How do professional traders use tech like AI without losing their personal touch in trading decisions, especially when the market changes quickly?
AI trading
- Thread starter eleanorK80
- Start date
In the context of your question, 2 different use cases come to my NOT all-knowing brain, both allowing the trader to maintain a personal touch.
1) AI-based screener.
2) Ai-based specific instrument comparatives.
Im sure there are MANY other uses.
Ai in its current form, is merely an if/else lookup table, with links to other if/else lookup tables, and so on and so on.
The personal touch lives on in ...
1) the construction of the ai data/learning model AND linkages.
2) the validity of the data and the results of inclusions.
The dangers for trading are obvious.
https://tylervigen.com/spurious-correlations
https://tylervigen.com/spurious/cor...diana_correlates-with_apples-stock-price-aapl
Robert Morse
Sponsor
"Professional Trader" is a very broad category. If your sole income is from trading, or you work for a Prop firm or hedge fund or on the trade desk of a GS, or a MM on an Option or Futures exchange, these are all Professional Traders. Some work with $35,000 and some work with $350mm. The key is who is in control of your strategy. When you work for others and their money, in general, you are hired and provided with an allocation to do one thing. To do another, you need to get permission. When it is your money, all you care about is risk/reward. You can do whatever you want to get the results you expect or need. Your personal "touch" and feelings should not be an obstacle.
That's why a balanced approach is preferred, you can make all the adjustments and analyses you want with bots and AI and what have you, but in the end, you make the decision to execute a trade or not, and furthermore, I doubt people actually leave everything in hands of bots and if you are indeed a "professional" trader you are always ready for the worse.
I've always thought using a bot in the first place was to remove such "personal touch" from trading.
As for your question, it's simple. You pull the plug when the loss exceeds x%
Shouldn't that be "Labrador" 
EleanorK80 the AI, and more specifically machine learning, is used in many aspect of a trading system. Some used for researching only, others for portfolio optimization, execution optimization, etc. I think, and maybe I am wrong, that you assume the AI is used for generating trading signal (buy/sell) but that is not the case in many occasions.
Anyways, depends on where and how the traders use it, they can give it their "touch" They could trained it for high volatility from the beginning... The casuistic is soooo big that it is impossible to even scratch the surface here.
Anyways, depends on where and how the traders use it, they can give it their "touch" They could trained it for high volatility from the beginning... The casuistic is soooo big that it is impossible to even scratch the surface here.
%%
SEVERAL things to consider;
AI is one name for it[Artificial= fake, untruthful , fake , untrue a lot of the time]
But heard some in the industry said he preferred ''machine learning'' since some of it helps+ is accurate, machine learning helps.
So i will use both names; names are important .
Good thing for me I trade metals some; therefore i dont pay pro fees + should not, diversification in business .
So as you asked , how could some use to AI or machine learning to profit, since cant predict a ,market like SPY, SH some or QQQ......................?
WELL lets see we know FEB \SPY is usually down \machine learned data + Stock Traders Almanac. So far so good; oops FEB was super strong in QQQ+ better than average SPY.
So good thing some did not just trade invest, short based on history AI or machine learning, worst FEB would have been result.
Most March months are up based on Stock Traders Almanac [book by Hirsch] , based on History machine learned ;
but since the humans invented AI , I never trade blindly based on machine learned history.
[AS a practical point super swings-whipsaws in a single day, too easy to for me get slaughtered , so I find a good trending market or ones i understand like SPy + QQQ.]
%%
Something else, market makers can do well, even with super swings % intraday, real markets like stocks + ETFs, so again, so far so good ,for them.
Machine learning can work real well for the market makers, so far so good ,again.
Knight Capital, except for some fines, did real well;
except the time when thier machine learning was more mistakes ,AI\ big mistakes , then Knight Capital firm has such loss \ had to sell co for a bargain price, matter of public record.
So as we say on Sunday, ''never check your brains off @ the door'' LOL-true
,
My comments only apply to real markets. Good plans help, like Mr schizo noted s you get out[pull the plug, as planned ] when loss exceeds -x%.
And even in real markets \ JPM buyer, said they did not pay up for Bear Stearns, ''don't pay up for a house on fire'' Amen.
Good questions + comments.
Something else, market makers can do well, even with super swings % intraday, real markets like stocks + ETFs, so again, so far so good ,for them.
Machine learning can work real well for the market makers, so far so good ,again.
Knight Capital, except for some fines, did real well;
except the time when thier machine learning was more mistakes ,AI\ big mistakes , then Knight Capital firm has such loss \ had to sell co for a bargain price, matter of public record.
So as we say on Sunday, ''never check your brains off @ the door'' LOL-true
,My comments only apply to real markets. Good plans help, like Mr schizo noted s you get out[pull the plug, as planned ] when loss exceeds -x%.
And even in real markets \ JPM buyer, said they did not pay up for Bear Stearns, ''don't pay up for a house on fire'' Amen.
Good questions + comments.
Last edited: