(DOfactor * BerufsauslagenMin, min(DOfactor * BerufsauslagenMax, NetSalary * BerufsauslagenTarif)),
      Kids = NChildren * Kinder
    ) %>%
    transmute(AjustSalary = NetSalary - Verheiratet - Versicherung - DO - Beruf - Kids)

  TaxAmountFederal <- max(0, lookupTaxAmount(TaxableIncomeFederal, BundessteueTabelle, rate_group) - 251 * NChildren)
  TaxAmount <- TaxAmountFederal + TaxAmountKGC

  return(TaxAmount)
}


#' @title lookupTaxAmount
#'
#' @rdname lookupTaxAmount
#'
#' @description Search the tax amount to be paig given one income on the tax tables.
#' @seealso swisstax
#'
#' @param Income Annual stipend.
#' @param Tabelle Income - Tax rate table at Federal level.
#' @param CivilStatus Marital status.
#'
#' @return Tax amount to be paid.
#' @examples
#' \dontrun{
#' lookupTaxAmount(Income = 100000, Tabelle = BundessteueTabelle, CivilStatus = "A")
#' }
#' @export
lookupTaxAmount <- function(Income, Tabelle, CivilStatus) {

  # Define column to pick
  if (CivilStatus == "A") {
    CivilStatusColumn <- "taxAmountSingle"
  } else {
    CivilStatusColumn <- "taxAmountMarried"
  }

  # Get closest bin
  salary_bins <- Tabelle$I
  nearest_salary <- salary_bins[findInterval(Income, salary_bins)]
  TaxAmount <- Tabelle[Tabelle$I == nearest_salary, CivilStatusColumn]

  return(TaxAmount)
}

#' @title buildTaxBenefits
#'
#' @rdname buildTaxBenefits
#'
#' @description All inputs are scalars. Builds a data frame as long as the years to retirement.
#' Calls 'getTaxAmount()' through 'calcTaxBenefitSwiss()', therefore, it assumes objects on the global enviornment.
#' @seealso swisstax
#'
#' @inheritParams getTaxAmount
#' @param RetirementAge Age of retirement.
#' @template given_bday
#' @template P2
#' @template P3
#' @template salary
#'
#' @import dplyr
#'
#' @return data.frame tax benefits path.
#' @examples
#' \dontrun{buildTaxBenefits(
#'  birthday,
#'  TypePurchase,
#'  P2purchase,
#'  P3purchase,
#'  returnP3,
#'  Salary,
#'  SalaryGrowthRate,
#'  postalcode,
#'  NChildren,
#'  churchtax,
#'  rate_group,
#'  MaxContrTax,
#'  givenday = today("UTC"),
#'  RetirementAge = 65)
#' }
#' @export
buildTaxBenefits <- function(birthday,
                             TypePurchase,
                             P2purchase,
                             P3purchase,
                             returnP3,
                             Salary,
                             SalaryGrowthRate,
                             postalcode,
                             NChildren,
                             churchtax,
                             rate_group,
                             givenday = today("UTC"),
                             RetirementAge) {
  TaxBenefitsPath <- data.frame(Calendar = getRetirementCalendar(birthday, givenday, RetirementAge = RetirementAge))
  ncp <- nrow(TaxBenefitsPath)

  TaxBenefitsPath <- TaxBenefitsPath %>%
    mutate(
      BVGpurchase = calcBVGpurchase(TypePurchase, P2purchase, ncp),
      P3purchase = c(0, rep(P3purchase, ncp - 1)),
      TotalContr = BVGpurchase + P3purchase,
      ExpectedSalaryPath = calcExpectedSalaryPath(Salary, SalaryGrowthRate, ncp),
      TaxableIncome = pmax(ExpectedSalaryPath - pmin(TotalContr, MaxContrTax), 0),
      AgePath = as.integer(sapply(Calendar, calcAge, birthday = birthday)),
      TaxBenefits = calcTaxBenefitSwiss(ExpectedSalaryPath, TaxableIncome, rate_group, AgePath, NChildren, postalcode, churchtax),
      t = buildt(birthday, givenday, RetirementAge = RetirementAge),
      TotalTax = calcAnnuityAcumPath(TaxBenefits, t, returnP3),
      ReturnTax = TotalTax - cumsum(TaxBenefits),
      DirectTax = cumsum(TaxBenefits)
    ) %>%
    select(-c(ExpectedSalaryPath, P3purchase, BVGpurchase, TaxableIncome))

  return(TaxBenefitsPath)
}

#' @title calcTaxBenefitSwiss
#'
#' @rdname calcTaxBenefitSwiss
#'
#' @description Calculates the tax benefits as a difference of the taxes paid with and without retirement contributions.
#' Calls 'getTaxAmount()', therefore, it assumes objects in the global environment.
#' @seealso [getTaxAmount()]
#' @seealso swisstax
#'
#' @param ExpectedSalaryPath Vector of annual salaries until retirement.
#' @param TaxableIncome Vector of annual taxable income until retirement.
#' @inheritParams getTaxAmount
#'
#' @return Single tax benefits (tax relief) of one contribution.
#' @examples
#' \dontrun{
#'   calcTaxBenefitSwiss(ExpectedSalaryPath = seq(90000, 100000, 1000),
#'                     TaxableIncome = seq(88000, 98000, 1000),
#'                     rate_group = "A",
#'                     Age = seq(55, 65),
#'                     NChildren = 0,
#'                     postalcode = 8400,
#'                     churchtax = "Y")
#' }
#' @export
calcTaxBenefitSwiss <- function(ExpectedSalaryPath,
                                TaxableIncome,
                                rate_group,
                                Age,
                                NChildren,
                                postalcode,
                                churchtax) {
  assertthat::are_equal(length(ExpectedSalaryPath), length(TaxableIncome))

  TaxAmountGrossIncome <- sapply(seq_along(ExpectedSalaryPath), function(i) {
    getTaxAmount(ExpectedSalaryPath[i], rate_group, Age[i], NChildren, postalcode, churchtax)
  })

  TaxAmountTaxableIncome <- sapply(seq_along(ExpectedSalaryPath), function(i) {
    getTaxAmount(TaxableIncome[i], rate_group, Age[i], NChildren, postalcode, churchtax)
  })

  TaxBenefits <- TaxAmountGrossIncome - TaxAmountTaxableIncome

  return(TaxBenefits)
}