|
Title: System39...was: 'Rule Based' versus 'Discretionary' trading...
Recently I have been pushing the idea that imo: trading systems do not need to be complex, that not all systems require special in-depth market expertise to trade, that mechanical rules can work, and that lots of experimenting can lead to success.
Below is a simple (one line of code really) EOD system idea that was traded with real money, 100% mechanical, many years ago, on a single stock that had a unique quality. This quality was found through testing lots of ideas that failed! This system was traded daily, for about 7 months, giving 80-90% winning streaks. The system failed when the company was bought out by another company. Obviously those trading it and giving me the pattern I needed, stopped trading it. The last few days we traded the system were losing days. We were so used to the long winning streaks that we couldn't believe it stopped working. I believe there are plenty of such trading anomalies around; they come and go, it is just a matter of finding them at the right time. To exploit temporary patterns you have to be looking ALL the time. There are no patterns that last forever.
The code below has been trimmed. I have no time to test it now but I think the basic idea is close to what we traded. You can run this on your DB to see what surfaces, then experiment and add frills (optimize,stops,portfolio,MM). DO NOT TRADE WITH REAL MONEY AS IS. Let me know if you can make it work for you.
Good luck,
herman
SetTradeDelays( 1, 1, 1, 1 );
BuyPrice = ShortPrice = SellPrice = CoverPrice = C;
thL = Optimize( "thL", 39, 0, 100, 1 );
thS = Optimize( "thS", 39, 0, 100, 1 );
S = ( LinearReg( C, 3 ) - L ) / ( H - L ) * 100;
Buy = Cover = Cross( thS, S );
Sell = Short = Cross( S, thL );
E = Equity( 1 );
Plot( C, "", 1, 128 );
Plot( E, "Equity", 1, 1 | styleOwnScale );
Thursday, August 21, 2008, 2:19:45 PM, you wrote:
> Bill,
> Those are interesting articles. Connors has some unusual views (and
> very expensive books <g>).
> Below is an example of how a perspicacious DT can gain a trading edge.
> It's a quote from Joe Ross' Trading Tidbits.
> IIRC, the S&P contract in 1997 had a tick size of 0.05 worth $25 and
> Joe's normal trading size for the S&P was 25 contracts (from "Trading
> By The Minute").
> How much money was Joe making?
> " In 1997 I experienced a fantastic winning streak day trading the S&P
> 500, prior to electronic trading, and prior to the e-mini S&P 500.
> In January of 1997 I noticed that at about 10:45 EST the market
> moved up about 200 ticks, four days each week, better than 90% of
> the time. It never happened on Friday, but Monday through Thursday
> were great. I had no clue as to why it did it or who or what caused
> it. All I knew is that it worked. So at about 10:40 EST I would
> begin watching. Within minutes the attempt to move up those 200 or
> so ticks would take place, and I would grab off about 100 ticks,
> call it a day, and head for the Bahamian beach.
> Then, sadly, in August of that year it stopped working. The exchange
> cut the S&P in half and introduced the E-mini. I have never again
> seen that phenomenon work, but believe me I have looked. "
> Bill
> --- In amibroker@xxxxxxxxxxxxxxx, "wavemechanic" <timesarrow@xxx> wrote:
> http://www.tradingmarkets.com/.site/stocks/commentary/lcbattlep/08272004-39801.cfm
> http://www.tradingmarkets.com/.site/swingtrading/commentary/lcbattlep/09022004-39899.cfm
>> ----- Original Message -----
>> From: sidhartha70
>> To: amibroker@xxxxxxxxxxxxxxx
>> Sent: Wednesday, August 20, 2008 10:52 AM
>> Subject: [amibroker] Re: 'Rule Based' versus 'Discretionary'
> trading...
>> I think you're right Brian. We do all use rules of some sort.
>> But I guess discretionary traders don't use 'hard and fast' rules and
>> can't always define the same set of rules by which they choose to
>> define an entry or exit.
>> For example, as we all know, something as simple as defining a trend
>> programatically can be more problematic as you might at first think.
>> However, a good trader can see very quickly what state the market is
>> in by looking at various time frame of chart.
>> Likewise, divergences of various sorts can be easy to see with the
>> naked eye but difficult to code in their entirety.
>> Like driving a car, or a golf swing, you learn the 'rules' but when
>> you get really good you are no longer thinking rules... you've
>> effectively let go of the rules and are just 'doing'...
>> --- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@> wrote:
>> >
>> > Here is my definition:
>> >
>> > We are all rule based traders.
>> >
>> > Mechanical Traders are a specialist group who have programmed
>> > computers to autotrade their rules OR automatically announce their
>> > rules via computer communications (audio, email, chart prompts,
>> > spoken text etc).
>> >
>> > I am prepared to continue the discussion with any seers,
> inituitives
>> > etc, who come forward, and adjust my definition to meet anything
> new
>> > that comes out of that.
>> >
>> > In advance I admit to the possibility of exceptions to the rule.
