I am 100% convinced that there are people doing this. I know an ex-Google engineer who's doing it for stock options. I think, however, that to be successful, you'd need to have some comparative advantage, e.g. one or more of the following:
1. access to a data source others don't have easy access to;
2. a reasonably deep understanding of statistics and, particularly, a deep skepticism and conservatism about any
data you look at; or
3. time to invest in looking for oddities in the market.
(3) is probably the "easiest" for a newbie, although it's not as conducive to algorithmic trading, since it requires manual research (though, if you were clever, you could probably come up with something to make this algorithmic). Here the example that comes to mind is the two guys Ledley and Mai mentioned in this chapter of the Big Short: http://www.bookcaps.com/the-big-short-chapter-summaries---ch...
I don't see why (e) teaching scientists to be statistically literate so they don't abuse or misunderstand these tests, and/or (f) focusing on reproducible results and shaming researchers with sloppy methodology, wouldn't work. The hypothesis test has known limitations, but it's not clear that we should blame null hypothesis tests for people mis-using them, when researchers untrained in stats are just as likely to mis-use any method you give them.
1. access to a data source others don't have easy access to;
2. a reasonably deep understanding of statistics and, particularly, a deep skepticism and conservatism about any data you look at; or
3. time to invest in looking for oddities in the market.
(3) is probably the "easiest" for a newbie, although it's not as conducive to algorithmic trading, since it requires manual research (though, if you were clever, you could probably come up with something to make this algorithmic). Here the example that comes to mind is the two guys Ledley and Mai mentioned in this chapter of the Big Short: http://www.bookcaps.com/the-big-short-chapter-summaries---ch...