Lessons Learned From Profitable System Development

WDGann · Sep 13, 2003

LOL...

Yes, in another words, the forum directly above us is all BS...

and 95% of what you read in the Forum on the far top is BS....

HE he he!

vanilla2 · Sep 13, 2003

Quote from brokerj:

Amen to testing any system, Paper Trade in a realistic fashion until you can produce CONSISTENT results before tossing cash on the table. Good trading to all.

just want to clarify, my advice earlier in the thread to trade the system and not sleep on it or forward test it for months assumes that one has already performed rigorous backtesting and evaluation. if a system has so many degrees of freedom that a forward test kills it dead, it was a bad system from the start, and not worth consideration in the first place.

in my systems, i'm looking for extremely simple rules. embarrasingly simple rules.

Bolts · Sep 13, 2003

Quote from WDGann:

LOL...

Yes, in another words, the forum directly above us is all BS...

and 95% of what you read in the Forum on the far top is BS....

HE he he!

Yes, there must be an awful lot of traders who never backtest at all. It becomes so immediately obvious once you start backtesting just a little bit. In my very brief backtesting experience the only things I've found that work well at all are very unorthodox or counter-intuitive.

WDGann · Sep 14, 2003

Quote from vanilla2:

just want to clarify, my advice earlier in the thread to trade the system and not sleep on it or forward test it for months assumes that one has already performed rigorous backtesting and evaluation. if a system has so many degrees of freedom that a forward test kills it dead, it was a bad system from the start, and not worth consideration in the first place.

in my systems, i'm looking for extremely simple rules. embarrasingly simple rules.

For the 1st time in ET, I agree with ya...

My setups and tests I run are very simple. But the evaluation process of the edge is somehow relatively complicated.

but also

WDGann · Sep 14, 2003

Quote from Bolts:

Yes, there must be an awful lot of traders who never backtest at all. It becomes so immediately obvious once you start backtesting just a little bit. In my very brief backtesting experience the only things I've found that work well at all are very unorthodox or counter-intuitive.

Yes.... working with discretionary traders... the profitable trades are usually the ones that break the rules for the indicators they use...

"Hey, but you bought Minis when the Stochastic was above 70, you told me you go short when a reversal occurs... blah blah blah"

"Ummm... I don't know... the market looked like it was going to go up."

When it comes to intuition and technical analysis... Getting an intuition against the rules are good. Breaking it and changing it is a completely different thing.

brokerj · Sep 15, 2003

acrary · Sep 16, 2003

Quote from vegasoul:

Some of your other sub-catorgories (mid day continuation breakouts, time of day based breakouts, trend day breakouts)
seem to be pattern-based( definition of a trend-day?) , from my relatively little experience, these system seems to trade less (less overall profit) because the patterns act as stict filters and they also "seem" ( I am being very subjective here) to be less robust due to the increase in the number of parameters to discern the patterns.

I had some success in developing such system( their profit factor always beats the plain-jane breakouts) but they have never made into my arsenal because I always thought I was curve-fitting. But you have really made me think again. Are such system less robust to a extent than other breakouts systems?

I define a trend day as a day where the open is within 10% of the low/high and the close is within 20% of the high/low. The % is based on the range of the day. Ex. In a day with a 20 pt range, the open must be within 2 pts of the low/high and the close must be within 4 pts of the high/low. In the first 6 months of this year there were 18 trend days that met this definition.

I've found the other types of breakouts to be just as robust as the PROD5 model I posted. They do have a smaller frequency of trades, however. Below is the monthly report for the past couple of years using trend day breakouts so you can get an idea of how it's performed.

My other question is how do you allocate capital to these system.
Because the trading frequency/overall profits of the models differ, do u give different weights to system to balance it out or do u risk the same 0.5 percentage( I remember you said this somewhere) on all models?

As another complication, do u use a seperate equity curve for each model or lump them together for position sizing purposes. So that if a system does better, the profits are invested back into the system?

More...

I was just using .5% of total account equity per-trade per-model. If one model had a trade, then I'd have .5% risked. If two models were in a trade, I'd have 1% risked, etc. All my models have a edge so I don't favor one over another.

Remember, I'm trading against the models using discretion to improve the results beyond the performance of the individual models. This gives me something to do during the day and keeps me from being bored to death while waiting for the day to end.

My only real work with the models once they're in production is to do the edge test quarterly to make sure the edge hasn't deteriorated. If it has, the model is replaced with another one.

OddTrader · Sep 16, 2003

Quote from TriPack:

1) Filtering out smaller moves is important. This does two positive things for the system, reduces the # of whipsaws (entry in a flat market) and increases the % profitable. One side-effect is that whenever you filter you introduce lag.

Anyone would like to explore/discuss the above challenging lesson, one of the very interesting points in this thread?

Walther · Sep 16, 2003

Quote from OddTrader:

Anyone would like to explore/discuss the above challenging lesson, one of the very interesting points in this thread?

I disagree on that lag thing . If one will use daily or intraday cycles as a filter, there is no lag . Trade will be taken only in vicinity of a time specified . You know , ahead of a time , that there will be a trade if your method confirms it.
Walter

CharlesXT · Sep 16, 2003

This is my first post here and I would just like to say that I have been overwhelmed with so much great advice and help. This board is fantastic. I would especially like to thank acrary for so many outstanding posts. This information is priceless and in many cases not available in any book, but rather comes from experience.

I began trading in 1987, in fact my first trade was buying 10 put options on the friday before the black monday crash. It was as lucky as a person could be. I was running a very early version of metastock on a tandy 1000SL computer. It is so slow to get the charts up on the screen, but it was quite challenging and fun.

I had to stop trading for quite a few years due to many family problems, critical illness's, but I am now in a position to get back into some real time trading. I just feel so lucky that I found this board. It makes things a lot easier and I have found a clear direction for my research.

So once again, thanks to all of the great traders here and good luck to you all. Thanks to the Baron also for such a great board!

Lessons Learned From Profitable System Development

Attachments