How would you improve this Pullback Trend Following System?

babymush

Active member
Messages
221
Likes
1
A trend following system like mine performs best in liquid stocks that tend to trend in a relatively stable manner. The good liquidity enhances the effectivness of stop losses. (stops tend to become less effective in thinly traded stocks.)

The template I put on my WEEKLY charts is...

* 11 period ADX with a horizontal line set at 25.
* 3 period CCI histogram with a horizontal line drawn at the zero level.
* 13 week and 26 week EMA's.

Here's what I do in my once a week analysis each Saturday morning...total time spent is about an hour to analyse the 600 US blue chips that I follow.

First, check the WEEKLY chart of the All Ords (or S+P 500 if you trade US stocks). If the 13 EMA is above the 26 EMA, both are rising and pulling further apart, and the tops and bottoms of the price action are getting progressively higher, then you know the overall market is bullish and you ABSOLUTELY MUST trade ONLY long positions.
The mirror image of these rules apply to identify a downtrend. If the overall market is downtrending, you ABSOLUTELY MUST trade ONLY short positions.
I cannot overstate the benefits of being in tune with the overall market...its VERY important.

SCANNING.
If the weekly trend of the All Ords is bullish...Scan for 'CCI crosses below zero' This scan finds stocks that are in the process of retracing to the downside.
OR
If the weekly trend of the All Ords is bearish....Scan for 'CCI crosses above zero' This scan finds stocks that are in the process of retracing to the upside.

With just a couple of mouse clicks the stocks picked by this scan can be put into a separate workbook. My weekly template is put on these stocks. Now I look at each stock one by one, but at this stage I'm only interested in the ADX, not the price action. As each chart comes up I'm looking at the bottom right corner of the screen to see if the ADX is above the horizontal line which is set at 25. It takes only a one second glance to check the ADX reading, and if its not strong enough, a further two seconds to put the next chart up on the screen. It sounds inconceivable that you can check out a new chart every three seconds, but remember that the process so far has been mainly to weed out the undesirables rather than to pick the winners.
If I come across a potential winner (one whose ADX is reading 25 or higher), I then check it's price action, Moving Averages etc, and if it shapes up OK in those areas I transfer it to a workbook which I've called 'Watchlist'....this takes just a couple of mouse clicks.
Out of my workbook of 600 US blue chips the scanner usually finds about 10% - 60 odd stocks, that are giving the CCI signal. At less than 5 seconds per chart it doesn't take long to whiz through this list and cull most of them, and save the best 8 or 10 to a watchlist.

The final selection on the top 8 or 10 is purely visual. If the overall market is bullish I'm looking for the stock to have....
* Higher tops and higher bottoms, 13 EMA above 26 EMA, daylight (space) between the two MA's, both MA's rising and pulling apart.
* ADX reading of 25 or higher.
* Stock must be in the process of a pullback.

ENTRY...
Put a buy stop or a buy limit order 1 tick above the candle that pulls CCI below zero. Lower the buy order each week if necessary to just above the top of the last candle. If/when the uptrend resumes, your entry order will be filled.
OR
For a more conservative entry, put the buy order 1 tick above the first bullish candle following the pullback.
All buy orders are placed on a GTC (good till cancelled) basis.

STOP LOSS...
Attach an 'If done' stop loss order to your buy order. If your buy order is done (filled), your stop loss will automatically be put in place.
I like to place the stop somewhere between 8% to 10% from entry, but I've used a 5% stop on occasion. 8 to 10% gives the market a bit of room to fluctuate without stopping you out.
For long positions its prudent to pay a small extra charge to use a GSO (guaranteed stop loss order), if your broker offers this facility. A GSO is designed to protect you against a severe price slump that could gap past your stop loss order.

TRAILING STOPS.
As the trend moves in our favour we need to progressively lock in profit. I do this by moving my stop up to just below each swing low on the weekly chart. Each swing low is of course preceded by a swing high. When each pullback ends, and the revitalised uptrend moves above the last swing high, then and only then will I move my stop loss up to 2% under the swing low.
I repeat this process each time a higher swing low is made, until finally the trend dies and my stop is hit, and I'm taken out of the trade.
There's one other part to my trailing stop strategy....if a weekly candle penetrates the 26 week EMA, at the start of the next week I'll move my stop up to 2% below the bottom of that candle. A penetration of the 26 EMA can sometimes be a warning that the trend is getting tired, in which case its prudent to take the precautionary action of moving your stop up to just under that candle.

MONEY MANAGEMENT.
2% of my trading capital is the maximum I'm willing to risk on a trade.

SUMMARY.
Use weekly charts exclusively.

