DT-waw's general thoughts on trading

D

DT-waw

I'd like to summarize my view on trading and discuss it with more experienced traders.

1) The nature of markets
Dynamic, non-linear process. That's why there's no Holly Grail, no single formula which explains market's swings. There are many ways, methods to profitable trading. We can't say that one indicator/system/style is better than the other.
Most of the time, predictions about the future probably have 50/50 chance to be successful. But there're few moments when predictions can have an edge above 50%.

2) Systematic vs discretionary trading
Since markets are non-linear, trading systems will fail. Every system must experience a huge drawdown in the future. Optimization won't help, because markets change. Diversifying over a range of different systems and markets may be a good idea. Trading systems may not react in timely manner to fundamental news or events - they can appear suddenly and can have large influence on markets. Discretionary trading can be more flexible and ( who knows? ) more profitable.

3) Money management
Probably the fundamental thing in trading. After 50% loss, you must earn 100% to go flat. Math will never change, so proper money management is the key to profitability and can save you from bankruptcy.

4) Luck
Luck has some impact on trader's profit or losses. We can't calculate it, only estimate it. One thing is for sure: the more trades are made, the less significant luck factor is.

Markets have to be non-linear in order to exist. Trading is more an art than science ( now I agree with you, ChrisM! :) ).

That's all. Comments appreciated!
 
1) I think chaos theory describes the markets well. The market is constantly evaluating and digesting new information, so it is very hard to make a prediction of where it might be more than a few days ahead.

Re forecast accuracy, my winning percentage has been 67% over a long period of time. My average trade goes for 1-2 days, and average winner is greater than average loser. But I only trade when I feel absolutely sure I have got it right (about 15-20 trades/month).

2) Yes and no. I used to design trading models for a large successful CTA. They have a very consistent track record. Their core methodology hasn't changed much in 10 years. Having said that, I am purely discretionary. There is no way I could get as good results being purely systematic. Knowing how to read the news/fundamentals and being early to recognise possible paradigm shifts can make a huge difference in the long run.

3) Not rocket science, don't bet the farm on every (or any) trade. If you can make 50% per year, in 10-20 years time you will be doing very nicely.

4) Agree.

Also agree that it is more art than science. Almost all the truly great traders are highly discretionary.
 
Originally posted by zentrader


Also agree that it is more art than science. Almost all the truly great traders are highly discretionary.
More...



It would be interesting to hear who and where all the truly great traders are, apart from the ones on Elite that is.

Is successful trading encapsulated as discretionary because no one need give any other credence to these methods, than they are intuitive and these people are special in some way in which they see the market ?

In this context mechanical systems have to stand up to scrutiny over time. Discretionary traders can come and go.
 
1. Read in a Futures Mag article, about "artificial intelligence and the predictibility of markets"... Their conclusion was that there was SOME degree of predictability "for the next bar only"... all time frames. Beyond that, none.

2. Trading = Guessing. Over time, you will learn "what to guess at"... presuming, of course, you survive the "introduction".

3. Develop GREAT stop discipline. It's your life jacket.

4. You might be unlucky enough to go bust anyway. It's more likely if you violate #3.

I believe your "edge" will come in the form of (a) superior stop discipline, and (b) learning what NOT to guess at.

Good luck (from someone with 20 years in front of a real-time screen).
 
Thanks.

Originally posted by gnome
1. Read in a Futures Mag article, about "artificial intelligence and the predictibility of markets"... Their conclusion was that there was SOME degree of predictability "for the next bar only"... all time frames. Beyond that, none.
More...

Thats exactly what chaos theory tells - only some short term predictability.
 
1) no doubt. any trader should fully understand that. non linear it is.
just like the weather it is predictable very short term.
- the comparison goes further : although only predictable short term, seasons do apply to the markets too (it is statistically possible to know when in the year the market has a upward bias or downward one).

2) I don't fully agree with you regarding systematic vs discretion. it's not surprising.
my reasons
- first I do both and both make money.
- if it's predictable then it can be by a human or machines.
- large firm and hedge funds do both.
systematic is often misunderstood by momentum daytraders.
I do agree discretion may improve results upon systematic approaches. This is, I theorize, because the brain can process when things are not 'right'. the system does not care (although we try here to code as much auto detection of failing market conditions as possible. in systematic trading the idea is to keep the human out as much as possible)

3) of course. money management is key. You can't cheat math.

