TOOLS FOR TRADING
I am constantly displaying different things that attempt to measure and quantify market action. We calls these "tools of the trade." Regardless of your occupation, there are certain techniques you use to determine how you will go about your business, trading is no different. One of the newer ideas that is out there is projecting price movements by computer matching of bar patterns. The techniques use algorithms that go back and measure or match the recent activity to other occurences that have high correlations to that action. They then project what will happen going forward, based on what has happened during the prior occurences that have a high probability match.
Above you can see one I have run for the SP500. What this match is telling us is that based on an 80 percent correlation, there are 13 prior matches in the history of this contract to recent action. On average what you see is what has transpired after those prior occurences. This shows a small rally, then a small dip, then a steady rally for a couple of weeks beyond that. I have played around a great deal with this and have found it to be useful, but more as a tie breaker if I am undecided on something. When this tool was first released by Genesis last year, I thought it was just great. I do think it is accurate more than 60% of the time. As a result I used it on a weekly basis to confirm every trade. Alas, I missed a few very big ones where this was wrong and everything else I use said to do them, so I put it in storage.
Recently though, I have taken it out of the moth balls based on a comment friend a fellow trader about it, just to use as a general projector of what might happen. If it is against what I am considering doing I do not care if everything else says to go. However, if I am really torn on something, I am using this to push me into a trade. There are now versions coming out soon that will do this same thing not just on price, but on the indicators themselves. For example if you use stochastics you will be able to have your computer tell you what happened with price the last time the stochastics "looked" like they do right now. I have mixed emotions about this for a few reasons.
First, the more mechanical you get the more subject your approach is to being over optimized. The biggest nemesis to developing trading systems is "data mining." This is basically where you tailor your rules so precisely to prior outcomes that it is virtually impossible for them to work in the future. However, your test results look gangbusters. Second, for the most part indicators tell us what has already happened. They are generally not inherently predictive even though we try to make them so. Third, we are getting into a situation where we are trying to compete with the brain trusts of wall street. Large funds can hire alot of brainpower to develop and use tools like this in way most of us neither have the time or the skill to do. If we get too reliant on this we are at best using something in an inferior way to someone else.
It certainly is an interesting prospect to think about the computer being able to tell us when our favorite indicator is telling us something and to know exactly what that something is. Are the patterns we use as good as we think they are? This will tell us that. However, there is one thing that can get lost in all of this technology, the human element. As full of flaws as we all are, we do have moments of brilliance that can lead to very good successes. Making judgements at critical times is part of life. Trying to get around making those decisions so we can blame the bad outcomes on a "system" or tool is just a cop out. I am willing to live with the ramifications of my decisons, knowing at times the decisions will be lousy. I also know at times the decisions will be great and better than any machine can make.
When I made my transition away from systematic trading to discretionary trading it was painful. The first few months were very up and down, not at all what I was used to in my old mechanical days. However at this point I now make more money that I ever have trading with discretion, and I would not have it any other way. Tools of the trade are mandatory, just don't become slaves of them. Use them in a way that best suits your personality. I can assure you that if you have a conflict between the two, eventually it will really bite you. The worst part is that it will most likley bite you at a time you can least afford it.
Is that above forecast likely to be what happens? I have no idea. It is in conflict with some of the cycles I displayed over the weekend. It does represent what would be the bullish case here that we will find support at these levels and move back up again. Some of my cyclical work indicates a bounce here, then a larger dip than this shows, so we will have to see what transpires. That is what keeps this fun, not knowing.
No comments:
Post a Comment