I'm sorry, I honestly don't know the answer to that question. I wish I did it could make life a lot easier - it seems to vary so much. Some systems take not much to prove them or disprove, other take a lot more... Kind regards Natalie
For equities a 30 min chart is the basic defining chart. I use a quote sheet as well which is real time. Over the years many tools have begun to appear. The last three that are significant are: MA of volume of indexes in daily newspapars, a thing called unusual volume on the quote sheet of qcharts. A/D as an indicator available on Q.charts I regard all these appearances as a status report on the state of the art. LOL. It would be nice is alot of stuff were spread around to hepl people out. For equities, current beginners appear to be running at a bout 7% a week as they start out. The futures indexes work most easily on intraday charts (I describe the ES03Z here). the 5 minute chart is the easiest to trade and the 1 minute chart backs it up. An example of three levels of the tree working works out this way for the 18SEP03. Beginners would do three trades of 8,1,and 0 points; Intermediates had 5 trades of 8,2,0,0 and 1. Experts did many trades. Two of the three conditions that apply to a gross market assessment occurred. On the next level where the 5 min bars were examined, twenty trades were initiated and they ran their courses in all cases except 8. Those 8 required an intervention to be able to take proifts. Interventions are on the third level of the tree. In the 12 where no interventions were required two losses occurred. One was a 1 tick loss and the other was a 2 tick loss. The remaining 18 trades yielded profits in order of importance such as 7.2, 3.1, 2.1, 2.0, and all others under 2.0; one trade was a wash. The net was 17.0 points and the H/L for the day was 15.0 points for comparison. As posted byother traders elsewhere many took one trade during the day and a lot of people saw the day as limiting because the day's operation appeared to be mostly a slow long trend. Profits are more or less limited on such days. I run concurrent logs for beginners, intermediates, experts using everything (three tree levels) and experts not using interventions. It is a simple process. You can see that watching 81 bars a day has an assortment of rewards. The marginal improvement from beginner to intermediate to expert with or without intervention is a questionable consideration from the view point of effort in this example. the differences in 9, 11, or 17 points is not much variation. The slow long trending day is the reason, of course. An easy day to take money out of the mrket. The more frequent occurance is that the spread is vastly greater. This is because the day usually breaks up into more traverses (partial usually) of the daily H/L. You can see what just one 5 point partial traverse to and fro does. Two of such is fairly common where one is in the am and the other is in the pm. The primal concept of "staying on the right side of the market" is very persuasive. If a person makes an effort that focuses on three levels of consideration, it is an almost exhaustive effort to insure success continually. The gross level is what beginners and intermediates should adhere to. Experts, first, just add another level, and do not really progress much because they have losing trades still. The third, most refined level, provides one standard intervention which disposes of the probability of losses. The narrow battleground where losses occur is located at the place where intervention is not signalled and yet the market does not offer change over time. Limbo is a good description of this stand off. you do not make money and what you lose is the smallest increment available as one would expect. I hope this fills in some blanks for you.
...thanks for that detailed explanation. Perhaps I am missing something from your history of posts, but I don't understand what you mean by beginners, intermediates, and experts. Are those progressive levels in your decision tree? Best regards. - Mike
Does anyone else on this thread have a view on backtesting generally or on backtesting system trading .. what is a sufficient period? Currently I use the last 7 and half months. I have concentrated on trading of the Dow. Is there a best sampling period do you think? Can it be too short or too long? So what would you suggest .. 5 yr, 2 yr, 1yr, 6 months, 3 months? Any info on your practices or views as to backtesting periods would be very helpful.
...IMO that depends on what you are backtesting for. If you want to determine the robustness of a pattern- or time-based entry rule, I would say the longer the better, to verify that that rule represents some enduring structure in the market. If you want to optimize price-based entries, stops, and profit targets which will vary as market volatility changes, I think shorter is better. Many sources say less than 20 trades is too short. In intraday trading I think 30 days catches the volatility about right, but I test to the maximum of 60 days that my service provides because that typically is enough to show a couple of drawdowns. How and what are you trading?
