« Trading Dilemma: “Creating and Sticking to your Fundamental Trading Rules” | Main | Fibonacci Trading »
How to Design a Simple Trading System
by John Piper
I have spoken a lot about system design and the next logical step is to write a more detailed piece on this subject. So how do you design a system. Well, the first step is to put the structure in place. By this I mean the aim of your system. What do you want it to do. Do you want it to catch trends? Do you want to trade ranges? How much risk do you want to incur? What success ratio are you looking for?Some of these parameters are going to affect each other. For example, if your stops are very close then your success ratio is going to be reduced. But as long as you have realistic expectations there is no reason why they cannot be met. Having decided what you want the next step is observation. Examine how markets behave and how you think you can take advantage of their actions. That is the key element to any system. As long as you can divorce yourself from the psychological effects that markets have on your psyche. You have to double guess the market, because otherwise you will be losing all the time. There are two ways of doing this. Either you look to trade ranges or you look to trade trends. The first means looking for extremes and entering when such extremes are reached. The second means looking to catch trends and entering when your system indicates that a trend is in place. Albeit you can also combine these two approaches.
In both cases you need to define your trading conditions. You need to define a range or a trend. Once you define what you are looking for you, simultaneously, you define how to catch it. You can define trends in many different ways. First you have to decide over what time frame you wish to define the trend. You must then use that time frame to give your trending signal, for example if you feel that you want to trade weekly trends then you must in some way define the trend using weekly charts. Once you have defined the trend you will have your trend indicator. So if you decide that a higher high on a weekly bar chart means that you have an uptrend then that is your indicator. Alternatively you may decide that acceptance above the previous day’s value area signifies an uptrend – again that is your trend indicator. Incidentally I am going to concentrate on trending systems for the purposes of this article, but the same basic approach will apply to whatever you decide to trade, be they trends, ranges, spreads, or whatever takes your fancy.
At this point let us list the requirements that you need to make up the entire system:
1. You need to define your objectives.
2. Therefrom you get your trending signal.
3. You must then decide on your money management system – this is critical.
4. That tells you how much you can risk on each trade and from this you can decide on your stop policy and position size.
5. You must then decide on your entry strategy.
6. The next step is how to move your stop as the trade progresses.
7. Finally you will want an exit strategy, albeit this may simply be to wait for the stop to be hit
So far we have covered items nos. 1 and 2. Money Management has been the subject of many articles in these pages and I don’t want to repeat this here. Suffice to say that ideally you should not risk more than 1% or 2% per trade with 5% as the maximum. Otherwise you will run the risk of being wiped out.
The stop policy follows on from this. Once you know what you can risk you can work out the stop policy and the position size. However your stop policy goes to the core of your system. it is the relationship between your entry point and the stop which defines the system. To an extent this is where the magic comes in. This is the essential element which we cannot reach by logic. This is an act of creation. You must decide upon your methodology which is the heart of the system.
This particular aspect requires observation, thought, creation, and then testing. Although the actual creation must be yours there might be a few guidelines which I can usefully give. We will start with Time Price Opportunities (TPOs). All charts are made up from these building blocks. The building blocks themselves merely refer to the fact that as time goes by, illusory thought it may be, different prices are available and traded. Hence TPOs, a term from Peter Steidlmayer’s Market Profile (MP). From TPOs you can build whatever form of chart you want. Don’t be constrained by the methods you know, bar charts, point and figure, candlesticks, MP itself. Be adventurous, devise new ways of displaying the raw data (TPOs) – you do not need to stick to the ways you know and you may find a method which suits the purposes of your system much better (maybe you will be the next Steidlmayer, discovering a better way of looking at markets!). Experiment with using time in a different way, and then price. As you do so you may find that different ideas present themselves to you.
