trading systems - the basics

  1. 1,368 Posts.
    If you’ve been getting hammered lately, and want to take yourself (your ego) out of the trading equation, a system might be for you. Ive investigated just about every one around over the last few years. Heres a quick look at the basics for those who havent got started in this area yet. I must say there are some very good arguments for not using a system at all, but we’ll leave all that for the random walk theorists. Personally, Im a contrarian. Note that all this is for stocks only, and not for warrants or options.
    Whatever you believe about trading, you can always benefit from listening to someone elses research, even if you decide not to use it. These are just the basics of setting up a system.

    1- Know your software. I recommend metastock, but anything with similar capabilities is OK, as long as it can do backtesting. Get to know your manual, and open up all the parts you’ve been putting off, like learning how to use templates and portfolio management. Set up portfolios for blue chips, and stocks under $1, as youll want to test those seperatley. A range of $5-20 stocks is also a good choice.

    2- Entries, wow , there a thousands of combinations. I’ll discuss some choices at a later stage, but lets start with setting up a simple Moving Average system, like buying when a 3day simple MA crosses above a 8d MA. Go to Trading systems, and Edit one of the existing basic system examples. There should be a few in there already. Learn the formula language. Buying when rsi breaks above 30% is also a common one. We can get more sophisticated after weve set up a basic system.

    3- Exit, ok heres where the rubber meets the road. You either exit technically, as the 3d ma crosses below the 8d ma, or you use money management. Good arguments for both choice, but we’ll leave that until weve backtested.

    4- Backtesting. Your software should allow you to test this theory on a stock, say tls for eg, and give you a choice of date range, say 1/1/98 to current, I like 5 or 3 years range, and a choice of dollar amount, say $1,000 per trade, and we’ll ignore commission for this exercise, although your software can show comm slippage. If it doesn’t do all this, then get some better software, cause youll need it to run systems.

    5- Refinement. After we run this and see it is a losing system (no surprise as tls went down, down) we ask our selves some questions. Would this system have worked better on most other stocks? To answer this we run a test over a batch portfolio of stocks instead of just one stock. Would this system have worked better if we short tls as well/ or instead of buying it? Backtesting can answer this. What is the best combination of moving avges to use? We can run a test on all short avge from 3-9 against long avge from 9-21. We may find that a 9day crossing a 16d ma is indeed more profitable. We may find that long avges are better on the blue chips and short avges are better on the penny stocks.

    6- Would an rsi system be better that a MA system. Would MA and the rsi combined be better? Probably would , but lets check the numbers. Save results on each test and use trial and error untul you find some thing that gets it right more than 50% of the time. We can run several layers of tests like a MA first, then an rsi on those handful of stocks that passed the first test. This is called filtering. Most good systems use several filters. Other systems use FA and trade by dividend returns instead of TA, and these are also valid systems.

    OK then youre back to money management. You need a stop loss, not a piece of 4 by 2 to hit yourself over the head with. Most start by setting stop loss at 10%. You can backtest differing amounts until you get something you are happy with. Tighter stops (lower %) will stop you out more often, but save you bank balance. Larger stops like 20% may actually win more, but is riskier.
    Then you need to consider when to take profit. Will we use a preset amount, like selling everytime we make 15%? That’s a good starting point. We can do better with a trailing stop loss. Maybe we’ll let profit run, and only sell if profit falls back by 5 or 8%. We can backtest these combinations and find out what would have worked best. After youve refined that part, most good traders progress to exiting technically, as they have refined and can trust their own indicators.

    Running a basic system will save you a lot of money by stopping you trading blind and getting faked out by every move. Those without a system are doomed to die a slow death (unless they are already pros). Its true that 90% of traders don’t survive, and lack of discipline, study and knowledge is one of the main reasons. Whether you like this appraoch or not, you should at least study and replicate it before you discard it. You must understand what the rest of the traders are using against you. (Reminds me of the old poker saying "if you look around the table and you cant work out who the sucker is, it’s probably you!")

    Those who are interested in this thread can discuss more complicated systems, over the next few days, but lets get everybody up to speed first.

    Cheers.
 
arrow-down-2 Created with Sketch. arrow-down-2 Created with Sketch.