A lot depends on what you mean by fake/predatory (these terms are not well-defined). Still if you demand a yes/no answer then I will say no.
As Prof. Santa Clause gave in a comment, the criteria to be listed in Scopus are given here. You need a lot of things to be listed, such as diversity in geographical distribution of editors, clarity of abstracts, and no delays or interruptions in the publication schedule. These things aren't easy to fake. The "no delays or interruptions" criterion sinks a lot of legitimate attempts at journals as well.
This particular criterion is especially hard to satisfy: Citedness of journal articles in Scopus. You can see what is needed also on that page: a journal's inclusion in Scopus is threatened if "the journal received half the number of citations, or less, when compared to peer journals in its subject field."
Based on this criterion alone, for a fake/predatory journal to still be listed in Scopus, they must somehow conjure citations from other journals that are also listed in Scopus. This isn't probable unless you go to extremes like postulating that there are a lot of fake journals in Scopus that are citing each other, but not only would the journals have to all get into Scopus in the first place, they would probably fail other criteria, such as "The journal's abstracts are used half as much, or less, when compared to peer journals in its subject field."
tl; dr: it depends on what you mean by fake/predatory, but if a journal is listed in Scopus you can assume that it is reputable until proven otherwise.