estimate_fama_macbeth: Estimate Fama-MacBeth Regressions

In tidyfinance: Tidy Finance Helper Functions

Estimate Fama-MacBeth Regressions

Description

Estimates Fama-MacBeth regressions (Fama and MacBeth, 1973) by first running cross-sectional regressions for each time period and then aggregating the results over time to obtain average risk premia and corresponding t-statistics.

Usage

estimate_fama_macbeth(
  data,
  model,
  vcov = "newey-west",
  vcov_options = NULL,
  data_options = NULL,
  detail = FALSE
)

Arguments

data	A data frame containing the data for the regression. It must include a column representing the time periods (defaults to date) and the variables specified in the model.
model	A character string describing the model to be estimated in each cross-section (e.g., "ret_excess ~ beta + bm + log_mktcap").
vcov	A character string indicating the type of standard errors to compute. Options are "iid" for independent and identically distributed errors or "newey-west" for Newey-West standard errors. Default is "newey-west".
vcov_options	A list of additional arguments to be passed to the NeweyWest() function when vcov = "newey-west". These can include options such as lag, which specifies the number of lags to use in the Newey-West covariance matrix estimation, and prewhite, which indicates whether to apply a prewhitening transformation. Default is an empty list.
data_options	A list of class tidyfinance_data_options (created via data_options()) specifying column name mappings. The date element is used to specify the date column. Uses data_options() default if NULL: "date" = "date".
detail	A logical value indicating whether to return additional summary statistics. If FALSE (default), the function returns only the coefficient estimates. If TRUE, it returns a list with two elements: coefficients (the usual estimates table) and summary_statistics (a one-row tibble with the average cross-sectional R-squared and the average number of observations per cross-section).

Value

If detail = FALSE (default), a tibble with columns factor, risk_premium, n (number of time periods), standard_error, and t_statistic.

If detail = TRUE, a named list with two elements:

coefficients

The same tibble described above.

summary_statistics

A one-row tibble with r_squared (mean cross-sectional R-squared) and n_obs (mean cross-sectional observation count).

References

Fama, E. F., & MacBeth, J. D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81(3), 607-636. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1086/260061")}

Newey, W. K., & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. Econometrica, 55(3), 703-708. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.2307/1913610")}

See Also

Other estimation functions: estimate_betas(), estimate_model()

Examples

set.seed(1234)

data <- tibble::tibble(
  date = rep(seq.Date(from = as.Date("2020-01-01"),
                      to = as.Date("2020-12-01"), by = "month"), each = 50),
  permno = rep(1:50, times = 12),
  ret_excess = rnorm(600, 0, 0.1),
  beta = rnorm(600, 1, 0.2),
  bm = rnorm(600, 0.5, 0.1),
  log_mktcap = rnorm(600, 10, 1)
)

estimate_fama_macbeth(data, "ret_excess ~ beta + bm + log_mktcap")
estimate_fama_macbeth(
  data,
  "ret_excess ~ beta + bm + log_mktcap",
  vcov = "iid"
)
estimate_fama_macbeth(
  data,
  "ret_excess ~ beta + bm + log_mktcap",
  vcov = "newey-west",
  vcov_options = list(lag = 6, prewhite = FALSE)
)

# Return detailed output including R-squared and observation counts
estimate_fama_macbeth(
  data,
  "ret_excess ~ beta + bm + log_mktcap",
  detail = TRUE
)

# Use different column name for date
data |>
  dplyr::rename(month = date) |>
  estimate_fama_macbeth(
    "ret_excess ~ beta + bm + log_mktcap",
    data_options = data_options(date = "month")
 )

tidyfinance documentation built on June 1, 2026, 1:06 a.m.