View Single Post
  #3 (permalink)  
Old 11-23-2007, 09:27 AM
iGoR iGoR is offline
Senior Member
 
Join Date: Jan 2006
Posts: 992
iGoR is on a distinguished road
Hi Newdigital,

I am offering some of my free time to evaluate commercial topics.
But could we agree on some important elements.
There are severall categories of commercial topics.
Lets go true some examples:
- someone who offers his skills as a fundmanager:
To make an evaluation of that, this person needs to show a trackrecord of at least severall years of results and that true REAL statements. To check out all the statements is nearly impossible because it takes way to much work.
So is it worth while to let a person like that open a commercial topic? The risk that is involved for the client or for the person who would be interested is very big. We do not talk anymore of a monthly subscription of a couple of dollars but of investing a lot of money. A fund manager that accepts amounts of money like 1000$ is not a fund manager but someone who dreams to make a quick bug with a system that is highly overleveraged.
If the client is losing his money then there is responsability for this to the forum and also for the evaluator of that topic. So I would not take that risk and sujest that the forum would not allow a topic like that.

- People who offer their programming skills.
It is impossible to know for me or other evaluaters how good a programmer can program. Or how good his after sale service would be.
I can only and strongely sujest that if one wants to have some programming done that they agree to pay half of the money with the delivery of the indicators or EA's and that the programmer gives ONLY the exe. file with a 14 day time clock on it. During that 14 day period the owner of that software can check-out for bugs and let them repair. After that the owner givies his OK then the programmer gives the mq. file without time lock and the owner pays the other half.

- People who offer trading rooms with a monthly subscription.
That highly depends on what the owner of the room is prommising.
If he is prommising that he will call trades and this way people need to take in trades that will filter or proof itself. If the owner calls his trades during a whole month and at the end of the month people are not pleased with the resulls or with what they received, they are not going to pay anymore for that room and they will also post on the topic that the room is crap. Of course there will always be people who find something crap but people will have the possibility to read over many comments (pro or con ) and also the reaction of the owner to those comments.

- People who sell indicators.
It is difficult for an evaluator to see or know if an indicator "works".
What should one expect from an indicator ?... if one expects that a singel indicator will make them profitable then the buyer has the wrong expectations. A single indicator will always and only give some indication. Nothing more nothing less. So the expactions of the buyer should be realistic. Of course if the seller promisses some a Holy grail indicator ,I can understand that a newbie buyer will expect big results. But as said, it is difficult for the evaluator to check the indicators out if his does not have the indicators.
An other problem that happend even here on your forum is that people steel indicators from this or an other forum put them in a new little jacket and sell them...Again difficult to know for the evaluator but he could have a look to it and give comments as: if you want to buy such or such indicator, have a look to this and this that is for free and is doing nearly or exactly the same job.

- Last but very least the people who sell EA's.
There is a lott to say and to write about EA's. And a lott of people will agree and a lott will disagree.
Most of the EA's that people will try to sell, is the code and the rules and the formula of course protected. That makes it difficult to evaluate an EA if one does not ave any clue how the EA works.
Most of the people who sell an EA will only wave with backtest results. I personaly did a zillion backtests and I know what is possible and not possible and where one can or would cheat or where one would do so much "curve fitting" that for shure in the future the chance of making any money is nearly zero.
So I would sujest 2 important elements that need to be looked at for an EA so buyers would be protected:
1) The seller should be cappable of prooving that he develloped his system over a period of data and then do a second test over "out of sample" data.
Example: he develloped and optimized his system over data between 2001 and 2005. Then he has to show the results without anymore optimizing or changing anymore setting over the period 2005-2007. Of course it is not easy to proof this or that an evaluator can judge the correctness of these backtest results because for all the same the seller does an other optimization on the last 2 years.
But 1 very important element is that the seller needs to show results over at least 5 year plus period. Results over 2-3year data have no meaning or no value what so ever. Or the seller does not want to show the results over 5 years ( because then it is not realy profitable) or he did not take the time to purchase that amount of data and should go back to do better home work so he can proof the results over 5 year data.
2) The seller needs to place the EA on a demo or a real account with some sort of a link on his topic that does a continous update so people can check-out the results. Severall people have done this already here on this forum ( also with the steinutz EA).
If the seller refuses to do this then conclusions are easy to make in this way to refuse that seller to have his topic. A second protection would be that the seller need to place that link at least 1 month before he can start to give any information on selling his EA. 1 month is a very short period but over a 1 month period many EA burn already or underperform in a big way (This competion is proof of that--after 1 month already half of the competitors were bunred or losing amounts of money).
So that would give both parties a little bit of time. For the seller to proof and for the buyer to witness. After 1 month the seller can start to realy sell his product by giving information how to purchase his EA. It should explicitly read on his 1st posting that people can not contact him in anyway within the 1st month because he will show proof how good his EA is doing before he will sell his EA.

Some people will say that a trial period is nescessary but I can say from very own experience that a 14 day trial does not give any proof what so ever of the robustness or the profability of an EA. For all the same the EA is performing great but during that 14 day trial it is doing bad. A good develloper knows that also and in many cases only for this reason he doesn't want to give a trial period.
So The best and most realible way is like I said place it on a server where people can check out continiously how the EA is performing.
If the seller is closing down his link after making bad results and comes back with a new link and a new account then that EA should be banned emmidiatly.

Most probably there will be other sort of commercial offers that I did not mension but if one is looking to it with commen sence and a logic aproach to protect the buyers as mutch as possible then the rules to evaluate should always be quite straigth forworth and based on proof of results.
And If some sort of a commercial offer can not be checked or the risk for the buyers of losing serious money is to big, then the risk and the responsabilty for the forum and the evaluators should not be taken and then the best solution is not allowing that offer or topic on the forum .

regards...iGoR
__________________
Succes comes with knowledge. Knowledge comes with experience. Experience comes with time and hard work... (iGoR)
To avoid a lot of false trades ask yourself the question:...Am I desperatly looking for a trend because I want to trade...or am I taking a trade because the trend is there...(iGoR)

Last edited by iGoR; 11-23-2007 at 03:07 PM.
Reply With Quote