For learning to trade, all persons have to go through a progression. It is mandatory to avoid and eliminate false starts in making money. Every person starts out at a time when they have incomplete knowledge, skills, understandings, and beliefs about how a market works. Therefore, they may not participate in money making unless the conditions are compatible with their knowledge, skills, understandings and beliefs. The beginner constraints are to only trade fast paced strong trends. This is like an edge sort of constraint. The edge is limited trading, minimum capital in the market (a one contract limit), and always exiting at the end of the fast paced strong trend. You know when you have beginner skills because you have accumulated several times your starting capital. Intermediates trade beginner trades and they also trade all slow trends. Slow trades are simple channel trades of any slope and the channel is a standard graphically defined channel. channels form and channels end. This is within the grasp of anyone who has, as a beginner, profited by several times their initial capital. Experts, are in the intraday market continually from the time the market settles down shortly after (15minutes usually) the market opens. hey normally use the tree on it's various levels and periodically during each bar go through: data gathering; data analysis; decision making and taking action. The usual action when an established money velocity is present is to "continue" to make money. Periodically, these people collect profits and continue along with the next opportunity that has arrived. Most of the time for any level of trading, the market tells you what to do. That is, you know the market to an extent based upon experience and you have a set of beliefs about it. gathering data and analyzing it provides information that is compared to beliefs you have gained. You make decisions that fit the situation. A decision based beliefs, allow you to behave accordingly. All of this in combination is "executing". You will observe that in ET a lot of people get "signals" and do something. They enter, hold , and exit because signals tell them to. They do this stuff because they made up a sweat of rules that they follow continually in a totally disciplined manner. And they probably probably backtested what they do. you will get the idea that what is going on is a mechanical type thing but usually not in a software state going along automatically. The alternative to this approach is actually monitor the market in a reasonable manner related to your knowledge, skills, understandings and beliefs. As you "grow" you have success. Success enables you to build more abilities. It is a process. I but it on one and 1/4 pages of text and Illustrated it with a guide illustration that has 9 pages of detailed word illustrations in columns and rows. This "document" illustrates how a person connects to a market at all levels and under all conditions. Arrows connect everything. You may if you wish, keep track of where you are at any moment. If you do you know where you are going to be next and if you keep track of time, you can ascertain when the next events are. I put up two stock examples recently. I marked the first illustration of each chart with X's to point out what was coming up soon. As the events occurred, I posted another chart showing how the price, volume, and all indicators filled in as I said they would. The charts now give a before and after kind of story. When you are investing this way, it helpful to make up the charts in advance so you can coordinate your accounts and trades. The two stocks I did recently are CHRS (I posted the daily stops to use there as well) and BBI (here I posted the amount of profit coming up as well.) In ET people are not used to having complete and comprehensive approaches on the table. That makes what I post seem uncommon, not the usual, and often not understandable. you night soya that it could be expected that a person would not accept what I suggest. I posted two charts on two different equities where people were "doing their thing" with them. In one case the initiator of the thread (on CHRS) was actually interested in gathering the thoughts of others on aspect of making money regarding CHRS. I posted what was going to happen, put in the trend lines and the stops and the point where the next trade would begin in advance. Then as time passed it happened. no one in ET would mark up their charts personally to see this unfold because it would be too much work or the copying is time consuming and such. So I posted the chart again after the time passed to do the trade. In this way people could see how investing works and they could see how I make money using what I do. For BBI the person wanted to trade it from where it was to a low value of 15. he thought up 15 and decided to do what he thought up. I simply posted a chart that had the information on it to make 6 to 8 % a week ad nauseam. the person thanked me and said he was not interested in trading it to make money. he was interested in doing a long term trade to short it to make about 6 points on 21 over an unknown as yet period of time. I do otherwise. I schedule stocks day after day so I can see them go along and Break out (buy time); I put in the slopes of their trends a head of time and I put in the stops as well; I mark when to sell them and at that time I have other charts ready to put money into. All of this is basically beginner stuff for equities. Right now beginners are knocking down about bout 7 to 9 % a week. they can go through "growth" and then be in a place where they make much more money. I compare the range of monthly income in the four streams of capital that is being used. the monthly ranges included 12, 22, 34, and a large one. Naturally, when a person is doing thing appropriately, then all streams run near "a large one". There is another thread here showing that no one in the world has done this consistently ever any where. No one can argue the point being made in the other thread. However it does appear to some extent that if one chooses stocks well and does the work,things turn out quite well. At ET people chose poorly (CHRS and BBI) occasionally and then they do not trade effectively. This is a tough thing for people to continually endure. I certainly would not do what they do.
Thank you Girlpower for your comments earlier. Answering you query Hypostomus .. trading the Dow based on direction and range. Thank you for your 'in-a-nutshell' advice on backtesting .. any additional comments on backtesting?
...that was very generous of you to compose such a thorough answer. To read it was a humbling experience. It is very much like reading about the standard progression of meditation practice when you are still struggling to attain the "access" level. Many thanks. - Mike
My testing to date indicates that profit targets generally reduce system performance (unless the system is extremely bad to start with). I prefer adaptive trailing stops or reverse signals.