Well, yes it's possible to do it using something like your approach. Your file_merge() subroutine would look like this:

sub file_merge {
  my $hash_ref = shift;
  while (<>) {
    chomp;
    my ($key, @values) = split /-/, $_;
    push @{$hash_ref->{$key}}, [ @values ];
  }
}
[download]

But actually, that's not right as you're reading from the default filehandle each time.

I'd do this completely differently.

sub file_merge {
  my $filename = shift;

  open my $fh, '<', $filename
    or die "$filename: $!\n;

  my %hash;
  while (<$fh>) {
    chomp;
    my ($key, @values) = split /-/, $_;
    push @{$hash{$key}}, [ @values ];
  }

  return %hash;
}
[download]

You would then call it like this:

my %msisdn = file_merge('/tmp/subscription_list');
my %calls  = file_merge('/tmp/Call_Det');
[download]

Another (perhaps better) alternative would be to return a hash reference rather than a hash.

As always strict and warnings will point out some of the major problems with your code.