R/viz_forecasts_cpi_wpi.R

In MattCowgill/macrodashboard: Shiny dashboard for Australian macroeconomic data

viz_forecasts_cpi_wpi <- function(data = load_data(),
                                  arg1 = NULL,
                                  arg2 = NULL) {
  cpi <- data$cpi_8 %>%
    filter(.data$series_id == "A3604511X") %>%
    select(.data$date, .data$value) %>%
    mutate(series = "cpi")

  wpi <- data$wpi_1 %>%
    filter(.data$series_id == "A83895396W") %>%
    select(.data$date, .data$value) %>%
    mutate(series = "wpi")

  actuals <- bind_rows(cpi, wpi) %>%
    mutate(type = "actual")

  # Load RBA forecasts
  raw_forecasts <- data$rba_forecasts
  raw_forecasts <- filter(
    raw_forecasts,
    .data$forecast_date ==
      max(.data$forecast_date)
  )

  smp_date <- raw_forecasts %>%
    pull(.data$forecast_date) %>%
    unique()

  forecasts <- raw_forecasts %>%
    filter(series %in% c("wpi_change", "underlying_annual_inflation")) %>%
    select(.data$date, .data$series, .data$value) %>%
    mutate(series = case_when(
      grepl("wpi", .data$series) ~ "wpi",
      grepl("inflation", .data$series) ~ "cpi",
      TRUE ~ NA_character_
    )) %>%
    mutate(type = "forecast")

  # Combine actual and forecast
  combined <- forecasts %>%
    bind_rows(actuals) %>%
    filter(date >= as.Date("2005-01-01"))

  # Title
  max_cpi_forecast <- max(forecasts$value[forecasts$series == "cpi"])

  title <- case_when(
    max_cpi_forecast < 2 ~
    "Inflation is forecast to remain below the RBA's target",
    max_cpi_forecast < 2.5 ~
    "Inflation is forecast to be in the bottom half of the RBA's target band",
    max_cpi_forecast < 3 ~
    "Inflation is forecast to be in the top half of the RBA's target band",
    max_cpi_forecast > 3 ~
    "Inflation is forecast to be above the RBA's target band"
  )

  latest_text <- paste0(
    "The latest CPI data is for ", format(max(cpi$date), "%b %Y"), "; ",
    "the latest WPI data is for ", format(max(wpi$date), "%b %Y"), "; ",
    "the latest RBA forecast was issued in ", format(
      max(raw_forecasts$forecast_date),
      "%b %Y"
    ), "."
  )

  # Graph
  forecast_band <- summarise(forecasts,
    min_date = min(.data$date),
    max_date = max(.data$date)
  )

  combined %>%
    mutate(series = case_when(
      series == "cpi" ~ "Inflation",
      series == "wpi" ~ "Wages growth",
      TRUE ~ .data$series
    )) %>%
    ggplot(aes(x = .data$date, y = .data$value, col = .data$type)) +
    geom_rect(
      data = forecast_band,
      inherit.aes = FALSE,
      aes(
        xmin = .data$min_date, xmax = .data$max_date,
        ymin = -Inf, ymax = Inf
      ),
      fill = grattantheme::grattan_lightorange,
      size = 0,
      alpha = 0.1
    ) +
    geom_text(
      data = tibble(
        series = "Inflation",
        x = c(
          stats::median(combined$date),
          mean(c(
            forecast_band$min_date,
            forecast_band$max_date
          ))
        ),
        y = c(3.15, 4.5),
        type = c("actual", "forecast")
      ),
      aes(x = .data$x, y = .data$y, label = tools::toTitleCase(.data$type)),
      # inherit.aes = FALSE,
      size = 16 / .pt
    ) +
    geom_line() +
    facet_wrap(~ .data$series) +
    scale_colour_manual(values = c(
      "black",
      grattantheme::grattan_lightorange
    )) +
    grattan_y_continuous(
      limits = c(0, 5),
      labels = function(x) paste0(x, "%")
    ) +
    theme_grattan() +
    theme(axis.title = element_blank()) +
    labs(
      title = title,
      subtitle = "Inflation and wages growth - actual and the RBA's forecasts (per cent per year)",
      caption = paste0(
        "Notes: Inflation refers to the trimmed mean. Wages growth is the Wage Price Index. Both series are seasonally adjusted.",
        latest_text,
        "Source: ABS Consumer Price Index; ABS Wage Price Index; RBA Statement on Monetary Policy ",
        lubridate::month(smp_date, label = TRUE, abbr = FALSE), " ", lubridate::year(smp_date)
      )
    )
}