>> >
>> > brian_z
>> >
>> >
>> > --- In amibroker@xxxxxxxxxxxxxxx, "brian_z111" <brian_z111@> wrote:
>> > >
>> > > >Descretionary traders make decisions that are based on personal
>> > > >knowledge and circumstances, perhaps using many factors
> unknown to
>> > > >themselves. Like which journal they read the night before.
>> > >
>> > > This is the nub of the question for sure, and the point that I am
>> > > investigating.
>> > >
>> > > I suspect that when they (self-nominated DT's) think they are
>> > making
>> > > discretionary decisions they are in fact making rule based
>> > decisions.
>> > >
>> > > That is why I asked for specific examples of 'discretionary'
>> > decision
>> > > making e.g. I haven't seen Bilbo's chart yet but I consider it
>> > highly
>> > > unlikely that the decision about whether a trend is in place is a
>> > > discretionary decision - I can define a trend in several
> different
>> > > ways - all of them can readily be written as a rule (in words or
>> > with
>> > > code) - I don't care if the definitions are 'correct' or not as
>> > long as
>> > > the system that they are part of works i.e. my rules for a trend
>> > depend
>> > > on the context.
>> > >
>> > > As Dennis said, our rules might be difficult to program,
> causing us
>> > not
>> > > to automate the trade, but mentally we are still running the
> rules
>> > and
>> > > if we are honest with ourselves we do know what the rules are.
>> > >
>> > >
>> > > >For a novice traders to try and mimic the techniques (of
>> > Discretionary
>> > > >Traders) without
>> > > >having similar backgrounds merits caution.
>> > >
>> > > What I am suggesting is that, over time, the sub-conscious mind
>> > will
>> > > automate what was intially habitual conscious behaviour, and even
>> > make
>> > > some improvements on it, so that 'we' can skip the conscious part
>> > for
>> > > some 'tasks' e.g. driving the car becomes second nature.
>> > >
>> > > That won't happen for new traders, in a short time, so they do
> need
>> > to
>> > > persevere, be patient and not try to mimic people who have been
>> > around
>> > > for years.
>> > >
>> > > IMO formal (written) rules based trading/backtesting/optimization
>> > is
>> > > the best place to start - it grinds the basic lessons in very
> well.
>> > >
>> > > If anyone can look at a chart, and without recourse to any rules,
>> > know
>> > > which way the price is going to move and trade successfully
> (long
>> > > term) on that basis then that is something else altogether.
>> > >
>> > > If it is at all possible to do that then it definitely can't be
>> > taught.
>> > >
>> > > That is why I asked, "Anyone doing it?".
>> > >
>> > > It is just like >100%PA returns - anything is possible but once
>> > someone
>> > > confirms that they have done it then it moves from the realm of
>> > > possibility into reality.
>> > >
>> > > In the meantime I will stick to my guns by saying that "except
> for
>> > > people who KNOW what the price is going to do everyone else is a
>> > rule
>> > > based trader and categorizing traders, as DT's or MT's, is
>> > arbitrary".
>> > >
>> > > brian_z
>> > >
>> >
>> ------------------------------------
>> Please note that this group is for discussion between users only.
>> To get support from AmiBroker please send an e-mail directly to
>> SUPPORT {at} amibroker.com
>> For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
>> http://www.amibroker.com/devlog/
>> For other support material please check also:
>> http://www.amibroker.com/support.html
>> Yahoo! Groups Links
>> No virus found in this incoming message.
>> Checked by AVG - http://www.avg.com
>> Version: 8.0.138 / Virus Database: 270.6.6/1623 - Release Date:
> 8/20/2008 8:12 AM
> ------------------------------------
> Please note that this group is for discussion between users only.
> To get support from AmiBroker please send an e-mail directly to
> SUPPORT {at} amibroker.com
> For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
> http://www.amibroker.com/devlog/
> For other support material please check also:
> http://www.amibroker.com/support.html
> Yahoo! Groups Links
> <*> To visit your group on the web, go to:
> http://groups.yahoo.com/group/amibroker/
> <*> Your email settings:
> Individual Email | Traditional
> <*> To change settings online go to:
> http://groups.yahoo.com/group/amibroker/join
> (Yahoo! ID required)
> <*> To change settings via email:
> mailto:amibroker-digest@xxxxxxxxxxxxxxx
> mailto:amibroker-fullfeatured@xxxxxxxxxxxxxxx
> <*> To unsubscribe from this group, send an email to:
> amibroker-unsubscribe@xxxxxxxxxxxxxxx
> <*> Your use of Yahoo! Groups is subject to:
> http://docs.yahoo.com/info/terms/
__._,_.___
Please note that this group is for discussion between users only.
To get support from AmiBroker please send an e-mail directly to
SUPPORT {at} amibroker.com
For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
http://www.amibroker.com/devlog/
For other support material please check also:
http://www.amibroker.com/support.html
![]()
__,_._,___
|