dplyr ifelse not working in mutate

I will start off by saying that I apologize for this example not being reproducible I don't think that I can get the necessary data into stack overflow for this to work.

Anyhow, when I run the following code theta still has negative values.

data <- data %>%
  group_by(ob_group) %>%
  mutate(
    pct_superior = sum(ma_mw, na.rm = T) / length(ma_mw[!is.na(ma_mw)]),
    theta = as.numeric(ifelse(benchmark > 0, (pct_superior - benchmark) / benchmark, 0)),
    theta = as.numeric(ifelse(theta < 0, 0, theta)))

All of the variables excluding ob_group are numeric.

Thoughts?