4) luck. well, I hate luck. because it has a mirror called bad luck. One should try to avoid luck in his trading. Luck is for the unprepared.
I agree when you say the more you trade the less it is important anyway (in number of trades). Said another way, if you don't know what you are doing, luck can't help you.

Interestingly, I trade 3 or 4 time frames. But the active trading is done with an average of 1.22 to 2.84 bars holding period. With day trading it's higher because some positions I let run when possible.
With market making the rules are totally different and that is only with discretion. then it's important to know when not to trade. It is very gratifying for a trader imo. But it is the least reliable results long term, so I would caution about it.
I do long term trading too, but then I use the seasons (remember the weather, it's about statistics) instead of usual trading methods.

art form or not, well, I would say it is not. it's exciting when you are in sync with the market, and feels great. But calling it an art form, I don't think so. One does not express itself when trading. anyway, whatever you believe, I simply think artform or not won't make you trade better or not. not really relevant imho. so we can agree to disagree.

tntneo
 
tntneo:
I think systematic and discretionary approaches both have advantages and disadvantages. I suspect many systematic traders will turn into discretionary at some point, probably at higher-than-expected drawdown.

Trading is a mix of art and science. Both theory and practice ( experience ) are needed to master trading. And even with them, everything can happen... beauty of life.

Hmmm chaos theory explains it very very well, but I wonder if there is some better theory.

I agree with importance of knowing "when not to trade" since most of the time market cannot be predicted.
 
Originally posted by DT-waw
tntneo:
I think systematic and discretionary approaches both have advantages and disadvantages. I suspect many systematic traders will turn into discretionary at some point, probably at higher-than-expected drawdown.

Trading is a mix of art and science. Both theory and practice ( experience ) are needed to master trading. And even with them, everything can happen... beauty of life.

Hmmm chaos theory explains it very very well, but I wonder if there is some better theory.

I agree with importance of knowing "when not to trade" since most of the time market cannot be predicted.
More...

An obvious one would be something like, "Fed surprised with a rate cut... SP + 20 points in 10 minutes... don't short just because the RSI is overbought". That's a stupidly obvious example, but one could find many much more subtle. The "obvious avoids" should swing more random probabilities in the trader's favor.

I think of trading almost like Bayesian logic... "every bar requires a new analysis and perhaps a new conclusion..." That's probably where automated system trades are weakest.
 
Originally posted by gnome
1. Read in a Futures Mag article, about "artificial intelligence and the predictibility of markets"... Their conclusion was that there was SOME degree of predictability "for the next bar only"... all time frames. Beyond that, none.

2. Trading = Guessing. Over time, you will learn "what to guess at"... presuming, of course, you survive the "introduction".

3. Develop GREAT stop discipline. It's your life jacket.

4. You might be unlucky enough to go bust anyway. It's more likely if you violate #3.

I believe your "edge" will come in the form of (a) superior stop discipline, and (b) learning what NOT to guess at.

Good luck (from someone with 20 years in front of a real-time screen).
More...

Between #3 and #4 above, I would insert (call it #3.5) the following: "Risketh not thy whole wad". Great stop discipline alone will not save you. If you are risking too much of your equity on each trade, you can stick to your stops religiously, and still blow out if (when) you hit a cold streak. Anything larger than about a 20% drawdown becomes exponentially harder to recover from. Look at your method's history and look at it's largest drawdowns. If they exceed 20% by much, you need to trade smaller or otherwise improve your parameters.

Also, a separate post made the following statement:

"I agree with importance of knowing "when not to trade" since most of the time market cannot be predicted."

IMO, the market can never REALLY be predicted, but the beauty of having a strict methodology is that if it has a positive expectancy (a.k.a., edge), you don't HAVE to predict the market. You follow your method and let the law of averages (your edge) work for you. As long as you keep each loss small enough to not hurt you, or a string of losses small enough to NOT blow you out of the game, then just follow the signals your method produces, for better or worse, and don't even bother with trying to predict the market. If I get in a coin-tossing game with someone where I win $2 on heads, and lose $1 on tails, I don't give a damn what the next flip results in, I just want to know how many times I can play. Should be the same with a trading method.
 
You must log in or register to reply here.
Back
Top