Why is that? Mathematics are always 'right'....
Find mathematics so you sell higher than you buy....; if you can successfully accomplish this with quick enough data and quick enough trades, which is a huge caveat, then what is the problem?
I wrote an algorithm that will give you a stock that will make .5% the next day...BUT
- Thread starter selrod0324
- Start date
Why is that? Mathematics are always 'right'....
Find mathematics so you sell higher than you buy....; if you can successfully accomplish this with quick enough data and quick enough trades, which is a huge caveat, then what is the problem?
An algorithm does not imply mathematics whether applied well or not. It implies a procedure. The procedure may be based on mathematics or it may not.
Market mathematics implemented in algorithms either applied mentally or by computer are a way of identifying market psychology and finding probabilities of future action based on past action under similar circumstances.
Not all mathematics is done well in this field especially. Just because there is math involved does not mean there is any 'right' involved.
Algorithmic trading (most of which indeed contains a heavy math component) is a far more competitive space than it was even 2 years ago. And I'll venture to guess that it will continue to get even more competitive.
I think you'll find that returns in the real world will be less than predicted -- perhaps even more than .5% less. Sketchy data in itself can point to "strategies" that are nothing more than flagging garbage data. And even among accurate data, a lot of equity volume is now off limits to anyone but the brokerage trading against its own client ("B/D Internalization").
I am not trying to be discouraging here -- if I were you, I'd pursue it till I figured out what's going on, out of nothing more than raw curiosity, at least for a time. But I'd also suggest not getting your hopes too high.
I haven't read the whole thread, but in response to the question of back-testing, why not just forward-test? Apply the algorithm to, say, the 500 tickers of the S&P for the next 3 months and see what you get. I know that may seem like a waste of time, but I have always found forward-testing to be much more valuable than back-testing. 3 months forward is worth more than 3 years back.
You are qualified to make that judgement, how? Or is this an illustration of transference?
I am ignorant to the market and everythign around it and have just discovered paper trading to my delight
Dude you're totally missing his point. He never claimed that he is guaranteed to make money. He just said that it looks promising. And you're wrong about algorithms not being rewarded, because they are. Combine a super algo with a super system and you've got a winner. If his strategy is market neutral (like mine is), then he really doesn't care which direction the market is going, so you saying that this market is the worst you've ever seen is basically irrelevant. Don't get hyper and start telling people they are going to lose money because they don't know what they are doing. He's here to learn so he won't lose money lol. Just take it in stride and go with the flow
