Today I decided to test a fairly new system - ES Logic. Even though it has a very short history it had already accumulated well over 100 trades of which the statistics are very favorable. Well, after about 10 trades through half of the day I sent the following message to the developer:
"Hello, I have to tell you that your system - at least with my set up which is trading at Interactive Brokers through TradeBullet - does not work. I subscribed yesterday. At the time I turned it off today at 12:45 ET your system page showed a profit of $212 while the actual trades in my account after commission had produced a loss of $291 and that does not take into account a profitable trade of $325 on your page which for some reason was not executed in my account, so the actual loss for me was $616. I cancelled the subscription. Please make sure that I am not charged for it. Thanks "
This was the response of the developer:
Please address all these questions to C2. All I do is provide signals.
This is the reason why it would be so helpful to see if the developer auto trades his system on C2 to show that he has enough confidence that it works in real life.
Please address all these questions to C2. All I do is provide signals.
The answer is correct. The system vendor isn’t responsible for execution of your trades. Also you agreed with terms of the subscription when you subscribe to the system.
Eu
Two thoughts:
1) Notice this stat: "Keep after worst-case slippage 49.2% " this is worrisome
2) Scalping systems (avg trade is 19 minutes) combined with autotrading never seem to work on C2. If you were going to trade this, I would suggest you NOT autotrade, but instead put the signals in manually as possible as Limit orders to attempt to get positive slippage. But, each to their own.
"This is the reason why it would be so helpful to see if the developer auto trades his system on C2 to show that he has enough confidence that it works in real life."
The problem with this statement is it implies if someone DOESN’T autotrade their own system it must mean they don’t have any confidence in it. This is absolutely untrue and highlights people forget that there are many many reasons why someone might want to become a vendor on C2 such as building an independent 3rd party verifiable hypothetical track record for potential clients, investors or employers.
One of the most important factors in selecting a system to trade should be that the vendors and subscribers objectives are aligned, this is one of the most basic premises to the building blocks of any system and yet is rarely explored by anyone.
Good and constructive points.
This is a valid point but it could be easily addressed by the vendor clearly stating why he is putting the system on C2
This is a valid point but it could be easily addressed by the vendor clearly stating why he is putting the system on C2.
It could but why should the vendor have to do that? Here’s an idea, how about the subscriber actually asks the vendor a question about his system before subscribing to it? How novel.
C2 provides enough stats for making a decision. What a system vendor suppose to do more? Lap dance?
Also, the system subscription terms is “Pay for performance” if Karl unsubscribed at the same day he won’t be charged. The system vendor answer is still correct.
Eu
If the vendor does not state this in the Systems Description - in this case the vendor did - I always ask him this question beforehand. But this does not answer the more important question, as this example shows, if the system works with auto trading at C2.
You guys look at it only from the vendor’s viewpoint, I wish you would put yourself in the shoes of the subscribers. A subscriber is not interested in a system which fails in real life auto trading no matter how brilliantly it performs on the developer’s own system.
I agree w/ Karl - there’s something that makes it hard to have confidence picking a system here.
For me, the first flag is the annual % performance calculation and the lack of subscriber testimonials that make me nervous, i.e., w/ all the systems making 100-500% per year, the subscriber testimonial bin should be overflowing because these are instant millionaire type returns (6 years of doubling from any starting point to a $1mm).
My first thought in helping subscribers begin evaluating systems would be a normalized return statistic that used 1x leverage as its base (a 1% move in the underlying market causes a 1% return in the P&L) and monthly compounding in its calculations to give subscribers a common baseline to start from.
I love the site and the concept, offered constructively,
–h
Agreed, but you have all this information at your disposal already.
1. real trade stats available - this system has been autotraded - show real trade from subscriber broker accounts
2. what this page displays -
overlay autotrade results
verified autotrades only
3. keep after worst case slippage.
Together with the plethora of stats available on the system page and the advanced stats section, is this not enough for you to work out whether the system actually made any money in real life or not?
"Agreed, but you have all this information at your disposal already. "
I wouldn’t even try to evaluate a trading system w/o calculating its t score (statistical significance) and optimal f$ (max leverage in contracts/$) values. Those two calculations require the standard deviation of trades and the largest losing trade - neither of which is available w/o a lot of manual calculation for systems on the site.
The point still stands, I think, that it is harder rather than easier to make good fast evaluations of systems on the site,
–h
