Optimizing or curvefitting

Hypo, your approach to all of this is something i do also. My system(s) are so simple, it is something i would be embarrased to market. They are based on ideas (or counter ideas) that i observe, and i have conviction in the reasoning and methods.
As far as Optimizing or Curvefitting, the amount of parameters involved are few - usually less than 3 (hence, simplicity). But, boy to i optimize those numbers.

I also stick to one contract / one market, with success. I think it all comes down to determining the personality of the market you feel comfortable with. i have tried to take these ideas which work well on one and test them on different markets, and get results which are so - so. I can adjust numbers, but i decide to stick with the one (market) i have and like. The comparison to a parisitic market is exactly what it is, and is how it feels every day.
Sure, I would like to find one that tests THE SAME across most major markets. But what i do now is ok, and as you said, i monitor it until it stops doing what the historical returns say it should do. At that point, it is wise to "tighten up" the parameters, and make changes which yield lower returns, but keep the thing alive.
Thanks for your posts.
 
...there is no hope for you if you think like me. I am making this all up as I go. Anytime I can debate with my betters like OddTrader (or OldTrader, or any others from the system mafia), I love it. Have to spin some outrageous bullshit, though, to keep 'em interested. Paranoid delusions, statistical mumbo jumbo, seemingly rigorous development rules, etc.

Re your three parms, though, I have to ask, there really are more, aren't there? I mean, anywhere in the code there's a Boolean test (even implied), it's a system parm. My systems have at least seven:

- the issue

- the bar period

- the earliest time to enter the trade

- the latest time to exit the trade (or the maximum trade duration)

- the entry threshold for the trade entry parameter

- the stop

- the target profit.

Sometimes I add one more, such as the latest time to enter the trade, or a second threshold on the test parameter. Are your three on that list?

Also, what on earth leads you to the paranoid theory that markets are manipulated? For example, you wouldn't expect todays collapse to imply that there's bad news tomorrow, would you?
 
Quote from hypostomus:

...do you think that increasing liquidity (more than NQ, say) translates to less manipulation?
More...

my 2 cents here:

Actually our servival in simple words is depended almost Completely upon the Big Guns' manipulations, provided the actions are in some replicate forms/patterns, statistically.

Without efficientlly high liquidity, the number per period and their associated odds of replicate forms/patterns would be lesser. This chiefly determines our win rate.

Possibly, the forms/patterns of actions could be changed from time to time, whenever they like.

Ideally imo, a good system should be effective enough to detect and recognise their actions for newly created forms/ patterns (Big Guns make decisions, we follow their decisions).

:confused:

PS: Sorry for I don't know what I'm talking about. :mad:
 
Oh boy. I guess you got me.

Your list:

- the issue

- the bar period

- the earliest time to enter the trade

- the latest time to exit the trade (or the maximum trade duration)

- the entry threshold for the trade entry parameter

- the stop

- the target profit.

I do use All of these. These are my only parameters, on three of my systems.
My "stop" is just max i can stand to lose. Never really alter that number - my systems dont need it. I limit my losses with entry and exit times.
BUT.....In my defense....
The reason i am thinking three: The toughest ones for me (the ones i really crunch the numbers on) are target profit, entry and exit times.

The other ones i dont play with. They are just set and work, based on i guess what you consider the "bar period" (the condition, which i consider to be the idea, which is a constant in the whole equation from the get go).

Man, you keep me on my toes. If in fact you spat jibberish,
I am a living example of your jibber jabber (development as a simple excursion)working. Because thats all i do. Your delusioned mind has alot of merit, according to my account statements (for now at least. I am always thinking of the next trade).

All in all, keep posting. Spew some more, if you can.
 
"For example, you wouldn't expect todays collapse to imply that there's bad news tomorrow, would you?"


Ok. I'm short now. Thanks for the tip. :)

When creating my systems, i ran into that dependency idea. I decided to not even persue it. In other words, I only look at whats happening now. Dont care about yesterday.

Now... that doesnt rule out that the NOW (the trade i enter) this market is not being manipulated. I believe it is. But the precept of the manipulation is important to know (the idea behind the manipulation for the moment - this is important to know). And thats the basis for all this hub bub. why we create systems. otherwise, it would just be noise. Damn it, Hypo, its not noise.
 
Quote from hypostomus:

... Anytime I can debate with my betters like OddTrader (or OldTrader, or any others from the system mafia), I love it.
More...

The truth is I am Really a newbie, something that is hard to confess publicly. :mad:
 
Quote from harrytrader:

http://www.elitetrader.com/vb/showthread.php?s=&threadid=28858&perpage=6&pagenumber=10

When under backtest such indicators seem to perform well, it's almost due to persistency of trend (see Arcsinus law: distinguishing trend from persistency of chance http://www.elitetrader.com/vb/showt...tency+and+trend) that is to say that they are meaningless inferentially speaking as for pertaining to their performance in the future ! That's why the backtest on these indicators are flawed from their very ground and all datas about returns and decorations to appear "scientific" are just cosmetics that won't erase that ! Only those who confuse EMPIRICAL statistics which concerns past performance only and ignore the INFERENTIAL statistics would be fooled.

