tests/testthat/test-backtest.R

In jmuhlenkamp/tidal.backtest: Package for Backtesting

context("backtest() basics")
test_that("backtest() handles errors correctly", {
    expect_error(backtest(1:26, LETTERS))
})


context("backtest() final nav within 1 bp excel calculation: daily rebalance")
etfs_5days <- dailycloseetf[dailycloseetf$date <= as.Date('2017-10-30'),]
etfs_5days_cash <- dplyr::bind_rows(dailyclosecash, etfs_5days)

daily_long100 <- dplyr::pull(dplyr::filter(
    backtest(prices = etfs_5days_cash, data = dplyr::mutate(dailycloseetf, w = 1/3))$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Daily Rebalance & 100% Long Exposure",
          expect_equal(daily_long100 / 1003182.5310446 - 1,
                        0, tolerance = 0.0001, scale = 1))

daily_long050 <- dplyr::pull(dplyr::filter(
    backtest(dplyr::mutate(dailycloseetf, w = 1/6), prices = etfs_5days_cash)$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Daily Rebalance & 50% Long Exposure",
          expect_equal(daily_long050 / 1001675.91084044 - 1,
                       0, tolerance = 0.0001, scale = 1))

daily_long150 <- dplyr::pull(dplyr::filter(
    backtest(prices = etfs_5days_cash, data = dplyr::mutate(dailycloseetf, w = 1/2))$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)


context("backtest() final nav within 1 bp excel calculation: weekly rebalance")
etfs_7days <- dailycloseetf[dailycloseetf$date <= as.Date('2017-10-31'),]
etfs_7days_cash <- dplyr::bind_rows(dailyclosecash, etfs_7days)
df_price <- etfs_7days_cash

week_long100 <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/3), '.test', 'week')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Weekly Rebalance & 100% Long Exposure",
          expect_equal(week_long100 / 1004273.34906276 - 1,
                       0, tolerance = 0.0001, scale = 1))

week_long050 <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/6), '.test', 'week')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Weekly Rebalance & 50% Long Exposure",
          expect_equal(week_long050 / 1002217.74233325 - 1,
                       0, tolerance = 0.0001, scale = 1))

week_long150 <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/2), '.test', 'week')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Weekly Rebalance & 150% Long Exposure",
          expect_equal(week_long150 / 1006249.14558189 - 1,
                       0, tolerance = 0.0001, scale = 1))


context("backtest() final nav within 1 bp excel calculation: monthly rebalance")
etfs_9days <- dailycloseetf[dailycloseetf$date <= as.Date('2017-11-02'),]
etfs_9days_cash <- dplyr::bind_rows(dailyclosecash, etfs_9days)
df_price <- etfs_9days_cash

df_test <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/3), '.test', 'month')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Monthly Rebalance & 100% Long Exposure",
          expect_equal(df_test / 1009969.6748769 - 1,
                       0, tolerance = 0.0001, scale = 1))

df_test <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/6), '.test', 'month')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Month Rebalance & 50% Long Exposure",
          expect_equal(df_test / 1005087.90897054 - 1,
                       0, tolerance = 0.0001, scale = 1))

df_test <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/2), '.test', 'month')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Monthly Rebalance & 150% Long Exposure",
          expect_equal(df_test / 1014748.29119619 - 1,
                       0, tolerance = 0.0001, scale = 1))


context("backtest() final nav within 1 bp excel calculation: qtr rebalance")
etfs_50days <- dailycloseetf[dailycloseetf$date <= as.Date('2018-01-03'),]
etfs_50days_cash <- dplyr::bind_rows(dailyclosecash, etfs_50days)
df_price <- etfs_50days_cash

df_test <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/6), '.test', 'qtr')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Quarterly Rebalance & 50% Long Exposure",
          expect_equal(df_test / 1036400.81989253 - 1,
                       0, tolerance = 0.0001, scale = 1))

df_test <- dplyr::pull(dplyr::filter(
    backtest(prices = df_price, data = dplyr::mutate(dailycloseetf, w = 1/2), '.test', 'qtr')$dailynav,
    dplyr::row_number() == dplyr::n()),
    price)
test_that("Final Nav: Quarterly Rebalance & 150% Long Exposure",
          expect_equal(df_test / 1106057.80874977 - 1,
                       0, tolerance = 0.0001, scale = 1))