first its there and then it ain't

H

hoodooman

I use technical indicators to trade very poorly, by the way. I use the Medved Quote tracker and right now my favorite setup is using 2 minute charts, candle sticks, trix and the parabolic sar. I also use the 50 ema and two 2ema's that track the high and low average of each candle. At the end of the day, my charts look absolute great. But in reality this is just a fantasy. In real time the trix crosses and then uncrosses before it decides what its going to do and the sar will give a buy signal and then change its mind and print out a sell on the same candle.

I would like to know if other people are having the same problem. Do their systems change signals or is this just a problem with QT. I would also like to know if there is a solution to the problem.

regards
peter daniels:confused:
 
This is why looking at old charts it seems like indicators are the holy grail as they mark every turning point with amazing accuracy but when you try to trade off them , as you sit there watching the tail on that fast line flipping up and down, you come to the realization that there aint no easy answer.
 
I use TradeStation, and it has an option to keep the indicators from updating on each tick. Any system that I test that uses an indicator always enters on the close of the bar so that the indicator reading it is based on is "final" for that bar. It is the only way to get a realistic idea of how a method will perform when I am backtesting something. This usually results in most indicator systems being worthless (which is how they would have performed in reality too, so better to find out now).
 
It seems to me that if the indicator was calculated discretely with only the closing price of the candle this phenomena might be eliminated. Of course the candle time interval would have to be small but the results might improve considerably. At least it may be a much better way to analyze the indicator's effectiveness. Right now, I wait for the price to move to the high or low of the moving average after the initial signal. This helps some with the signal descreptancy. I've emailed Mr. Medved about it twice. I can't recall what the reply was but I think that I will try again.
 
...a suggestion. Subscribe to a service which allows you to code strategies and backtest them. Pick ONE rule, parabolic SAR, enema crossover, momo, anything. Code it, and do parametric variations (ATR bars, enema period, momo period) and test it. Do this until you convince yourself that TA doesn't work. Then start randomly trying some pure price or time rules. What happens if I buy the Tuesday open and sell the Tuesday close? What happens if I sell a gap down? What happens if I short every morning at the open and get out an hour later? (None of these work, of course, or I wouldn't suggest them.) Do this enough times and like a blind hog who occasionally finds an acorn, you will find a profitable pattern. Strange things happen when you try strange patterns. The stranger the better. The more counterintuitive the better. The less logical, the better. This activity will also keep you out of the market and save you a lot of money.
 
Technical analysis, in the long run hasn't worked for me, intra day and you folks are probably right but I was hoping I might get a positive response from a few people. At least get some attention to the problem of these "now you see it and now you don't " technical indicators. Thanks for the replies

peter daniels
 
Quote from hoodooman:

I use technical indicators to trade very poorly, by the way. I use the Medved Quote tracker and right now my favorite setup is using 2 minute charts, candle sticks, trix and the parabolic sar. I also use the 50 ema and two 2ema's that track the high and low average of each candle. At the end of the day, my charts look absolute great. But in reality this is just a fantasy. In real time the trix crosses and then uncrosses before it decides what its going to do and the sar will give a buy signal and then change its mind and print out a sell on the same candle.

I would like to know if other people are having the same problem. Do their systems change signals or is this just a problem with QT. I would also like to know if there is a solution to the problem.

regards
peter daniels:confused:
More...

Try different charting package. Prophet.net has one week free trial on their gold package .
Intraday trading is like boomerang hunting . You see a pattern of a movement so you toss a boomerang ( order ) and wait if bird ( price ) will fly into a path of a boomerang ( your order ). You have to learn to recognize pattern which lead to the price change or a trend change . When you study your charts after the fact, pay more attention to what happened before your signal, not at the signal .
Walter
 
Quote from hypostomus:

...do you know how?
More...

I have slowly come to the belief that the market at any given point is almost random. (I recommend reading Fooled By Randomness). Therefore no matter how I enter a position, its near random as to the result.

As such I use support and resistance levels to make targets and place stops. If this supports a 2:1 risk reward ratio, with a 50% chance of success, over many trades this will be profitable. I use position sizing to make risk constant over all trades (ie 0.5% of account size). Otherwise you'd find yourself risking 2% on the losers and 0.5% on the winners, which would bleed you dry.
 
You must log in or register to reply here.
Back
Top