Discretionary vs fully Systematic

[virtusa]

I'm probably one of those whos brain isn't capable of doing that.
I know from my experience that I get into less trouble when I follow my rules.

I think less than 10% of traders are able to perform very good with discretionary rules. So you are perfectly normal. I am not normal as the majority defines what is normal and what is not.
I also follow my rules. But I have hard mathematical rules and discretionary rules. The combination of both keep me out of trouble.

 

[virtusa]

Can you give me an example of a logical rule that isn't one of your rules.
Maybe my rule set is larger than most. I hope that my trading plan answers the question "What do I do now?" If you have discrectionary rules it sounds like you are not sure.

I only apply discretionary rules if I have proof, after extensive testing, that they have a very high rate of success. So I am fairly sure that these discretionary rules will be right.

Give an example is difficult as I only know those that I apply...

I will give one that I use, but not in detail: when trends are very strong you should avoid taking profits too early, and also avoid to go in the other direction too early.
So in that situation I skip signals that would tell me to go already in the other direction. I see based on discretionary when to skip and stay in my position. Success rate is almost 100%, and additional profits, by ignoring the hard rules, are many times very big. Not using the discretionary rule would result in much smaller profits and also in losses as the system goes too early in the opposite direction. It is impossible for me to put it in a fix rule as it depends on a price. And it is impossible to put a fix number on that, as the price can vary. So I cannot say if X should be between +25 and +45... Because the value can be out of this range and still be valid. I watch the total formation on my screen to recognize what to do.

Compare it with driving a car. You have no fix rules when to break. All depends of the situation. You judge on that situation what you should do. And as there are millions od different situations you would need millions of rules. An experienced driver drives "discretionary" not with only fix rules.

 

[VSTscalper]

I - ALWAYS - follow my trading rules - win or lose.

Many years ago - when I first started trading - I thought - this should be easy. In the beginning - I was trading Stocks - made a lot of money. I got a little cocky - thinking I was such a damn good trader - it is so easy to make money trading. At first - I traded part-time - kept working - as I really loved the type of work I was doing. Finally - in the early to middle 1990's - I decided to quit work - go full time as a trader. Not long after - I started trading really STUPID - and damn near went out of the trading business. I guess you could say - I thought I was smarter than the market - got a little arrogant - and the market finally showed me - I wasn't even close to being - even a marginally good trader.

Going forward

Well - I had to decide - do I keep trading - or go back to work. As far as going back to work goes - it was a hard decision - because when I left - a damn good job and career - many in the company knew what I was going to do. Hell - would they even give me my job back. Finally - I decided - to make some changes - keep trading.

Without getting to deep into what I did - what I will say is - the type of work I did - was really - based on Rules. The funny and sad thing when I started trading - was I did have some Rules - but some VERY WEAK - and in a way - MEANINGLESS RULES. Excuse my example - but say you Fart - it smells Good - you Buy - if it smells Bad - you Short. So - you might say - basically - my RULES - were really STINKY.

Finally - I stopped trading for a short time - really put some thought into - WHY - I was failing as a trader. The type of work I did in my career - in the Aerospace industry - you couldn't make mistakes - as lives depended on the finished product. If I had done my job - like I did my trading - I am sure - I would have been fired.

E-Book

I better stop - as I am probably already boring the hell out of whoever reads this - so sorry - if you get a headache.

I finally realized - I needed to take trading SERIOUSLY - make some good RULES - then actually FOLLOW THE DAMN RULES. There is a saying - "Rules are make to be Broken" - well - if you have GOOD RULES - and YOU BREAK THEM - you might as well NOT HAVE ANY RULES AT ALL.

Here is an old e-book I put together years ago - hope it helps some traders.

 

[VSTscalper]

By the way - about the topic of this thread. I am a Day Trader - and I develop Algos. Even with my Algos - I still prefer taking the Algo Signals and trading them Discretionary. I would NEVER let an Algo trade - without watching it - as there are many things that can go wrong - that have NOTHING to do with the Algo itself. Things such as a power outage - even for a second or less - can cause the problems with the Algo. Yes - there are things that you can do - like having a backup source - but MOST traders do not have that.

I only trade for 2 hours in the morning - NEVER more. In my younger days - I would trade all day - make money early - give some or all back at times or go negative - the longer I traded. Even though I was younger - it really wore me out some days. You might think now - how old is this guy - well - let me just say - "time" had NOT been invented yet - so when people ask me - I just tell them - I was born - before "time" began - and even friends and family - look at me - and think - yeah - the way you look - that could be true.

 

[deaddog]

I only apply discretionary rules if I have proof, after extensive testing, that they have a very high rate of success. So I am fairly sure that these discretionary rules will be right.

Give an example is difficult as I only know those that I apply...

I will give one that I use, but not in detail: when trends are very strong you should avoid taking profits too early, and also avoid to go in the other direction too early.
So in that situation I skip signals that would tell me to go already in the other direction. I see based on discretionary when to skip and stay in my position. Success rate is almost 100%, and additional profits, by ignoring the hard rules, are many times very big. Not using the discretionary rule would result in much smaller profits and also in losses as the system goes too early in the opposite direction. It is impossible for me to put it in a fix rule as it depends on a price. And it is impossible to put a fix number on that, as the price can vary. So I cannot say if X should be between +25 and +45... Because the value can be out of this range and still be valid. I watch the total formation on my screen to recognize what to do.

Compare it with driving a car. You have no fix rules when to break. All depends of the situation. You judge on that situation what you should do. And as there are millions od different situations you would need millions of rules. An experienced driver drives "discretionary" not with only fix rules.

