Alternative ways to do it are.
This one will be fast:
julia> 100 * count(<(df_symbol.close[end]), df_symbol.close) / nrow(df_symbol)
72.83464566929133
or (this one is probably simple to understand)
julia> using Statistics
julia> 100 * mean(df_symbol.close .< df_symbol.close[end])
72.83464566929135
or (using a function designed for this means, the issue is that you need to learn it 
):
julia> using StatsBase
julia> percentilerank(df_symbol.close, df_symbol.close[end], method=:strict)
72.83464566929135