Breakout Trading Searching Methods?

[Newt]

Fairly new to trading, am currently breakout trading on paper. Short term stocks, about 1-4 weeks. But it seems my biggest issue is actually searching for stocks, none of the books or tutorials I've read really cover this properly. I use Market Analyst 7 and I have no clue what search parameters to use when looking for stocks that have just broken out of consolidation.

Tech and W, suspect after MANY years of trading experience you're probably thinking in very different way to krampster. When you're a newbie (heck, I'm pretty sure I'm still one) just having proof that trading is viable is a big deal. Agree experienced traders may well know how to manage an existing position toward expectancy of profit. Suggest Newbie thinking more likely to be "do I even get in the water?"

 

[tech/a]

It's an important part of developement as a discretionary trader.
A starting point for systematic development.

A breakout can become many things
(1) a short term movement in price
(2) the beginning of a trend
(3) the continuation of momentum
(4) reversion to a mean.

So the breakout itself can function in many " trading styles " and
Methods.

Put your foot in the water---dive right in!

 

[RADO]

Re: Break out Trading Searching Methods?

Is Amibroker popular? Would prefer to use something with a big user base to make learning how to use the program easier and to know that it's trusted.

I think Ninjatrader would probably have the biggest following out of all the platforms. But Amibroker is unique in its own way imo it offers the best value for money and there customer support via email is very good. If you do get a copy of amibroker this website: wisestocktrader.com will have just about everything you will ever need, there's heaps of free coded trading systems and scanning tools.

 

[Valued]

I would caution you about using other peoples' code and trading tools without actually understanding them first. If they use complex formulas you do not understand, you probably shouldn't start trading with them.

I do not understand programing and do not understand math beyond basic algebra that you would learn in middle school (I never took math past grade 10). I still like Amibroker. I just bought it a few days ago. It's relatively cheap. Since I can't do anything myself, I had to get the AFL Wizard too since I wanted to use it as a replacement for stockscan. This way I can just click on things and select them from a list. It's obviously more basic but it means I won't have to spend a monthly fee on stockscan which was even more basic.

The charting does what I need it to do, especially since I do not believe in elliot waves, gann, fib or patterns (although Amibroker will do all this stuff). It makes it very easy to select charting packages!

 

[tech/a]

The charting does what I need it to do, especially since I do not believe in elliot waves, gann, fib or patterns (although Amibroker will do all this stuff). It makes it very easy to select charting packages!

This " belief " you have
How have you developed it?

What do you have a belief in?
How have you developed that?

 

[Valued]

This " belief " you have
How have you developed it?

What do you have a belief in?
How have you developed that?

Easy. There is no evidence that suggests that any of these indicators are better than random entries. There is no evidence that there is a positive expectancy to them. One would expect that if you already have a profitable system, then adding on another indicator would increase positive expectancy on back tested data. There is no evidence that it does. The evidence may be there, but no one has found it and published it for definitive proof. Even if it increased the positive edge by say 2% this would be significant. It would indicate it doesn't work that much of the time but if you used it you would expect to make 2% more than someone who does not. The only way an indicator can work is if it increases expected value (+EV or positive expectation). It may be that one indicator is only +EV when combined with another. An example of this would be volume. This may be the case for other technical indicators too. Given how many there are, there are far too many combinations to assess. A computer could but then you would need the knowledge to code them all and have a computer compare every possible combination of indicator to back test data across different markets.

Without evidence, belief is just faith. I do not need to establish a reason for believing in a negative, it's just simply being skeptical until I have proof. Now something may make intuitive sense or adds up as being reasonable or logical and we may try it out without hard proof anyway since that's how humans are wired. The issue is things like Gann lines or Elliot waves offer no intuitive explanation for why it should work. We need hard proof this is the case. For this, the premise must be testable. You must be able to define what you do. I am not saying you can't make money from subjectively eye balling a chart and then reaching a conclusion. It would be a rare individual who can analyse more than a few variables like that but there are people who may be able to do it. The issue is that others cannot replicate it. We need to know exactly the parametres to test.

 

[tech/a]

Your reply is one I've seen many times before.
And would be my own reply if I was in your position.

Infact my arguement with ALL analysis is that PRACTICAL
APPLICATION of analysis is sadly lacking.

A great deal of theory and hypothesis but little profitable
Application ----- particularly evidenced based.

