Thread: System from scratch!

Ok, So I wanted to start a thread regarding building a system from scratch - My intentions of doing this is to:

• Show how much hard work /dedication goes into building a system.
• Show all the areas that I look at with building a system.
• Indicate how much time goes into preparation before any money is put onto the table.
• Discuss position sizing & stops.
• and hopefully learn something about myself / how I do business along the way.

At this stage I do not know if the system I build will be profitable, or even if I will be able to trade it. I will set out with very clear goals and intentions of what I am looking for from this system and attempt to build it.

Of course the disclaimer is that I am not giving out financial advice / if you wish to copy do your own due diligence.

Goals:

The first step of building a system really looks at yourself. You have to ask questions of yourself like the following:

1. What time frames work for me? (i.e are you someone who works a day job, does it suit you to look at small quick trades and be done for the day, maybe you want to spend an hour a week on this.
2. what profit targets do you want to set and do they fit into the time frame you want to trade? (i.e if you are looking at minute bars: 10-20 pips targets / 30min or 1hr: 50 - 100 pips / 4hrs: 100-200 / daily: greater targets)
3. What kind of probability can your psyche take? (i.e are you the kind of person that has to win? or are you able to take a string of losses before you get a win?)
4. how many trades per timeframe do you want to do? (i.e 10 per day, 5 per week, 5 per month ect)
5. Do you want to watch your trades like a hawk or just set and forget?
6. Do you want to use hard stops / trailing stops or an exit signal to take you out of a trade?
7. Will you stop trading if you have a string of losses (how will you know if your system has stopped working?)
8. Will you make this mechanical or discretionary?
9. How much money will you commit per trade?
10. Will the system I build let me sleep at night (HAHA! )

Those will give you an idea, but basically you want to tailor this system to suit you! If you don't do this step - you will end up either not finishing the system building steps , or alternatively ditching the system at the first sign of loss.

So I guess the next step for me is to answer those questions - and I'm going to do that by describing a system that (I believe) will suit me.

• I want to be able to check for entry signals through the day/evening but not have to watch every minute.
• I want a system that is easy to know when to enter but i don't need to monitor the trade constantly once in the trade.
• I want a system that wins about half the time - and one that is not prone to large loosing streaks.
• I like trading with the trend.
• I like pulling the trigger and i'm not the best coder so discretionary works for me.
• I would like a system that produces 1 or 2 trades per week.
• I would like the option to increase my wins / decrease my losses.
• I would like something that I could apply to multiple forex pairs (in particular EUR/USD , USD/JPY & AUD/USD)
• I want a system that does not suffer large losses from one trade. (i.e black swan event immune)
• I want a system that if I go on holidays and miss a 1000pip move I dont regret it for the rest of the year.
• I like the idea of being able to trade a coin toss with better odds than 1:1.

You get the picture - So from this I can translate it into real parameters:

Timeframe: 30min.
Entry: signal or indicator.
Exit: Take profit / stoploss.
Entry method: Discretionary.
Profit probability %: approx 50%.
Max consecutive losses: approx 5.
Trades per month: 10.
Pairs: Inital testing on 1 but ultimately multiple.
Position sizing: Fixed percentage of a/c.
System basis: Trend following.
profit factor: 2 approx

Additionally I wanted to touch upon position sizing (money management): as this is one of the hardest things to manage!

Ok so if you do a Google search for:

• Trading position sizing - 1,580,000 returns
• Money Management - 192,000,000 returns
• trade size - 89,000,000 returns

So as you can see its something that is talked about alot - however its one of those topics that is completely tied into one's psyche. The first system i traded was one of the usual stories: Paid my $$'s , went to a conference for 3 days, got taught the magic system, got email / sms trade entries, and put in my orders.

The one thing they did not teach in the magic system was how much money to put in! We'll I had my account fully loaded, and I had the magic bullet system, so why not put 50% into 1 trade?? Sounds good to me, I have the experts on my side - Massive growth here I come! .

What.... loosing trade? ... Huh another loosing trade WTF!@! I paid for my course, no one told me the market could do what it wants!!! PRICK OF A MARKET! ohhh another sms! hold up I cant pull the trigger to enter the trade because suddenly... i'm scared!

This happens so easily - and to be honest, I still don't know if that system is any good!

