Human Traders Are Trouncing the Machines

[luisHK]

I know plenty of discretionary traders who have massive losses, what does that say? Not much. Automation requires much more work, I'd say 7-8 years without any background in the field. This is because you have to master the trading part AND the technical side of things.

I'm a bit surprised you got to know plenty of traders who ever dealt with massive amounts of money, win or lose.
But yes, saying that automated trading is an easier learning process than discretionary trading sounds far fetched (discretionary trader here)

 

[d08]

I'm a bit surprised you got to know plenty of traders who ever dealt with massive amounts of money, win or lose.
But yes, saying that automated trading is an easier learning process than discretionary trading sounds far fetched (discretionary trader here)

Why surprised? But they are pikers, small-time. That shouldn't matter for this comparison, it was simply to prove a point.
Personally, if I have a choice of a data-scientist nerd type who has no strategy but an environment for proper research vs. a discretionary trader waking up every day saying "today I'm going to make it"...I'll bet on the nerdy one.

 

[volpri]

you should read: beating the machines.
written by a famous paper trader.

 

[lawrence-lugar]

This is ET...not AT, AverageTrader.com

We, or I, could care less about the collective, Average masses' returns.
If 1 out of a 100 traders excels greatly...I would naturally be curious. I don't care if they are considered a weird or temporary or unique outlier.

So for every Success, of course there are a whole bunch of failures.
But that doesn't mean the averageness of automation or machines or Whatever perceived safer method is better.

Do you have sex with a machine too,...high-end sex dolls have greatly improved, they're all the rage in Japan

 

[SoesWasBetter]

1. used as an excuse for the last 100 years
2. no master of the Universe has 5 losers in a row

 

[sss12]

1. used as an excuse for the last 100 years
2. no master of the Universe has 5 losers in a row

Why surprised? But they are pikers, small-time. That shouldn't matter for this comparison, it was simply to prove a point.
Personally, if I have a choice of a data-scientist nerd type who has no strategy but an environment for proper research vs. a discretionary trader waking up every day saying "today I'm going to make it"...I'll bet on the nerdy one.

While I trade with a rules based descretionary bent personally, I can see how certain styles certainly would lend themselves to more of an automated system.

But, my question is outside of RenTec'and a few high profile/huge infrastructure players, who is making all these returns with automation, etc. ?

I' m not saying it's not happening, but I don't know who. ..anyone have specific examples of small/midsize managers ? the ones I have come across (day job in Wealth Management) have ave returns at best even with supposedly lower deviation/Sharpe ratio factored in.

 

[ironchef]

Besides, the Princeton geeks were doing this before most of these quants were born and will tell you its nothing new and will be a long hard road to get & maintain good performance.

I googled Princeton geeks and came up empty handed. Who are they and what were they doing?

 

[ironchef]

Systematic trading has wider draw downs and their upside has taken a nose dive now that that arena is so crowded. Discretionary traders typically use a rules based approach to minimize emotions. Programming automated strategies is pretty easy these days. Even I can program on 4 different platforms and I am just playing with it for kicks.

Systematic traders do have to deal with emotions- after all they can have large draw downs and often shut them down at the wrong time - that was in Market Wizards. The problem is few retail will have the conviction and or account size to let them run in expanding draw downs. Systematic trading has its place - but to proclaim it is wildly superior to discretionary is just lying to yourself.

Barclays

View attachment 178645
View attachment 178646
IASG

Discretionary
View attachment 178647
Systematic
View attachment 178648

Can you explain your numbers:

1. Discretionary started with 1000, ended with ~8600 CAROR=18.99%

2. Systematic started with 1000 ended with ~ 8500 CAROR=22.88%

Shouldn't the compounded annual rate of return be the same?

Thanks.

 

[schweiz]

1. used as an excuse for the last 100 years

Wrong, as 100 years ago the computer did not yet exist. So tbis argument could not be used against automated trading.

2. no master of the Universe has 5 losers in a row

200% correct. 5 losses in a row is proof of a horrible trading plan, or a trader who is, in reality, a loser rather then a trader.

 

[schweiz]

Programming automated strategies is pretty easy these days. Even I can program on 4 different platforms and I am just playing with it for kicks.

what you say is like: playing basketball is very easy.
it all depends of what you exactly mean. you can play basketball like you do and you can play like Rodman. there is a huge difference between the two, who both play basketball.

i have seen strategies that could not be programmed by +25 years experienced ICT wizards. the wizards were even not able to summarize what exactly should be programmed in an understandable way to start from. theoretically everything can be programmed. however theory and reality are sometimes in conflict, and reality always wins.

so "pretty easy" is at least an understatement.