There are exceptions.
Dr Bruce Vanstone has done some fantastic work
In evidenced based method application.
Nick Radge has volumes of evidence of profitable application
Of analysis on his website ---- available for all to see.
Righly so it's only available to members. Yes I'm one.

Then I have 20 yrs of practical application of my own --- so I've
The belief and proof ---- I hope one day you find it also as it is empowering.

BUT
I've come to the conclusion that while I'm lucky enough to be capable of rudimentary
Systems design ---- I'm just never going to be up to speed with the likes of say
Howard Bandy or Bruce Vanstone. I guess you and many reading this are in the
Same boat.

I am lucky enough to have a Son who is a Doctor of Physics --- Kris.
Kris has just broken free of conventional employment and now heads up a
Start up Company with 2 other Doctors.

I'm also lucky enough to be able to employ people far more intelligent than me
To help me in business ---- and with so many questions in the field of investment that
Both Kris and I have. I have employed him ( as one project of Many he and his team are working on)
To join me in the search and finally application of an automated neural network ( fuzzy logic is PERFECTLY suited to technical analysis methodologies ) method
Which trades automatically in the background while we both take advantage of the time we have
On this planet to actually ---- enjoy it.
Dr Vanstone has already proven massive increases in return with the introduction of neural network computing.

So some of the questions you have I already have the answers to--- well enough for me to
Put my hard earned up for trading.
Others I have engaged the help of experts to review and work with those who have studied and answered the
Same questions we all have had from time to time.

 

[krampster2]

Oh wow, thanks for the replies, thought this thread would not continue. I guess I'm to used to forums where threads only last a few hours.

Anyway, I'm trialing a whole bunch of software and on Incredible Charts I made a basic scan. I know it's not great, but I'm just experimenting.


I found two interesting stocks, the first, SRX, has a big gap and volume support. It's also closing above previous high from late September. The second SDG, does not have a gap on it's breakout nor does it have overwhelming volume. So I'm interested to see if it can succeed. What can you guys make of this?




EDIT: My mistake, was not sure how to attach images without having show in the post.

 

[Newt]

Hi Krampster

You can probably answer your query on these 2 stocks yourself 7 days later. Sorry I haven't looked at them closely (I can tell at a glance they're not the sort of pattern I like to chase). If you're serious about a breakout system its worth researching until you grasp the likelihood of:

- roughly 50% (or lower) win rate
- able to take accept many losses for small number of winners
- you must take all trades or risk missing the small number of "winners"
- you have much greater odds of success with a defined system with backtested and out-of-sample data
- at some point you may well move on to become a successful discretionary trader
- how many lots of what size with what risk should you trade to minimise the likelihood of blowing up

Many people find it very hard to keep at such a method consistently over long periods, waiting for a small number of winning stocks, so lastly, does your personality and expectation match you're considering?

 

[krampster2]

Hi Krampster

You can probably answer your query on these 2 stocks yourself 7 days later. Sorry I haven't looked at them closely (I can tell at a glance they're not the sort of pattern I like to chase). If you're serious about a breakout system its worth researching until you grasp the likelihood of:

- roughly 50% (or lower) win rate
- able to take accept many losses for small number of winners
- you must take all trades or risk missing the small number of "winners"
- you have much greater odds of success with a defined system with backtested and out-of-sample data
- at some point you may well move on to become a successful discretionary trader
- how many lots of what size with what risk should you trade to minimise the likelihood of blowing up

Many people find it very hard to keep at such a method consistently over long periods, waiting for a small number of winning stocks, so lastly, does your personality and expectation match you're considering?

I'm reading several resources on breakout trading at the moment and over the next year or so plan on putting together my own system like I have seen so many people recommend. I'm objective and think I would be less susceptible to the emotional roller coaster that trading takes you on. It's interesting what you said about being able to accept having a less than 50% win rate. So the risk management system you recommend is to aim for betting more on winning trades and putting less on trades with less potential right? But what about putting roughly the same amount on each trade and focusing more on having a high win rate? I've heard that those are the two main ways to do risk management, put more money on your trades more likely to win and not focus on win ratio so much. Or apply the same amount of money to each trade and focus on having a positive win ratio. Why do you use the system you do?

 

[Newt]

I personally aim to trade > 10 positions with each a fixed percentage of the available pool and the same risk management on each.

The more positions you trade:
- the less likely a series of losing trades will take you out
- the more work you have to do entering and exiting positions
- less chance of not being on a winner
- less % win from the rare winners

There are numerous way of setting position size. Many advocate have less (or smaller) positions when you're not doing so well, but ramping up size and risk when on a winning streak.