Once a week check the weekly trend of the overall market, and trade only in that direction.

Find stocks that are trending in the same direction as the market as a whole, but are currently retracing.

Use the ADX to measure the trend strength of these stocks.

Buy these stocks as they finish their retracement and resume their trend.

ALWAYS place a stop loss order, and make it a guaranteed stop loss if you're trading long.

Use trailing stops to progressively lock in profits as the trend moves your way, and to take you out of the trade when the trend dies.

Use the 2% money management rule to prevent serious depletion of your trading capital from a losing trade.

If the overall market is bearish, use the mirror image of the system to go short in bearish stocks.

CONCLUSION.
By cutting losing trades quickly and letting winning trades run as long as the trend remains healthy, by trading only in strongly trending stocks, and by trading only in the direction of the overall market, this system virtually ensures an acceptable ratio of winning trades to losing trades, and that the average profit will be considerably larger than the average loss.
Its difficult NOT to make money with this type of system, PROVIDING its applied consistently.
The icing on the cake is that a large number of stocks (600 in my case) can be analysed in just an hour once a week.
 
BM - I really don't want to rain on your parade, but where did THIS all come from?

A few posts ago you were asking for someone to suggest a profitable trading system.

Is this it?

Confused... :confused:
 
Started to read your post m8y but got bored after the first sentence. You couldn't paraphrase it could you ?

But I did actually catch the bit about 2% risked on any one trade. I wouldn't even risk 0.5% of my capital on any one trade. I find 2% too much to lose in one go and you wouldn't want 50 consecutive losing trades would you, lol ?

Yawn, roll on Monday.
 
Have you got historical performance data for this method?
What software do you use to scan so many charts so quickly. Nothing I know of, without some bespoke coding, would allow me to do that.
Only comment would be that by the time you get a signal on this system you are likely find the trend is already too mature. A 13 & 26 week EMA cross as a starting point is certainly too long term for my liking.
 
twalker,

Using Prosuite you can effectively "scan" hundreds of instruments for specific criteria without having to look at the specific chart unless I have misunderstood you ?


Paul
 
OK, I misunderstood "babymush" then, I thought he looked at each as it flashed across the screen and was able to click to drop them into specific directories accordingly.
I did not realise this function existed on prosuite however, that sounds useful. I will have a look at it, thanks
T/.
 
I'd improve your process and perhaps performance by using a good backtester such as AmiBroker. Put your indicators into a system and then backtest against whatever metrics you consider to be important (e.g., equity, drawdown, sharpe, etc.). Your universe of stocks will be tested in a very short time.
 
Salty,
Generally speaking, I agree with you and would prefer not to risk more than 1% tops of my equity. However, according to Alpesh B. Patel, babymush may not be risking as much as we think. In his book 'trading online', Mr Patel writes:
"Imagine you risk 2% of your equity on any one trade. Allowing for the fact that your account size drops on each occasion, your initial equity would be down 50% after 34 consecutive losses. What are the chances of 34 consecutive losses? Well, if there is a 50-50 chance of profit on any one trade then the chances of 34 consecutive losses would be 0.5 (3/4) or one in 17 billion. Makes you wanna trade doesn't it."
;)
Tim.
 
start again and find a true insight , optimising doesn't work , in my humble opinion.
 
Yeah looks ok, Babymush are you new to trading? As there's not much mention on how your gonna exit your trades. Its all to do with finding trades and setups, some stop loss work. But nothing on exiting trades that are in profit. You'll notice when your becoming an expert, as you tend to concentrate on exits much more. Even a random entry with a good exit strategy and money management (position sizing etc) can work.
 
Hi

Yes I am new to trading. :) Quite new. Where can I read more on Exit Strategies? Most that I read are only a trailing stop loss to exit the trade.
 
I was full of a sense of foreboding when I started reading it, but changed my view fairly quickly. It doesn't look bad at all to me, I must say. Are you going to tell us whose system this is that you've posted here, babymush? My guess is that if you monitor every trade closely and keep your trailing stop in a reasonable position (maybe to be derived as a function of the stock's ATR?) that won't be such a terrible exit policy. To try to come up with something more sophisticated will be really complicated, especially proving that it's better than the above plan. Do you have back-tested results, and over how long a period and what were the markets doing overall when they were collated?
 
one thing is certain good money management cannot make an unprofitable system profitable!
 
I am currently using Metastock and cannot seem to do a Backtesting properly - Any metastock users here can teach me how to use its system tester properly ..
 
Are you going to tell us whose system this is that you've posted here, babymush?
 
apparantly babybush a lot depends on which version , i'm 7.2 and backtesting is poor to say the least though i'm led to believe later versions are better!!
 
Top