[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Metastockusers] Re: Scott---Building a Profitable System



PureBytes Links

Trading Reference Links


@Scott,

ok - I try to show an example.

I've uploaded a file "mcs_example.gif" in the files section. Pleas 
look at the screenshot and read the explanation below.

Explanation:
1. you have tested EOD data with Metastock and use the results as 
input parameter for "zen monte carlo simulator v3"
2. your system test data was 1.000 OHLC bars (nearly 4 years of EOD 
data)
3. you have 100 trades in these 4 years (60 winning = 60% and 40 
losing = 40%)
4. your average profit per winning trade was 100 points (because we 
tested DAX index future) and your average loss per losing trade was 
75 points.
5. these system test results of Metastock resulted in the blue 
fields for TR (trade ratio = 60% winners...), PR (Pay off ratio: 
1,33) and PX (Profitability index 2,0)
6. each point of this DAX index future ha´s a currency value of 25 
Euros (if you test stock markets this value is 1 and you have to 
type the dollar or euros amounts in the average profit/loss fields)

...until now all the data come from your own test process in 
Metastock etc.!

Now you make two additional inputs specially for the MCS process:

7. you select a simulation range of trades. In this example I select 
100 trades - that simulates a trading of 4 years, because the system 
has generated only 100 trades with 1000 bars = 4 years (EOD)

8. you select the number of simulation runs

(in reality it's better to select a very high number, but the 
freeware is restricted to 5.000 simulation runs. Here I selected 
only 1000 runs. That is: we simulated 1.000 (4 year lasting) trading 
periods.

As Result we have the min, max, average values of the simulation for 
profits and accout drawdowns and additional estimated values 
(statistical reliability of two standard deviations in the freeware 
version)

In our example the system generates an average annual simulation 
profit of 18,796 Euros, but the statistical estimated annual profit 
is only 8.186 Euros.

More Important: you have to expect an account drawdown of estimtated 
9.515 Euros. That means that you need an account of minimum 9.000 
Margin for the DAX index future) + 9.515 + some intraday range 
amounts => minimum 20.000 Euros to trade this system with that 
Future. 

And even more, more important: in one simulation the account 
vdrawdown was 25,625 Euros, so that you need nearly 40.000 Euros to 
trade only one contract of the future with your system!


I hope, you can now see how to use the simulation software. It's a 
safety belt for your trading system development career - but you 
have always to think, that the future has this "little" piece of 
uncertainty and there a bunch of situations, where even a safety-
belt don't help...