Howard Bandy writes some great books on this sort of thing BTW.

 

[Valued]

If you want to use a system without doing any work, you can always just do whatever Nick Radge is doing. He designs systems and they have proven to be profitable in back testing (although that does not mean they will always be profitable). His website is The Chartist or you can just get his book Unholy Grails. He seems to be highly recommended around here. Personally, I think there is no evidence of price patterns and elliot waves so I am skeptical of the methods. I would at least buy Unholy Grails though so you get an idea of how simple profitable systems can look like.

I think it would be hard to design a breakout trading system and far easier just to look at the breakout yourself and use your discretion. It's very hard to define trading ranges, price spreads, volume and then factor in whether this is after an extended upmove or an extended downmove. After all of that, you then need the system to dictate when the breakout has failed - how many bars and how wide spread and on what volume has a breakout failed? Difficult stuff to code into a computer system unless you have significant experience in that regard or can afford to pay someone to do it for you.

 

[Valued]

I personally aim to trade > 10 positions with each a fixed percentage of the available pool and the same risk management on each.

The more positions you trade:
- the less likely a series of losing trades will take you out
- the more work you have to do entering and exiting positions
- less chance of not being on a winner
- less % win from the rare winners

There are numerous way of setting position size. Many advocate have less (or smaller) positions when you're not doing so well, but ramping up size and risk when on a winning streak.

Howard Bandy writes some great books on this sort of thing BTW.

By trading more though, you increase brokerage costs. You have to be trading with a large account. Even if you were trading with 20k you would lose on brokerage trading a couple of hundred times per year. Even at $10 brokerage you would be losing 10% for 200 trades. Further, you increase the chance of picking bad prospects just to maintain your quantity of trading. You might feel forced to take razer thin profit edges that just add to down draws and may only increase ROI of less than 1%, or worse losing money on some trades. Once you factor in taxation, some trades may not even be worth the risk.

 

[krampster2]

If you want to use a system without doing any work, you can always just do whatever Nick Radge is doing. He designs systems and they have proven to be profitable in back testing (although that does not mean they will always be profitable). His website is The Chartist or you can just get his book Unholy Grails. He seems to be highly recommended around here. Personally, I think there is no evidence of price patterns and elliot waves so I am skeptical of the methods. I would at least buy Unholy Grails though so you get an idea of how simple profitable systems can look like.

I think it would be hard to design a breakout trading system and far easier just to look at the breakout yourself and use your discretion. It's very hard to define trading ranges, price spreads, volume and then factor in whether this is after an extended upmove or an extended downmove. After all of that, you then need the system to dictate when the breakout has failed - how many bars and how wide spread and on what volume has a breakout failed? Difficult stuff to code into a computer system unless you have significant experience in that regard or can afford to pay someone to do it for you.

Yeah I'm not a very experienced programmer, and I have been weighing up the benefits of a very tightly guided trading system which operates within strict rules compared to a more flexible system. I'll give that book a look, thanks for the resource.

 

[krampster2]

By trading more though, you increase brokerage costs. You have to be trading with a large account. Even if you were trading with 20k you would lose on brokerage trading a couple of hundred times per year. Even at $10 brokerage you would be losing 10% for 200 trades. Further, you increase the chance of picking bad prospects just to maintain your quantity of trading. You might feel forced to take razer thin profit edges that just add to down draws and may only increase ROI of less than 1%, or worse losing money on some trades. Once you factor in taxation, some trades may not even be worth the risk.

As I'm new I'm stepping in tentatively so am not investing much initially. So trading few trades may be the best way to maximise profit. Don't like the thought of the burden of the brokerage costs looming over every trade I make. On the other hand it seems daunting making so few trades.

 

[Newt]

By trading more though, you increase brokerage costs. You have to be trading with a large account. Even if you were trading with 20k you would lose on brokerage trading a couple of hundred times per year. Even at $10 brokerage you would be losing 10% for 200 trades. Further, you increase the chance of picking bad prospects just to maintain your quantity of trading. You might feel forced to take razer thin profit edges that just add to down draws and may only increase ROI of less than 1%, or worse losing money on some trades. Once you factor in taxation, some trades may not even be worth the risk.

Therefore, breakout/trend trading may not be a good idea without a large account size (>$20,000 perhaps, but many would argue on decent account size and suitable position sizing). Bottom line - you have to be pretty sure how your planned approach will survive the tough times that happen when you least appreciate it.