Search results

Thanks for the quick answer :) If you don't mind.. do you use daily or intraday data to develop/test range bound model? I can hardly imagine testing range bound 1-day holding period model with daily data spanning over many years.

Acrary, I have but one last question. In your system development thread you used three models where you explained importance of correlation. You said the first model had an edge and the other two had no edge. Nevertheless, how do you come up with three models that trade the same market with...

If you want objective analysis, learn to test your ideas with statistical tests. Test every idea to see if it's better than random, i.e. significant.

There is nothing special about 2%, it may be overfit to historical data. You need to adjust variables according to current volatility. Fixed values don't work in the long run (cost me to learn this fact two years ago). One day 2% may sound like a small value, a few months later it may look like...

You may try to look at volatility normalized MAE and MFE. If you exit at EOD, you're trying to catch fat-tail moves. Look to place initial stop narrow enough to cut off negative outliers, but wide enough to leave positive outliers intact. Again, your stops should be expressed in volatility...

In that case, if I had to do this stuff, I would use Hull moving average to approximate price action. Then convert time series of moving average into input data as described in the youtube example: The rest is training the multilayer perceptron to recognize the patterns you want. P.S...

Are you looking for something like this? http://www.youtube.com/watch?v=2XOMkkGafGc

This video is useful for insomnia treatment :D And the last words were: "so, it's quite basic".

Impossible to tell without knowing the details. But certainly, looks like a bug somewhere.

This is a simple problem that can be addressed objectively. What you're looking at is 3 risk weights. You need to find an optimal combination. The simplest approach would be to maximize Sharpe ratio. This is how you can do it: 1) take monthly results of the systems 2) run through all...

A similar story I heard from my uncle :) Still don't believe in dowsing. Regarding TA I think you have to define what it is, and then decide if it has any validity. I think there are two categories of TA. One is "magic" BS, such as use of fibonacci numbers and similar fantasy. The other is...

What do you mean by "optimization run"?

I don't think it has any correlation with profitability (whichever way you define it). But it sure correlates with trading style. If you're daytrader/scalper that makes tens-hundreds-thousands trades per day in many different instruments then you probably wish to have as many monitors as...

It would be interesting to see if all this consumption of unnecessary chemicals has any correlation with trading results. But I guess it's impossible to measure.

Testosterone effect :D

Old news

Shouldn't trading be emotionless and rational as much as possible? Isn't testosterone affecting your emotions/aggressiveness? If so, how does it help trading then? By the way, high levels of testosterone causes baldness.

My opinion about backtesting is that out-of-sample test should be done in the last step of development. Otherwise, if you tweak your system after out-of-sample test, you're just overfitting it to the whole dataset, and out-of-sample testing becomes meaningless.

A few notes on your generator: 1) you use excel's RAND() function, which generates values of uniform distribution. 2) therefore you get daily ranges from 0 to 1 which are uniformly distributed, i.e. chance of any range is equal to any other range between 0 and 1. 3) in reality...

They're random in a way that they're not 100% predictable.