Vendor compare performance
- This topic has 36 replies, 8 voices, and was last updated 1 month ago by
.
-
Hahahahaha
I recall the first “chat” with you where I coincidentally clicked that wrong link and said so. You said something along the lines of liking my wrong click anyway so you weren’t going to do anything about it fore sure. 🙂 🙂Enjoy your weekend !
1 user thanked author for this post.
After 1 full week running (no System stoppages / Rejections) on my Demo Account, the results of the ProfitAlgos are attached.
Re the Comments column … 48 T3 means an average of 48 per trade over 3 trades … the averages per trade include currently open trades (pretend closed at time of this post).
Overall a very respectable performance!
2 users thanked author for this post.
Hello,
We have been publishing since March 2021 a public and open demo allowing to import and launch backtests, with ITF files that not only show past operations, but also continue working and showing ongoing operations, because they have no expiration date). Additionally, we include ITF files with past versions and systems that are still in validation (not yet for sale). The key point is that, for example, if there is an ITF file from May 2023, by launching a Backtest, what you are observing is the Out Of Sample period from May 2023 onward (and I assure you, that’s provides already a lot! This ensures the validity of the version and gives a good approximation of the performance on the period).
More than a year after we started publishing that demo, in May 2022, and following various conversations and proposals at different levels, we published the following post: https://www.prorealcode.com/topic/results-validation-with-our-demo-and-proposal-to-add-demos-in-the-marketplace/ –> We recommend that all interested parties read the post and download our demo at https://cfdautotrading.com/demo, also reading the attached PDF, which explains in detail what the demo provides and why we considered it the best validation mechanism possible under our possibilities (which is in line with what we indicated in the first paragraph). And for information about versioning you can see https://cfdautotrading.com/en/timeline-en/
I trust in the development in progress that Nicolas mentioned (I would say, besides previous personal conversations with us and others, that as far I remember he also commented on it in other posts). It seems to be the most “pure” option and could even be natively integrated in some way. From our point of view, that mechanism should provide the real performance data, demo or OSS (the best that can be technically achieved), primarily focusing on accuracy and the impossibility of data manipulation. Also if possible, It’s desirable that the tool considers and clearly shows the development dates, versions, performance, and other details for each product. Something like generate an accrued data graph of the OSS of each version, for each product.
As you know, there are multiple factors and data to consider when analysing a strategy, and then each person has their own criteria, experience, and preferences… so the critical aspect, from my point of view, is to provide an impartial and objective analysis, not only the final profits / losses.
For instance, we were listed a while ago on best-trading-algos.com because we thought it was a great idea (if I recall correctly, we were the first commercial vendor to appear on it, because we contacted the owner of MetaSignalsPro requesting it). Later, we requested to be removed from the site, as we did not share the vision of the analysis and scoring system, along with other reasons that are not relevant now.
In fact, as I was writing this email, my own conclusion is that I still thinking that my proposal could be beneficial if included as a part of the official solution under development:
– When any vendor publishes a new system or version through the “MASTERFILE”, an ITF file for backtesting should be uploaded at the same time. This would allow for a visible list of ITFs and versions on the product page, with their publication date (something like our demo + our timeline, in this case accessible all from the product page at MarketPlace)
– The evolution would be, if it’s possible, to make some sort of integration where the MASTERFILE (in a single file) are processed internally to both publish the downloadable product and also in parallel run it in server, in a demo or a “semi-real” environment, and the data could be extracted from the server (and I interpret this is what Nicolas is currently working on).However, I am not sure if this tool that is on its way could provide all the information that the detailed report from launching a backtest offers. As I mentioned, we all know that there are multiple validation criteria and preferences (just look at the many threads in this forum or any trading community). Somehow, what I would not like, for example, is for something to display 0-5 stars based solely on the performance of the last three months or something that could mix or confuse Out Of Sample data with ‘nice backtests’. I mean that the preferable could be prepare the same detailed report that we see when backtesting, but using the data from other source.
To conclude, as said, I trust that Nicolas and team will do its best, and that includes to work under the technical limitations (and other kind of limitations) that they found in the way.
1 user thanked author for this post.
Find exclusive trading pro-tools on 
