My entry into forecasting was in 1995 when I read an article in AI in Finance Magazine called "Tahiti or Bust." It outlined the experiences of a neural net expert who worked with a trader to develop a system for trading so they could both retire to Tahiti.
One of the things that they did was according to the article was use neural networks to forecast the markets, and then they combined this with the trader's expert knowledge of trading the SPX Index to develop a trading system.
The article ended, "And then we retired to Tahiti."
Since then, I've been researching different methods of market forecasting, including indicators, neural networks, genetic algorithms, and other methods. I still haven't found the silver bullet yet that will let me retire to Tahiti.
But, I have put together enough information that I think that is IS possible to forecast the markets, and trade successfully. However, I have yet to find all the methods that will allow me to do this.
This blog is to post some of what I've found, and hopefully some of you will find it interesting and share what you've found out, so I can finally get to Tahiti.
Thanks for reading!
Saturday, August 22, 2009
Market Physics - Looking at Speed
One of the things that I ran into when I first started trading was that there were many methods to use to trade. I decided not to wait because of the article that I read, and because it was about a SPX trader, and decided to jump right in and start trading OEX options because they were cheaper than SPX options.
Right away, I started using charts, and ran into a problem which was -- which indicators to use? One of the simplest things that I looked at was the change in day to day value, which meant that my option values went up or down depending upon the market. Then, I started looking at the speed that the market moved at from day to day to try to determine the best times to buy and sell. I did this just the same way that I'd drive a car -- I paid attention to the charts, how the market was moving, and then bought and sold according to how the market moved from day to day.
When it started going up, I'd jump in and buy, and then try to sell to when it slowed down before it changed direction too much. This strategy worked great, but the stress was immense. At times my option positions would change thousands of dollars in just a few hours. The stress was so great that I had to quit trading so I wouldn't lose my day job (which at the time was computer programming).
Here is a simple chart that illustrates this.
(Data source: Yahoo Finance).
What I've done is plot the daily change, and then plot the change versus time, known in physics as speed. One of the things that this plot can be used for is to time entries and exits.
I will never forget when I was working at Fuji Capital Markets one day, they were losing money trading Eurodollar futures, and the head trader said, "The market is moving against us." The way he said it, it sounded like they made the market. I never in my trading tried to be against the market. I always tried to be with the market.
If you look at the graph below, if you look at the daily change, and the speed, what I tried to do was never trade against the market. I always waited until the change was a positive one, and the speed went to positive in that direction, before I entered a trade. I always waited until the change was a negative one, and the speed was positive in that direction, before entering the market. This was my first entry into indicators, which are calculated values based upon prices that can help you with timing trades and analyzing the markets. I started looking indicators while I was starting to do my forecasting work with neural nets. Indicators can give you an idea of how the market is moving in the short term, in a way that is easier on the stomach and less stress than trying to track it in realtime the way you would drive a car.
Right away, I started using charts, and ran into a problem which was -- which indicators to use? One of the simplest things that I looked at was the change in day to day value, which meant that my option values went up or down depending upon the market. Then, I started looking at the speed that the market moved at from day to day to try to determine the best times to buy and sell. I did this just the same way that I'd drive a car -- I paid attention to the charts, how the market was moving, and then bought and sold according to how the market moved from day to day.
When it started going up, I'd jump in and buy, and then try to sell to when it slowed down before it changed direction too much. This strategy worked great, but the stress was immense. At times my option positions would change thousands of dollars in just a few hours. The stress was so great that I had to quit trading so I wouldn't lose my day job (which at the time was computer programming).
Here is a simple chart that illustrates this.
(Data source: Yahoo Finance).What I've done is plot the daily change, and then plot the change versus time, known in physics as speed. One of the things that this plot can be used for is to time entries and exits.
I will never forget when I was working at Fuji Capital Markets one day, they were losing money trading Eurodollar futures, and the head trader said, "The market is moving against us." The way he said it, it sounded like they made the market. I never in my trading tried to be against the market. I always tried to be with the market.
If you look at the graph below, if you look at the daily change, and the speed, what I tried to do was never trade against the market. I always waited until the change was a positive one, and the speed went to positive in that direction, before I entered a trade. I always waited until the change was a negative one, and the speed was positive in that direction, before entering the market. This was my first entry into indicators, which are calculated values based upon prices that can help you with timing trades and analyzing the markets. I started looking indicators while I was starting to do my forecasting work with neural nets. Indicators can give you an idea of how the market is moving in the short term, in a way that is easier on the stomach and less stress than trying to track it in realtime the way you would drive a car.
Subscribe to:
Comments (Atom)