bye,
zentrader

--- In Metastockusers@xxxxxxxxxxxxxxx, "Scott Mariani" 
<mariani@xxxx> wrote:
> Zentrader,
> O.k. I remember now. I did try the freeware version before. I 
couldn't figure out exactly how to use it, I still cant. I read the 
mcsquickref_e.pdf but the different input parameters for MCS3.0 are 
not clear to me.
> Scott
>   ----- Original Message ----- 
>   From: zentrader22 
>   To: Metastockusers@xxxxxxxxxxxxxxx 
>   Sent: Sunday, March 13, 2005 5:09 AM
>   Subject: [Metastockusers] Re: Scott---Building a Profitable 
System
> 
> 
> 
>   Scott,
> 
>   i take this part of the message of superfragalist:
>   "The percentage winners is really not the problem. The bigger 
issues
>   are the profit per winner vs loss per loser, the frequency of 
trading
>   and the drawdowns."
> 
>   Exact this was the reason for me to develop a monte carlo 
simulation 
>   tool (as an additional tool for system tests with e.g. 
Metastock), 
>   which takes the system tester results and simulate possible 
future 
>   scenarios. So you can see easily the dependencies between 
percentage 
>   winners, profits and losses per trade and the trading frequency 
on 
>   one side and the expactions of future profits and possible 
drawdowns 
>   on the other side.
> 
>   There is a freeware version, which works for EOD:
>   http://www.zentrader.de/download.html
> 
>   I've also downloaded the english version of Zen Monte Carlo 
>   Simulator v3 freeware into the files section of this usergroup.
> 
>   Try it!
> 
>   good luck,
>   zentrader
> 
> 
>   --- In Metastockusers@xxxxxxxxxxxxxxx, "superfragalist" 
>   <jackolso@xxxx> wrote:
>   > 
>   > Read this entire Charlie Wright article from beginning to the 
end. 
>   In
>   > the later parts he shows you how to develop a system. His 
simple
>   > system trades very well. 
>   > 
>   > http://www.elitetrader.com/tr/index.cfm?s=17
>   > 
>   > To help you understand more about what you are doing, read a 
few
>   > systems development books. I recommend reading all of them. 
But you
>   > can start with books written by Kaufman, Stridsman and others. 
I've
>   > learned more about trading from systems development books than 
the 
>   pop
>   > books written by the "guru's" who don't trade much at all. 
>   > 
>   > The percentage winners is really not the problem. The bigger 
issues
>   > are the profit per winner vs loss per loser, the frequency of 
>   trading
>   > and the drawdowns. 
>   > 
>   > Roy's newsletter is about to start a series of articles 
written by
>   > traders that will describe how they trade and what MS tools 
they 
>   use.
>   > You could benefit from that a lot.
>   > 
>   > www.metastocktips.co.nz
>   > 
>   > You could also benefit from the back issues of MSTT which 
explain 
>   how
>   > to use the systems tester, etc. 
>   > 
>   > I've developed a number of systems and it's not very hard to 
do.
>   > What's hard to do is to learn when to use each of the systems. 
No
>   > single system works well in all market conditions. That's a 
really 
>   big
>   > mistake people make.
>   > 
>   > I have a couple of systems for up trends, down trends, sideways
>   > typical and sideways with rotation. Each one has a different
>   > expectancy and different values for P/L ratios etc. 
>   > 
>   > After I use the system to find high probability trades I fiter 
the
>   > charts through my best computer--the one in my head. In 
general 
>   that
>   > improves the ratio of winners to losers by about 50%. In other 
>   words
>   > if my system has a 50/50 ratio, my evaluation of the charts 
>   improves
>   > the ratio to 75/25, and often better. 
>   > 
>   > Can you reduce the chart evaluation to a mechanical system---
no. 
>   You
>   > need experience to evaluate the charts, especially in the four
>   > different market conditions. 
>   > 
>   > A very simple and very profitable method of trading is simply 
to 
>   use a
>   > prefiltered list based on TA, momentum and fundementals--the 
>   valueline
>   > T1 stocks, the IBD list or the S&P neural fair value list. 
>   > 
>   > Use a moving average of any type, and one simple indicator 
like the
>   > IFT of the RSI (Roy's newsletter). Buy the momentum stocks on 
one 
>   or
>   > two of those lists whenever they're above the MA and the IFT, 
if 
>   it's
>   > above it's thresholds. That's all you have to do to make money.
>   > 
>   > That being said, 98% of the want-to-be traders want to do 
exotic
>   > explorations of 3000 stocks a night for patterns, breakouts or 
>   other
>   > crap they can't even define muchless find. Of course they're 
going 
>   to
>   > have 40% or 50% winners when the market is in an uptrend. When 
the
>   > other three conditions exist they're only going to have 20% 
>   winners. 
>   > 
>   > If the market is in an up trend advancing stocks always lead 
>   declining
>   > stocks, so a monkey could throw darts at a copy of the wall 
street
>   > journal and wind up picking 50% winners. 
>   > 
>   > Trade from the defined lists, learn to read the charts of the 
>   momentum
>   > stocks so you can tell when the momentum is likely to continue-
-
>   there
>   > aren't any indicators that are going to tell you that any 
better 
>   than
>   > your eyeball after you've looked at something around 1000 
momemtum
>   > stock charts.
>   > 
>   > There are simple methods to use in all four market conditions. 
I 
>   think
>   > Roy will have articles about the methods in up coming 
newsletters.
>   > There isn't the time or space to post everything every newbie 
>   should
>   > know--not to mention having to repeat it one hundred times. 
>   > 
>   > Start with the Charlie Wright articles and work your way up. 
>   > 
>   > We are in a sideways market with rotation or a slight 
downtrend 
>   with
>   > rotation. That's the worst market there is to trade in so 
buckle 
>   up.
>   > On Friday, I saw some signs that the rotation might be about 
to 
>   stop.
>   > If it does, I'll determine what kind of market we have after 
that 
>   and
>   > then use the appropriate system. This is good time to read 
because
>   > without a lot of experience you aren't going to be making any 
money
>   > right now anyway in this kind of market. Maybe after August. 
>   > 
>   > If you don't believe me regarding market conditions, here's a 
>   simple
>   > test. Set up the explorer to find stocks that have gone up two 
>   days in
>   > a row. Run it everyday for the last two weeks and see how much 
over
>   > lap there is between the lists on a daily basis. For this test,
>   > confine your lists to the S&P 500, 400 and 600. That's 1500 
stocks.
>   > Now if you can't find many stocks that have gone up two days 
in a 
>   row,
>   > and with lists that overlap, it's time to read a book and stop 
>   trading. 
>   > 
>   > Have fun!
>   > 
>   > 
>   > 
>   > 
>   > 
>   > 
>   > 
>   > --- In Metastockusers@xxxxxxxxxxxxxxx, "Scott Mariani" 
>   <mariani@xxxx>
>   > wrote:
>   > > O.k.
>   > > I have been building system test after system test, trying 
>   several
>   > different combinations of indicators. Since I use EOD data I 
have
>   > given up on getting accurate exit values. I get a bunch of 
high 
>   losses
>   > which I attribute to the EOD data not exiting until the close.
>   > > 
>   > > I assume the object of a system test is to have more winning 
>   trades
>   > than loosing ones? I have resorted to adding one optimization 
so I 
>   can
>   > look at the overall system performance on a total $$ basis. 
The 
>   only
>   > problem with this is that with the exits being what they are, 
my
>   > losses are humongous. I have been trying to find a system that 
is
>   > better than 50/50 but have yet to stumble on anything that 
yields 
>   more
>   > winners than losers.
>   > > 
>   > > How do others go about analyzing system tester results?
>   > > Thanks, Scott
> 
> 
> 
> 
>         Yahoo! Groups Sponsor 
>               ADVERTISEMENT
>              
>        
>        
> 
> 
> -------------------------------------------------------------------
-----------
>   Yahoo! Groups Links
> 
>     a.. To visit your group on the web, go to:
>     http://groups.yahoo.com/group/Metastockusers/
>       
>     b.. To unsubscribe from this group, send an email to:
>     Metastockusers-unsubscribe@xxxxxxxxxxxxxxx
>       
>     c.. Your use of Yahoo! Groups is subject to the Yahoo! Terms 
of Service.





------------------------ Yahoo! Groups Sponsor --------------------~--> 
What would our lives be like without music, dance, and theater?
Donate or volunteer in the arts today at Network for Good!
http://us.click.yahoo.com/Tcy2bD/SOnJAA/cosFAA/zMEolB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/Metastockusers/

<*> To unsubscribe from this group, send an email to:
    Metastockusers-unsubscribe@xxxxxxxxxxxxxxx

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/