valuations.R

library(rvest)
library(ggvis)

trim <- function (x) gsub("^\\s+|\\s+$", "", x)

to_number <- function(string) {
  string <- trim(string)
  len <- nchar(string)
  last_char <-substr(string, len, len)
  suppressWarnings(
    if (last_char == "B") {
      number <- as.numeric(substr(string, 1, len - 1)) * 1E9
      } else if (last_char == "M") {
        number <- as.numeric(substr(string, 1, len - 1)) * 1E6
      } else if (last_char == "K") {
        number <- as.numeric(substr(string, 1, len - 1)) * 1E3
      } else if (last_char == "%") {
        number <- as.numeric(substr(string, 1, len - 1)) / 100
      } else
        number <- as.numeric(string)
  )
  return(number)
}

symbol_key_stats <- function(symbol) {
  url <- paste0(
    "http://finance.yahoo.com/q/ks?s=",
    symbol,
    "+Key+Statistics"
  )
  page <- html(url)
  tables <- page %>% html_nodes("table")
  stats <- tables[[9]] %>% html_table()
  for(i in c(12, 14, 16, 18, 20, 22)) {
    table <- tables[[i]] %>% html_table()
    table <- table[-1,]
    stats <- rbind(stats, table)
  }
  df <- data.frame(
    symbol = symbol,
    market_cap = to_number(stats[1,2]),
    ps = to_number(stats[6,2]),
    ev = to_number(stats[2,2]),
    ev_rev = to_number(stats[8,2]),
    ev_ebitda = to_number(stats[9,2]),
    profit_margin = to_number(stats[12,2]),
    operating_margin = to_number(stats[13,2]),
    revenue = to_number(stats[16,2]),
    revenue_growth = to_number(stats[18,2]),
    earnings_growth = to_number(stats[23,2]),
    gross_profit = to_number(stats[19,2])
  )
  return(df)
}

category_stats <- function(symbols, category_name) {
  n = length(symbols)
  stats = symbol_key_stats(symbols[1])
  stats$category = category_name
  if(n > 1) {
    for(i in 2:n) {
      line = symbol_key_stats(symbols[i])
      line$category = category_name
      stats = rbind(stats, line)
    }
  }
  return(stats)
}

baskets <- function() {
  saas_symbols = c("CRM", "HUBS", "ZEN", "SHOP", "BOX")
  cloud_telco_symbols = c("RNG", "FIVN", "EGHT", "JCOM")
  telco_symbols = c("T", "VZ", "TMUS", "S", "TDC.CO", "VOD")
  telco_software_symbols = c("DOX", "CNSI", "CVG","BSFT")
  saas_data = category_stats(saas_symbols, "SaaS")
  cloud_telco_data = category_stats(cloud_telco_symbols, "Cloud Telco")
  telco_data = category_stats(telco_symbols, "Telco")
  telco_software_data = category_stats(telco_software_symbols, "Telco Software")
  baskets = rbind(saas_data, cloud_telco_data, telco_data, telco_software_data)
  return(baskets)
}

stocks = baskets()
plot = stocks %>% ggvis(x = ~revenue_growth, y = ~ev_rev) %>% 
  group_by(category) %>%
  layer_text(text:=~symbol, dx:=-5, dy:=-5, stroke = ~category) %>% 
  layer_model_predictions(model = "lm", stroke = ~category) %>%
  layer_points(fill = ~category) %>% 
  add_axis("x", title = "Revenue Growth", format ="%") %>% 
  add_axis("y", title = "Revenue Multiple (EV/Revn)")
print(plot)