Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
This is a really interesting thread. I'm not sure quite how to respond yet, but here's sort of my ideal answer (reality varies ):
I want to have an "outside in" analysis process to start. That is, I look at larger timeframes and steadily narrow down to the level I'm going to trade. This can create big problems, because I may develop a bias based on a longer-term view that doesn't play out at my actual timeframe; but if I don't do it, I forget about the forest for the trees, so I try to correct for my biases and put the long-term view to good use.
What I'm looking for initially is what is elusively termed "context" -- what the market is up to. For instance, I did not trade yesterday, 8/19/14, which is a good thing, because I know from experience that such a slowly-grinding, low volatility steady trending day (in equity futures, at least), with no significant pullbacks would just grind me up. But, in the ideal case, if I can determine the character of the market fairly early, I can (I hope) adapt my trading to what is happening. In yesterday's case, if the trend is identified early, and if I could see the very limited action against it, the simple thing to do would be hop on the long side and just ride for a while. In other situations, you'd want to do more in-and-out stuff (my unfortunate default; I'm working on it.)
This is a work in progress, at this stage, but when I get it right I can usually trade much better.
After hopefully figuring out the context (which is obvously very discretionary), I have different sets of rules for different market situations. I try to keep them dead simple, since I do not want to be thinking too much while trading. I have to say this is a work in progress also.
But then, here's where we get down to your question. Other than loss control and risk/reward, I think of the trading rules I have as advisory, not absolutes. For example, in a trending situation I want to go with the trend (duh), coming off support or resistance in the trend direction (also duh), but I am going to rely on my read of the price action at the time to make an actual decision. This is because I think, at least, that real situations will always vary a bit from pre-conceived patterns, and I want to be alert to what is happening more than to what I think will be happening.
Long answer -- sorry, I do that.
So, if this is at all helpful, I can say that I have rules I take seriously but don't want to apply mechanically. I use them to try to focus my attention on what probably matters in what is going on with price now, and then act on what seem to be the probabilities at the time.
It seems to me, after observing many successful traders over the years, one thing they have in common is trading with the trend. Some claim to have a system, some claim not to have one, but both camps utilize discretion a lot, if not all the time, even though many wouldn't admit doing it.
And to even start with the basic concept of trend - it is highly discretionary to identify it. It is different depending on the timeframe or how you count HH/HL or how you draw a trend line, etc.
I have been trading a method in a discretionary way.
I then have started to automate it by writing a strategy.
My goal/plan for this was :
- back test the method (validated, get statistics of success/failure, ...)
- multiply activity in more instruments
- let it run while i can sleep or do other things
- eliminate the psychology
I have learned that this is a very very tough job and when you have to explicitly write rules that it is very difficult.
Maybe the method i have chosen, needs too much of 'context' analysis, or doesn't lean itself optimal for automation.
The first results are are quick and fairly easy, after that... wow !
I haven't given up on it, but the learning curve/progress path is extremely steep.
I wanted to share that with you to kind of 'touch' the connection between discretionary and automated.
I've always been a discretionary trader, but over time have developed more and more mechanical tools. The mechanical parts like automated entries/exits, ideintifcation of specific setups, etc, allow me to do "more" , since attention, focus and time are all finite resources.
So if I am heavily trading two different instruments at the same time, some mechanical aspects can allow me to still efficiently trade several other instruments as well without going to a full blown automated system.
One thing I have definitely noticed over the years with many retail traders is they want mechanical type signals and trade setups with discretionary results and profits. They are essentially trying to remove the "discretionary" part in discretionary trading, but at the same time not wanting the limitations of a mechanical system.
I feel most of this is probably psychological, since they might be unsure of the efficacy of their edge/method and if they can get enough signals to fire, bars to paint, lines to cross....overall, enough technical confirmation, then they can finally place that trade.
Some of the best advice given to me over the years was the suggestion to automate as much of the trade hunting and trading process as possible, with the end goal of increasing overall productivity and gains.
If you can keep your wits about you while all others are losing theirs, and blaming you....The world will be yours and everything in it, what's more, you'll be a man, my son. - Kipling
For me, if algorithmic execution helps my strategy then I am trading at too high a frequency.
I want things to be slow enough that algo execution only adds additional risk of something going wrong that I may as well just put buy and sell orders in myself.
For those that replied they are mostly / all discretionary, I would be interested to hear how you measure your performance. What sort of things (metrics?) are important to you in terms of knowing if you're doing well or not.
Or is it more a case that when trading mostly discretionarily, analysis of stats to determine potential areas of improvement in your plan is not as important. (edit: Just thought I would add. I have no doubt analysis is done. But from a purely stats analysis point of view is it perhaps not as important as to those who trade a mechanical plan?) Is it more a case of looking back over your performance and seeing if you were in tune with the market? (and if that is the case, how do you do that methodically?)
Thanks for the replies, im finding the differing approaches quite interesting. So far there don't seem to be many people who use a blended approach (eg: identifying their areas of interest is discretionary. Then the trade entry and management is mechanical)
For myself, the main measure I look at it is how the mechanical method or portion of my trading is benefiting me in ways that a pure discretionary approach cannot.
For example, a mechanical approach might open up doors to trade additional instruments, which would not be possible if they were to be trader directly by myself.
The biggest thing for me to overcome when first using mechanical methods was accepting the limitations of automation. I needed to overcome the notion that 'I could trade it better'....and as true as it might be, if that was truly the case, then I would of never looked into automation to begin with. So in the end, it was really a matter of looking at the tradeoffs between the two within the context of the overall edge or method.
If you can keep your wits about you while all others are losing theirs, and blaming you....The world will be yours and everything in it, what's more, you'll be a man, my son. - Kipling
I keep it simple, 60-70 trades evaluated over a month - did the account rise or fall? I plot the equity curve and log or post it. Then I ask why? So far the answer is always the same, when I am being sensible I can trade as well as anybody. When I am in tilt mode the mean will always revert to zero.
At this stage I don't need any more complex statistics. I hope to reach the stage where I will, and then they can make the next level of difference, but for now it's all about my trading behaviour, nothing else.
On a related theme that is my software focus for the year ahead, building mechanical or semi-automation tools that can influence and improve my trading behaviour.
Interesting that so many people lean towards discretionary trading.
I too trade discretionary however my strategy is 100% mechanical.
I use to trade purely discretionary, stops and all but figured emotion was too big of a variable and it was too hard to monitor what works and what doesn't.
So I decided to develop an edge that eliminated emotion or any decision making, a strategy that I could execute instinctively.
Took me 3 years to create a strategy that I am confident with (capital was also a factor) but much prefer this way of trading.
I suppose though it depends on what works for each individual and the traders personality also.
If I knew how to code I am sure my strategy could be traded as an automated strategy however this early in the game I prefer to execute my system manually due to experience gained and the kick I get from it!
):
