Going public with a system ruins it?

dbs119 · Feb 23, 2005

I have seen a lot of widely available systems that actually work..
I have mentioned here that i have read 2-3 books from the 80's whose systems are still performing nice. OF course they sometimes have their down perids ( like 1-2 years of drawdowns) but hey, that's what you get for an "ancient" not-very-optimized system.
I do believe that even if i post rules for an extremely profitable system, at the first losing trade everyone is going to turn to the next holy grail...Its just human nature..Greed and fear is name of the game..
There's nothing new under the sun....... Everyone that has read any history knows that..
Good night to all!

stock777 · Feb 23, 2005

None of those 'indicators' mentioned are systems. H&S, Bollinger , are not systems.

A system can be executed by a machine or a trader acting like a machine. Everything else is seat of the pants.

Systems that are 'automatic' can't survive in the wild very long for obvious reasons.

nitro · Feb 24, 2005

Quote from stock777:

...Systems that are 'automatic' can't survive in the wild very long for obvious reasons.

I am not sure what you mean by this, but if you mean what I think you mean, I don't agree.

nitro

NickelScalper · Feb 24, 2005

Systematization:

Input (market data) .... System (processing of data) .... Output ("buy", "sell", etc.)

stock777 · Feb 24, 2005

Quote from nitro:

I am not sure what you mean by this, but if you mean what I think you mean, I don't agree.

nitro

I don't know what you thought I meant so I don't know if I disagree with your disagreement.

To add to the confusion I'll clarify.

A reliable winning money machine made PUBLIC will not survive for long.

rickty · Feb 24, 2005

Quote from stock777:

A reliable winning money machine made PUBLIC will not survive for long.

To prove your above statement incorrect I will provide one counterexample. Richard Saidenberg's "R-Breaker" S&P futures trading system was released to the public in 1993. Since that time it has consistently been rated as one of the top ten S&P systems by Futures Truth, as it is presently:
http://www.futurestruth.com/topten.htm#S&P Systems

Richard

NickelScalper · Feb 24, 2005

Quote from electron:

That is the problem here as some of us noticed it already too. Catching phrases such as 'there are no secrets in the markets' are nice, but the reality is dirty and hiding it beyond the facade of catchy phrases gets you nowhere... The demand and supply will catch with you sooner or later depending on the system popularity, the market liquidity, the system entries, etc. Every system and every market should be considered individually and if you believe that you are better off keeping your secret to yourself then do it even if some people would tell you that there no secrets. That does not yet mean that the system will be ruined, but it will be affected in some way and thus less efficient.

Of course, it's a relative matter. The more liquid your market, the greater tolerance your system will have for multiple users.

stock777 · Feb 24, 2005

Quote from rickty:

To prove your above statement incorrect I will provide one counterexample. Richard Saidenberg's "R-Breaker" S&P futures trading system was released to the public in 1993. Since that time it has consistently been rated as one of the top ten S&P systems by Futures Truth, as it is presently:
http://www.futurestruth.com/topten.htm#S&P Systems

Richard

So you use this system and make good money. Congrats. Oh, by the way, if you release 1000 systems , do you imagine that all of them will fail at the same time, or ever? The trick is to figure out the ones that won't fail before you lose your shirt.

But since you've been on that roll since 1993, you don't have that problem!

TSGannGalt · Feb 24, 2005

Non-public systems get ruined. So?

AnomalyResearch · Feb 24, 2005

http://www.mechanicaltrader.de/randomequity.htm

Go play with this equity curve simulator using nearly random but nontheless edge containg ratios such as win/loss ratio of 1.01 and trade probability of 50%. You can see that even with an edge there are some sample runs which produce losses. Does an edge ever go away or does it just hit a bad "sample run"? Of course if one cranks up the ratios, i.e. creating a stronger edge, it is almost impossible to produce a losing run. I believe some successful, long lasting systems dance at the edge of non-randomness and manage to shake off less confident and less capitalized adherents during random periods of seemingly non-edge performance.