Hi David,
Thanks for your reply. You have given me plenty to think about. I will try
out what you have suggested and let you know how it works out.
Thanks.
dbw451 <[EMAIL PROTECTED]> wrote:
If you are trading a trending or swing type system, try scaling
out with two targets. The first target is to capture a small profit and move
your stop to breakeven once triggered. This guarantees the trade will be
profitable once the first target is hit. The second target should be a logical
distant target at something like a fib number, support or resistance, or a
previous swing high or low. Once price moves significantly towards the second
target (maybe ½ to ¾ of the way), change your breakeven stop to a trailing stop
using SAR or an MA or a few ticks above/below the previous bar on a 5x higher
timeframe (e.g. 1 minute timeframe would be 5 minute bars, 1 day would be
weekly, etc.). The trailing stop allows you to lock in some of the larger
profit and if the market is trending well it will reach your second target. If
the market pulls back and takes out your trailing stop, so what? Youve locked
in profit, start searching for the next trade. The
second target should be at least 2x your initial stop size and preferably 3x
or more.
Once the first target is hit and your market or stock is trending well, you
have the option of taking a new entry signal with half your usual position size
and adjusting the second target further out. You can essentially take several
smaller target 1 trades while riding a big trend to your distant target.
Finally, implement a time based stop (note: this is not an n-bar stop). A
time based stop exits the trade at market if the entry is not profitable within
a predefined amount of time. If the trade is profitable at the predefined
time, the time based stop is ignored. This one trade management technique
enhances a systems performance more than any other that Ive tested.
This should give you plenty things to think about and test. J
Regards,
David
---------------------------------
From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of a a
Sent: 02/21/2007 5:49 AM
To: [email protected]
Subject: RE: [amibroker] Can this be done
Hi David,
Thank you for your excellent reply. You have hit the nail on the head. I do
manage my entries well but the exit is a problem. I have been toying with
parabolic SAR. Ema 3,6 crossover etc. but have not been able to find a sell
strategy which manages to differentiate between a bull-trap and an exit. Many
times prices decline only to go back to new highs (i.e bull trap) and then one
feels stupid for selling and at other times the decline is for real.
Any ideas on this problem?
Thanks.
dbw451 <[EMAIL PROTECTED]> wrote:
Wow, thats a loaded question. Im not overly familiar with the
RSI indicator, so I cant offer specific suggestions about what youre trying
to do. However, I can say that in general a system based on a single indicator
is only profitable when the market is conducive to that type of indicator. For
an oscillating indicator like the RSI, the market either needs to be ranging or
in a pull-back of a trend for the oscillator to be useful. Trying to buy an
oversold oscillator in a down trending market is a recipe for disaster.
Identifying market conditions is the hardest part of trading, IMHO. You need
to decide if your system is ranged based or trend based and focus on
identifying when the market is conducive to your system. In other words, focus
on whats going on in the left hand side of your chart.
The golden rule of trading is cut your losses short and let your profits
run. If you think about it, the golden rule is primarily about where to exit
a trade. If you put your focus on exits rather than entries, youll be much
better off. Good entries are important, however its been proven that a coin
flip entry with tight stops and good profit management can be a profitable
system. All thats needed to win this game is a positive expectation system
and the discipline to follow it. Chapter 5 of Van K Tharps Financial Freedom
through Electronic Day Trading is the best write up Ive read on the golden
rule and system expectations. The book also has a great chapter on how to
write a trading plan. These two topics apply to any type of trading (i.e. day,
swing, etc.).
Best of luck in your trading education.
Regards,
David
---------------------------------
From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of a a
Sent: 02/20/2007 2:52 AM
To: [email protected]
Subject: RE: [amibroker] Can this be done
Hi David,
Thanks for your help. Do you have any suggestions on how the trading
strategy can be further improved upon.
Thanks.
dbw451 <[EMAIL PROTECTED]> wrote:
AmiBroker can do this very well and you have a good backtesting
strategy. ABs optimization basically iterates a variable over a range of
values. For each variable you want to optimize, you specify a starting,
ending, and increment value. An AFL example would be:
dRSIvalue = Optimize("RSI value", 14, 4, 30, 1);
dRSI = RSI(dRSIvalue);
I dont think AI software is needed for what you want to do.
Regards,
David
---------------------------------
From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of
swptec
Sent: 02/19/2007 1:57 AM
To: [email protected]
Subject: [amibroker] Can this be done
Hi,
I am new to Amibroker and don't know even if I am wording my question
properly.
I want to optimize an indicator (say) like RSI for (say) a two year
period from Jan 2003 to Jan 2005 and then do an out of sample backtest
for the optimized indicator for the period Feb 2005 to Jan 2007.
Can such a thing be done?
Is it worth while trying such things?
Am I better of buy some Artificial Intelligence s/w for this purpose?
Thanks.
---------------------------------
Check out the all-new Yahoo! Mail beta - Fire up a more powerful email and
get things done faster.
---------------------------------
Need Mail bonding?
Go to the Yahoo! Mail Q&A for great tips from Yahoo! Answers users.
---------------------------------
Bored stiff? Loosen up...
Download and play hundreds of games for free on Yahoo! Games.
