It must be those "great forensic skills" that produce blunder after blunder. Here is small example: I predict this share price will be under 5 cent later in 2016 / early 2015. 2016 to early 2015 ???????????? If true you would have to count backwards. Again, less than persuasive.