well said. it is not: humans or machines. it is: humans and machines. like in everything you need the best of both worlds. use the benefits of both worlds and avoid the bad parts of them.
One true edge isn't enough with the capital most people have at hand. Say $50k which seems to be a lot for many even on ET - it's unreasonable to expect to make multiple millions with normal risk (with this amount). It's possible if the market is just perfect for what you're doing but the likelihood of that isn't good. With AI, humans are becoming unnecessary (we're still not there though). I'm also not entirely sure you need to think outside the box - I'd say you just need persistent hard-work to analyze and exhaust all ideas. There's creativity involved and some smarts but you don't need to be the smartest guy to do that. There are probably geniuses who don't need to look at data and can just create amazing trading ideas out of the air but I haven't met anyone like that yet.
That's one institution with a wealth of information at their disposal comparing 2 funds and approaches. What about industry averages? It would be very interesting to see how retail systematic and discretionary (manual) traders compare due to information asymmetry.
I'd have to say that's only partially correct. 50K is not enough, unless you're trading out of a professional account. Some props do allow more than equities. (Haircut is really the only way.) But leverage and risk are two very different things.
Exactly. including the vaunted machines that will rule the world of finance - cue Dr Evil Wahahahaha.
%% VIRT, most market makers, specialist$, Knight Capital . Well Knight Capital got 7 out of eight computers right = right=tough business. Dont know any small systems managers. Some one said i was making a prediction this year when i said markets can not be predicted+ i dont consider anything in this post a prediction. LOL
%% Well they were not masters of the universe; + some had more than 5 losses in a row. BUT market trended better then.OK 2017 is not over yet Frankly , Ironchief,with my trading plan; markets dont trend enough for me to have 5 losses in a row LOL.NOT making a prediction-thats a plan....
Interesting thought.... so the implementation is to backtest your strategy and then execute with some discretion ?
it's all about your own abilities and knowledge: how good are you in programming? can you write down your tradeplan in such a way that it is programmable? do you trust your programmer and will he not steal what you created? do you have the discipline and nerves to trade discretionary? maybe the only option is discretionary for certain reasons or maybe automated programs are the only option etc... so for each person the answer can be different. the best solution is of course fully automated. start the machine, go on holidays all year and check every day how many millions the machine made for you. just check the trading companies that go for the machines and watch their real performance. discretionary can still beat most of them in my opinion.