although the contracts themselves are continuous in time, trading hours
on a stock exchange are not. I think the function reports the right
data, so let me mention a possible economic reason for your results:
Stock data is commonly reported only on trading days. Therefore it is
correct to get a time-series of only trading-day data. Think about e.g.
closing prices: On a non-trading day it would be the same value as the
closing price the previous trading day, so the return of any financial
asset on a non-trading day would be zero. Any estimation of paramaters
(beta-factor,...) would be biased by these data generated on non-trading
days. So don't worry about missing data on a sunday.