Looking at price action rather than TPOs themselves we find different aspects to view. Is the action determined in one direction or the other? Has it exceeded previous highs or lows? Has it moved sharply away from previous highs or lows? Has a particular price level been sharply rejected? Has a particular price level been accepted? All these questions, and their answers, can give you a clue as to how your system might work. Going on from price action we have all manner of indicators which present us with different forms of squiggles and dashes with which to conjure.
Out of all this various information the system designer needs to choose what to use to achieve his own objectives. My own preferred “indicator” of choice is Minus Development (MD) which is generally, but not always, a spike. I like the spike because it is an extreme and the stop placed beyond the spike is relatively secure. I think systems built around spikes have a better chance of success than other systems. The placement of the stop is vitally important because if the stop is vulnerable then so is the entire system and this brings us onto the concept of data stability. Longer term data is more stable and will give you better signals, on the other side of the coin you will be forced to use much larger stops, thus either risking more or trading less contracts.
So here is the big dilemma. If you want the biggest bang for your buck you are going to be looking at short term systems, thus using less stable data and hitting a greater number of losers. But the better systems are going to be longer term. But many traders I know are already in catch 22. They trade precisely because they haven’t got much money and they see trading as a means of getting rich. Thus they are forced to look at shorter term systems, because they cannot afford to take the longer term signals, and they are also in too much of a hurry to wait for such signals in the first place. Those who are rich have great respect for their money, which is why they are rich in the first place, and will only trade in a very cautious way, meaning longer term systems, and they will have the patience to wait for the signals. Many traders have little respect for their money and until they learn that respect are probably doomed to fail. However the actual design of a system need not be different for longer term or shorter term purposes. Except that longer term systems tend to be likely to be trend orientated as markets tend to trend over the longer term. This might lead the system designer to thinking along two distinct lines. The first might be “to get from point “A”, where the trend begins, to point “B” where it ends, the price must move through many points in between”. Thus he will set about defining such a point at which the system will enter the trend, preferably with stops beyond an extreme of sorts. The other thought might be “when the old trend ends and a new trend begins there are often similar types of price action”. Thus he will set about defining the type of price action he will look for to define such a turning point. By price action, I include all the various indicators, etc, and how they might behave at turning points or as trends continue. It would be useful if readers wanted to write follow up pieces to this. That way we might get some very interesting insights to how systems may be designed.
Finally let’s design a system! Our objective is to catch and stay with the big moves. We are going to use the weekly chart to determine trend. The rule will be that we must see 2 weekly highs or lows taken out. So an uptrend means that 2 weekly highs must be penetrated, we must also see acceptance above that second high. The stop is the last major high or low. We will also use major failed re-tests as a trading signal, as seen on 16th and 24th July. Entry can either be simply when the signal comes in or you might use shorter term techniques to try and catch the move. The stop will be moved to breakeven once we are 50 points in profit and will stay there. Otherwise we will only exit when we see a signal the other way. This simple system will catch virtually all good moves. The stop is secure, and, relatively, close. It allows profits to run. It would probably only trade a few times each year, maybe just once. In fact once would be ideal, because the best trades never end, they just keep on making money.
If you like this concept then check it out on some back data over a number of markets. The last trade on FTSE would have been triggered around 3775 (on the cash) on 5th August when the high of the week ending 2nd August was taken out. The stop would originally have been at 3610, but would now be at breakeven. We now await either a reversal signal (either by the weekly chart or by seeing a major reversal signal) or getting stopped out. Let’s call this the TTT Weekly system and I will give any new signals over the hotline, so we can monitor how it does. Some of the rules above would need to be more closely defined but my objective is to give an example of what might work in the markets. Anything that allows big profits and takes small losses fits the bill. In this case the stop is rather wide but on the other hand the market never looked back after the signal on 5th August. But you cannot judge a system on one isolated incidence.
This is an article submitted by John Piper of The Way to Trade.
Topics: Forex, Futures, Minis, Options, Stocks | No Comments »
Comments
You must be logged in to post a comment.