So first of all your position sizing ties you psychologically to your trading system- It contributes to how much you win, and how much you lose. It helps compound winners, and hopefully protects you from losing everything in a losing streak. Your money management strategy can even go as far as telling you to sit the next trade out!

My strategy is simple: I use fixed % position sizing.

When you are winning, you compound your winnings.
When you are losing, you defend your capital by naturally decreasing your position size.

Some systems I allocate a large chunk of $$ to and then use small % position sizing.

Example: 10k initial investment, 10% position sizing, 150 take profit / 50 S/L


Other system I start with a very small initial investment and use 100% position sizing.

Example: have 100k account, but only invest $100 (1% risk) into the system. 100% position sizing, aim to double account size, or loose 50% per trade. Goal to get a string of wins. Compounding interest with this is magical if you hit a hot streak. For example - this is 12 trades win from starting value of $25 ---- $50 / $100 / $200 / $400 / $800 / $1600 / $3200 / $6400 / $12800 / $25600 / $51200. And hit that streak and you were only risking 0.25% of account to receive massive % gain (haha cant be bothered working out) oh and if you dont believe its possible - Google search: mystic genie (and click the macrotactics link - it'll get you to the right forum)

It really depends on how the system comes across in its testing stage as to what fixed % i will use.

Unfortunately my testing software (forex tester 2) does not include a method for doing fixed % position sizing.

Anyway im off for now to toy around with a few indicators / ideas for my next step!

Do have a look here : http://www.traderszone.com/trading-b...imisation.html

Also a Trading Plan Template here : RapidShare: 1-CLICK Web hosting - Easy Filehosting

Look forward to further posts from you.

Nice work doczero!

Your entry method: Discretionary.

That's IMO one of the big mistakes traders make. They think they can 'wing' the entry. The traders (real traders, doing this for a living) I know all have a very detailed plan on when and where to enter. Some have a few setups, others have a lot of setups but they all have their strict rules for entry.

Very good info doczero.

Entry is important and a lot easier then the exit/targets. Mostly because your judgement gets clouded while in a trade.

You have addressed the 2 major parts for judging a trading system: risk vs reward and accuracy. You need both to calculate your expectancy. If this is positive you have a winning system. I have seen successful traders with high accuracy but lower risk vs reward (scalpers) and vice verse, lower accuracy but high risk vs reward (swing traders). Never I have found a method or trader that can have both.

I'm a low accuracy but high risk vs reward kind of trader myself.

So I figure with this post we might outline a couple on indicators we might try out. As I outline in my goals: I want a system that is easy to know when to enter. Additionally I stated that i would be crafting a trend following system.

When looking at trends, I believe you need to look into - A) Direction of trend, and B) strength of the trend (keeping in mind you have to be thinking about your Timeframe- do 1 or 5 min trends exist??)

I have been doing a lot of reading from Charles "Chuck" LeBeau - recently I was reading an article regarding the misconceptions of the ADX ( Direction is the Key to Using ADX Correctly ). Chuck states in a lot of his writing that the ADX is a highly reliable and robust trading tool. The readings have sparked my interest to try out this indicator.

So A bit about the ADX - it was originally introduced in J. Welles Wilder's book New concepts in technical trading systems. The ADX (Average directional index) comprises of:

• The +DMI line and the -DMI line. This is used for picking direction.
• The ADX Line. This is used to show the strength of a trend.

So there is heaps of information out there if you want to read up on it.

I think I will also be using 2 MA's for direction. One longer / one shorter.

So the next step from here, is to put these two indicators on my charts and then try and find some visual confirmation that I can get these to work together.

I Will be looking for a common setup with them before a move. I'll be looking at the size of the move. This isnt testing - however its just getting a feel for the tools before we put some rules to them.

OK- So I've had a play around with the ADX / Moving averages on my screen - and I think I've settled on the following:

• ADX set to time period - 30, Both +DI / -DI have been set to invisible so I only use the ADX line as the trend strength.
• MA's being used are the following: SMA 30 / SMA 15

So once your setup on the 30 min charts, you'll look something like this:





From here we can pick entry signals. Going back to what Chuck Lebeau's article, the ADX value was not critical - what mattered was how much acceleration the ADX had from the previous bar. The way I will be using this is entering on what I call an ADX Hook. Where the ADX has turned and is starting to accelerate. It will look like this:




And here again are a heap of entry points:




Next step we will outline some rules to use this entry with.