I believe we are arguing semantics. You have discretionary rules based on your past experience. If/Then scenarios. It this situation presents itself; Then i will react this way.
I would call that a rule based system.
To me a discretionary system is having the same set-up present itself and you decide whether to take it or not based on a hunch or on how you feel at the time.
If you can rationalize why you ignored a rule based on some sort of logic, then that's fine. If you can answer the question "Why did you do that?" with anything besides "I don't know" or "I was afraid that ..." you are probably allright.

 

[virtusa]

I believe we are arguing semantics. You have discretionary rules based on your past experience. If/Then scenarios. It this situation presents itself; Then i will react this way.
I would call that a rule based system.
To me a discretionary system is having the same set-up present itself and you decide whether to take it or not based on a hunch or on how you feel at the time.
If you can rationalize why you ignored a rule based on some sort of logic, then that's fine. If you can answer the question "Why did you do that?" with anything besides "I don't know" or "I was afraid that ..." you are probably allright.

Sometimes it is not clear what will happen. I then think and take the most obvious choice.
That choice is always based on past knowledge. Only idiots take decisions based on "I don't know".
"I was afraid that ..." is not just agamble. It is still based on some past experience or pattern that sometimes happens.
To me discretionary means ecisions without fix rules, so a situation where multiple outcomes are possible.
I clearly have decisions based on two different approaches: fix rules and rules (or rather logical decisions) that are not fix but can vary. If I replay these situations I sometimes take different decisions, so to me that is discretionary. It would be a rule if i would always take the same decision when repeating that situation.

 

[deaddog]

To me discretionary means ecisions without fix rules, so a situation where multiple outcomes are possible.
I clearly have decisions based on two different approaches: fix rules and rules (or rather logical decisions) that are not fix but can vary. If I replay these situations I sometimes take different decisions, so to me that is discretionary. It would be a rule if i would always take the same decision when repeating that situation.

OK You seem to be able to profitably combine rules based and discrestionary trading. Like you said you are one of a few. Would you recommend your methodology to new traders?

 

[virtusa]

OK You seem to be able to profitably combine rules based and discrestionary trading. Like you said you are one of a few. Would you recommend your methodology to new traders?

Each trader has different abilities and weaknesses. So there are three different possible outcomes if other traders would use that methodology:

1. make less money;
2. make the same amount;
3. make more money.

There are a lot of factors that have an influence on the final result. What works for A can be detrimental for B.
Each trader should try to find out what works for him. It took me more than 20 years to get to the point where I am now. I am already many years consistenly profitable, but not at the level that I am now. So new traders will have to go thru a very long learning cycle and maybe be successful.

If you would do the same learning cycle as a worldchampion in any sport, would you be as good as he? Probably not. There are a lot of external influences or requirements that can have a serious impact on the final result.

 

[alphadillo]

Sounds like we're in a very similar place. My background not in finance. I also started trading part-time in 2020 but haven't found consistency yet. I also think I know exactly what you mean by "fighting at the borders" of your system. Although I've built a pretty detailed system, I still find myself on too many days convincing myself today's scenario is something my system hasn't yet "accounted for," and that leads to some discretionary judgment (which either works or doesn't) followed by an updating/expanding of my rules after the fact. That dangerous habit has led me to seek out a more systematic approach where I have less and less say over things, and yet there always seems to be these moments where interpretation (of single trigger candles, or larger patterns) is needed, and that's where things usually go awry for me. I still have a long way to go, I've come to realize, but I do think backtesting and writing out your plan IN DETAIL holds one more accountable and thereby improves discipline. As others have said, you won't obey a system you don't believe in, and that's where thorough backtesting comes into play and can really help your discipline improve. One question for you: are you trading multiple things? For relative beginners like ourselves, I'm a big believer in really getting to know a single instrument, only trading one thing and backtesting it in a way where you get used to all its main scenarios, its quirks, etc. This is all very general, but hopefully helpful in some way.

Thank you & also all the other participants of the thread for all the replies! I appreciate the feedback!

I 100% agree that a written out plan helps tremendously. In the beginning I traded different markets (Dax, US, OTC) and today I only focus on US stocks. Discipline wasn't too bad in the last weeks, only a few slip ups, but overall I'm following my system. Simultaneously I started my backtesting journey and took a look at Python. Currently I work with templates, but I want to acquire more skills to develop my own scripts for my particular needs. The idea of having a clear statistically backed edge that gives out signals and then act on it according to median stats etc. as guidelines, combined with a small part of discretion for example in terms of profit taking, interpreting the bigger picture, etc. really speaks to me. 80% systematic, 20% discretionary is the goal.

Wish everyone all the best & will keep you updated in here!

 

[Laissez Faire]

Compare it with driving a car. You have no fixed rules when to break. All depends of the situation. You judge on that situation what you should do. And as there are millions of different situations you would need millions of rules. An experienced driver drives "discretionary" not with only fixed rules.

This is a great analogy which really underlines what's required of a discretionary trader (experience and screen time), the potential pitfalls of discretionary trading and also why it's generally hard to make a good strategy that's 100 % systematized (too many different outcomes and edge cases).

Thank you & also all the other participants of the thread for all the replies! I appreciate the feedback!

I'm not sure how much my opinion is worth as I'm not yet rich from trading, but I'm a mix between systematic and discretionary.

I've taken steps to remove as much discretion as possible from my trading, but at the end of the day, my trades are still entered manually.

For example, I no longer draw any lines on my charts. It's all drawn automatically from my trading software. I also heavily rely on statistics and know what to expect on any given market day. Yet, that still leaves a lot to interpretation and that's where experience comes into play. For example, how do you differentiate between a false and a real breakout of an objectively drawn level systematically? How do you incorporate time, pace and momentum in your system? What if the market approaches a key level from N days ago which I've proven to be statistically significant through back-testing? And that N is not a consistent number?

I hope to be able to systematize my approach even further in the future, but it's not easy.