splat, it is certainly possible to trade the markets with mechanical strategies. I do every day. The trick, however, is that you really cannot code anything remotely close to the minute by minute skills you learn over time. So I suspect that I will gravitate more and more towards tape-based trading over the systems over time. The game you play with mechanical systems is one of statistics: you have to take all trades as indicates or you mess with your outcome. This is actually more difficult than it sounds
it is possible to do pure mechanical, but it's very hard... I have been working on some intraday systems and the results are not consistent to say the least... seems the market has a natural way to figure out any edge, and eventually your system stops working... so I guess the trick is to have multiple systems to trade multiple markets on multiple time frames... you will find out it's a LOT of work, a continuous research process.. the Holy Grail does exist, and that is relentless research in finding an edge under any market conditions. some while ago I was naive to think system trading = figure out the magical system and set forever, boy was I wrong or what !!
I wrote a previous post about a fund I was selling that had over a 1000% return over there last five years. That fund was a system trading fund that traded futures and commodities every market, every time frame the system was being worked on from 1978 to 2001 and was perfect until 2002 then it blew up. The people that were running the fund weren't armatures these people knew what they were doing. Some of the people involved in the fund were ex floor traders they spent over 20 year for a holly grail and in the end they got screwed. The other funds I was selling were discretionary and have been around a long time there is no substitute for learning how the markets work. Look at schonfeld they have black box trading but the traders that make the most are the discretionary traders, then look at George Soros he is also a discretionary trader. If there was a perfect system every marketmaker and specialist would have it along with every big firm.
I think where a lot of people go wrong is trying to come up with a system that allows them to daytrade 4 or 5 times a day, every day. It is possible I believe to have a profitable daytrade system for the stock index futures but it will not trade everyday, in fact it may not trade every week.
SPLAT "I've read so many negative things about commercial Trading systems. Personally I wouldnât use one. But I am thinking of maybe looking at developing my own." Good luck mate, I bet you will be optimising and calculating for years. I spent years tying to build automated end of day systems a until I just started spotting good chart opportunities using trend analysis and visual inspection. The only indicator that I do use is 200-day relative strength as I like to look for long term divergences in the trend. I like to go through 100 or so end of day charts and narrow down a short list or interesting looking patterns (or forming patterns). How do you all feel about THAT then!! PS i hope you do really well at trading splat
the RSI divergence stuff looks good on books (so do most other indicators), but I back tested an end-of-day S&P system purely based on RSI divergence.... guess what, there is NO edge what so ever. and yes... I have the C++ code and data file and will be glad to send to anybody interested. I understand some people may be successful trading the RSI divergence... but I suspect the real edge is just some sort of positive expectancy, but the trader is somehow made to believe the divergence works. One thing I can tell you.... forget about using ANY canned indicators... the only thing that might have a slight chance of working is based on signals purely derived from price actions itself. Man look at the bookstore shelf with all the TA publications, there is so much bullshit in this industry !! :eek:
I do not just see a divergence and say "oh look a divergence lets trade it" thatâs like saying its cloudy outside it is defiantly going to rain. It is only an indication of what may lie ahead the, it is up to the individual to use this indicator in conjunction with other charting methods before forming a solid opinion on the stocks technical condition. I use the RSI merely as a tool to help me find potential weekness and strength but I like to see a nice conformation E.G double tops bottoms, head and shoulders, trend breaks etc.
I developed a mechanical trading system for trading the e-minis.Since 1997 I've kept very detailed records of it's performance.UP until 2001 I traded it off and on.Not because it wasn't performing well but because I wasn't making the effort to trade every signal.It typically generates 10-14 trades a month and is a market on close (entry)to market on close(exit next day)system.I finally woke up and made the committment to trade every signal for 2001 after 2000's trading signals would have generated 353 points in profits.I guess thats something like a 300% plus return on capital.For 2001 I still kept detailed records based on a mechanical trading basis but in actual trading I used my discretion in terms of exiting.When I did my final recap of 2001 performance based on a total mechanical approach I was amazed the Total Net gain point wise was Zero.Figures after 2000's 353 point gain.But my real-life actual results with my discretionary exiting was over 260 points.This re-inforced what I always felt about trading a mechnaical system....using entry points and stop points based on a mechanical system is definitely worthwhile but when it comes to exiting and taking profits,I think trader judgement can enhance the results
if you read my post carefully, what I am trying to say is the divergence itself does not have any EDGE... your profit comes some other posive expectancy (DT,H&S etc), and you can easily get confused thinking the profit comes from the divergence (or at least partially), because without back testing, how would you know??
I don't use a system to actually place my trades, however I do use OMNITRADER to automate the process of collecting buy/sell signals that I set. I feel it is a waste of my time to stare at the charts all day when I can glance over, only when something is "happening." An no, the signals don't always work. I have found that the signals work best only when they are "confirmed" by my brain. For example: TA doesn't take into account FED MEETINGS, NEWS and the like. I only place trades: 1.) Based on simultaneous TA signals 2.) With the trend 3.) Absent of major news or economic events Here is an overview of my strategy: I scalp the mini-ES 4-8 time a day, 2-4 contracts each. My goal is 2 ticks. $200-800 per day gross I use the following charts: 5-Minute candle chart for TREND (up or down) 1-Minute candles for buy/sell signal (with the trend) Tick chart for the trigger. These are the basic 4 indicators I use: RSI peaks DMI Crossover Commodity channel Kirshenbaums I chart with QCHARTS and use IB as a broker. Hope this helps.