So pure MECHANICAL SYSTEM based SOLELY on these kind of indicators are due to be DOOMED SOONER or LATER. Of course there is always a chance that they don't fail but it is due to chance, for example such a system for daily scale could outperform during the past 20 years of Bull market but on the next "sampling" of twenty years you don't know. Being impressed by such a performance, especially if it is compounded, is to be fooled by probability fallacies.
More...

Same thing for trend's model (since it is the very subject of the thread). Remember or not this thread about the "genious" ma "predicting" trend marvelously whereas it is after hindsight and a persistent trend of market.

http://www.elitetrader.com/vb/showt...=16684&perpage=6&highlight=trend&pagenumber=1

Yang on Moving Averages
Astrikos is free this week so I cut and pasted something from there here that should interest many.

Astrikos hardly ever has a free week, so it's definitely worth checking out. To get in the rest of the week the username is free and the password is pass.

Cheers.

By Rainsford Yang
Wednesday, April 23rd 2003 9:00pm ET

There's been a lot of press recently concerning the fact that the S&P500 has closed above its 200-day moving average, which sounds a lot more impressive than it really is. We've done extensive research on moving averages, and as far as the major stock market averages are concerned, a close above the 200-day moving average is absolutely meaningless. In fact, simple moving averages in general, except for the very short-term ones, should be disregarded in my opinion. Our studies clearly show that the best performing moving average (the 1-day average) isn't an average at all - it's simply today's close. If it's higher than yesterday's, the trend is up. If it's lower, the trend is down. This most basic of 'strategies' blew away all longer-term moving average strategies by a mile, as you'll see by the results of the studies below.

And my answer was:
http://www.elitetrader.com/vb/showthread.php?s=&postid=259983&highlight=trend#post259983


Re: Yang on Moving Averages
SQRT((261*(2002-1970))-1)*1*100*3 = +/- 27415 %

this is just a very rough order estimation of performance for buying or selling randomly between 1970 and 2002 that is to say a person could make a POSITIVE performance of an order of 27415% and an other a NEGATIVE performance of - 27415% and not be more or less competant than the first one whose people could think he has a financial genious touch whereas it could be just by chance . And since I used an understimated law by taking a random normal law this order should be much greater in reality and reach 50000 or 100000% so that the numbers below are not significant statistically and that's what economists have already said .

This is a classical flaw in Stock Market similar to the gambler's flaw in casino gambling. The more frequent the trading and the more expanded the total time period the more big the cumulative percentage. As an other consequence and for exactly the same reason it is also a flaw comparing 1 or 2 days moving average results with 200 days moving average because the law of variation is not the same and is more volatile by definition for short than long MA.

This is kind of myth like the martingale for gamblers keep people abreast of finding the holy grail for making eally rich without any effort ... and real edge . Of course counting on chance it's always possible but it is also possible to get ruined as rapidly : -50000% could be as probable as -50000% cumulating all the years.

The vendors of trading systems use the same kind of trick for showing superformance. In fact many if not all trading system testers are flawed with that kind of presentation tool when it is the main if not single possibility of visualising performance. This is only "apparent" performance not true that is to say comparable performance.
 
Part I

Quote from OddTrader:

my 2 cents here:

Actually our servival in simple words is depended almost Completely upon the Big Guns' manipulations, provided the actions are in some replicate forms/patterns, statistically.

Without efficientlly high liquidity, the number per period and their associated odds of replicate forms/patterns would be lesser. This chiefly determines our win rate.

Possibly, the forms/patterns of actions could be changed from time to time, whenever they like.

Ideally imo, a good system should be effective enough to detect and recognise their actions for newly created forms/ patterns (Big Guns make decisions, we follow their decisions).

:confused:

PS: Sorry for I don't know what I'm talking about. :mad:
More...

Part II

Q

Moreover, UK foreign exchange reserves were fast being wiped out in that defensive effort. In 1992, the global currency market's daily turnover was the equivalent of USD880 billion, according to the Bank of International Settlements (BIS) tri-annual survey.

--- Currency Strategy (by Callum Henderson)
UQ

Relative: ALMOST every trader in currency markets is merely a relatively small trader. :D

PS: I'm more confused now! :mad:
 
Hi, indahook

Probably the only way to create a good trading system is to use an AI software, curve fiting being cutt of and the sistem self-adjusting with every new tick.
If you trade Futures, see also THE ENCICLOPEDIA OF TRADING STRATEGIES by Jeffrey Owen Katz and Donna L McCormick. The title is missleading, it is not an enciclopedia, but it is a good book for entrance and medium level, using C++
 
You must log in or register to reply here.
Back
Top