R/data_fundamentals.R

In GregoryBrownson/EAPR: Empirical Asset Pricing Research Platform

# Implementations of fundamental variable computations. These functions will be internal and should NOT be exported.

#' @importFrom data.table set ':=' as.data.table setnames
#' @importFrom dplyr summarize
#' @importFrom lubridate month

bookEquity <- function(dt) {
    # comp Dependencies: pstkrv, pstkl, pstk, txditc, seq
    
    dt[["pref_stock"]] <- ifelse(is.na(dt$pstkrv), dt$pstkl, dt$pstkrv)
    dt[["pref_stock"]] <- ifelse(is.na(dt$pref_stock), dt$pstk, dt$pref_stock)
    set(dt, which(is.na(dt[["pref_stock"]])), "pref_stock", 0)
    
    set(dt, which(is.na(dt[["txditc"]])), "txditc", 0)
    
    dt[["seq"]] <- ifelse(is.na(dt$seq), dt$bkvlps + dt$upstk, dt$seq)
    dt[["seq"]] <- ifelse(is.na(dt$seq), dt$assets - dt$lt, dt$seq)
    
    dt[["book_equity"]] <- dt$seq + dt$txditc - dt$pref_stock
    set(dt, which(dt[["book_equity"]] <= 0), "book_equity", NaN)
    
    dt[["log_book_equity"]] <- log(dt$book_equity)
    
    return(dt)
}

marketEquity <- function(dt) {
    # crsp Dependencies: prc, shrout
    dt[["market_equity"]] <- abs(dt$price) * dt$shares_out
    
    dt.grouped <- group_by(dt, date, permco)
    
    max.me <- as.data.table(summarize(dt.grouped, sum_me = sum(market_equity), market_equity = max(market_equity)))
    
    dt <- merge(dt, max.me, by = c("date", "permco", "market_equity"))
    
    dt[, `:=`(market_equity = sum_me, sum_me = NULL)]
    
    set(dt, which(dt[["market_equity"]] <= 0), "market_equity", NaN)
    dt[["log_market_equity"]] <- log(dt$market_equity/1000)
    
    rebalance_dates <- sort(unique(dt$rebalance_date))[1:2]
    
    if (rebalance_dates[1] + years(1) == rebalance_dates[2]) {
        dt.dec <- dt[month(date) == 12, .(permno, rebalance_date, market_equity, log_market_equity)]
        setnames(dt.dec, c("market_equity", "log_market_equity"), c("dec_market_equity", "log_dec_market_equity"))
        
        dt.dec[, `:=`(rebalance_date, rebalance_date + years(1))]
        
        dt <- merge(dt, dt.dec, by = c("rebalance_date", "permno"), all.x = TRUE)
    }
    
    size <- dt[date == rebalance_date, .(permno, rebalance_date, market_equity, log_market_equity)]
    size[, `:=`(rebalance_date, rebalance_date + years(1))]
    setnames(size, c("market_equity", "log_market_equity"), c("exp_size", "size"))
    dt <- merge(dt, size, by = c("rebalance_date", "permno"), all.x = TRUE)
    
    return(dt)
}

bookToMarket <- function(dt) {
    # merged
    dt[["book_market"]] <- dt$book_equity/dt$dec_market_equity * 1000
    
    dt[["log_book_market"]] <- log(dt$book_market)
    
    return(dt)
}

assetToBook <- function(dt) {
    # comp Dependencies: at
    
    dt[["asset_book"]] <- dt$assets/dt$book_equity
    
    dt[["log_asset_book"]] <- log(dt$asset_book)
    
    return(dt)
}

assetToMarket <- function(dt) {
    # merged Dependencies: at
    
    dt[["asset_market"]] <- dt$assets/dt$dec_market_equity * 1000
    
    dt[["log_asset_market"]] <- log(dt$asset_market)
    
    return(dt)
}

earningsToPrice <- function(dt) {
    # merged Dependencies: ib, prc
    
    dt[["earnings_price"]] <- dt$earnings/abs(dt$dec_market_equity) * 1000
    
    dt[["earnings_price_indicator"]] <- as.integer(dt$earnings_price < 0)
    
    dt[["positive_earnings_price"]] <- ifelse(dt$earnings_price >= 0, dt$earnings_price, 0)
    
    return(dt)
}

# comp
operatingProfitability <- function(dt) {
    # comp dependencies: ebitda, xint
    dt[["oper_prof"]] <- (dt$ebitda - dt$interest_exp)/dt$book_equity
    
    return(dt)
}

investment <- function(dt) {
    # comp dependencies: act
    dt[["investment"]] <- (dt$assets - lag(dt$assets))/dt$assets
    
    return(dt)
}

cashFlowToPrice <- function(dt) {
    # merged dependencies: act, ib, txdc
    dt[["equity_share"]] <- dt$dec_market_equity/(dt$assets - dt$book_equity + dt$dec_market_equity)
    
    dt[["cf_price"]] <- (dt$earnings + dt$equity_share * dt$depreciation + dt$deferred_tax)/dt$dec_market_equity
    
    return(dt)
}

dividendYield <- function(dt) {
    # crsp
    dt.jun <- dt[month(date) == 6, .(permno, rebalance_date, adj_ret, adj_retx)]
    
    dt.jun <- dt.jun[, `:=`(c("rebalance_date", "div_yield", "adj_ret", "adj_retx"), list(rebalance_date + years(1), adj_ret - adj_retx, NULL, NULL))]
    
    dt <- merge(dt, dt.jun, by = c("permno", "rebalance_date"), all.x = TRUE)
    
    return(dt)
}