ProRealAlgos? Real or fake

• 05/26/2021 at 8:35 AM #170438
  Eric

  “Max Drawdown: 3 536€”

  Is this with 1 contract on DAX? ( 1 Euro per point?)

  https://market.prorealcode.com/product/prta-dax-m2/

  

  Eric

  Participant

  Topics: 18

  Replies: 441

  Been thanked: 84 times
  05/26/2021 at 9:02 AM #170445
  PeterSt

  I say : Fake.

  Always the same guys reviewing. Too many Trustpilot reviews entered on one day. And mind you, Trustpilot reviews will emerge from negative reviews (at least at first). Not from positive ones.

  Always the same amount of Likes in the Market Place.

  Even the same amount of purchases in the Market Place.

  In my view, someone is playing nasty games. But please judge for yourself.

   

  

  PeterSt

  Participant

  Topics: 73

  Replies: 2544

  Been thanked: 829 times
  05/26/2021 at 9:12 AM #170448
  PeterSt

  Oops, one attachment disappeared (probably my bad).

  

  PeterSt

  Participant

  Topics: 73

  Replies: 2544

  Been thanked: 829 times
  05/26/2021 at 10:45 AM #170472
  ProRealAlgos

  Hi Peter,

  Let me ask what exactly do you think is “fake”. Is it the results of our algos? Is it the trustpilot reviews? Is it the marketplace likes? Is it the purchases? Is it our customers?

  Maybe you think it’s all of them so, what the hell, I will just respond to all those topics.

   

  The results of our algos

  This first topic is pretty easy. All our algos can be downloaded on our website. And all our algos have a 21 free day trial period.
  Hence, our results can easily be verified by just downloading the algos and make a backtest or running them live.
  Aside from that we the results are be verified. e.g. directly by our customers and other. More on that can be read on our website

   

  Trustpilot reviews

  You say “Too many Trustpilot reviews entered on one day. And mind you, Trustpilot reviews will emerge from negative reviews (at least at first). Not from positive ones.”
  What do you mean too many on the same day? The first trustpilot review is from February 7th and the latest one from May 7th. I counted the day with most reviews, it was February 10th with 5 reviews and then February 16th with 3 reviews.

  And what would that indicate? Fake reviews?  I don’t understand. We claimed our Trustpilot profile in February 2021 and after doing so we sent out an e-mail to all our customers to please review our business. Of course there will be a peak of reviews after that point. But as mentioned there are reviews ranging over a couple of months.

  And for those who doesn’t know what trustpilot is:
  1) It’s third-party reviewing site. We started using as it’s the most trusted trustpilot site in the world. Other competitors to us have reviews that’s just added by themselves directly on their website. With trustpilot you have NO control or power over ANY of the reviews on your profile.
  2) Trustpilot also doesn’t allow for persons to have more than 1 account.
  3) Trustpilot enforce reviewers to prove that they are a customer of the company they are revieweing. Meaning that they have to send in receipts and other stuff from their purchase.

  I agree to some extent that happy customers are less likely to leave reviews than unhappy customers. But NOT if you kindly ask your customers to. And so far, not a single negative review.

   

  The amount Market place likes are the same

  Today (May 26th 11:42 UTC+02) we have the following number of likes on our four products 32, 36, 35, 35.
  And what is fake with this? Fake likes from fake profiles? I suspect and expect that the prc marketplace would notice and prevent this in some way.
  The marketplace opened just two days ago and we asked our customers to go into the marketplace and give us a like for the algos they enjoy. I don’t know what more to say, apparently some algos were more liked than others.

  If it’s the similar number of likes that annoys you, over the coming weeks you will probably see the number of likes diverge between the algos.

   

  The amount Market place purchases are the same

  They are not really the same, today (May 26th 11:47 UTC+02) we have the following number of purchases on our four products 3, 5, 4, 4.
  The marketplace opened just two days ago so ofcourse the number of purchases will be pretty low and similar.
  But anyway, please elaborate, what is fake with this? Most purchases are from clients that have waited for the PRC Marketplace, and as soon as the marketplace opened we let them know.

   

  Always the same guys reviewing? 

  So as I see it there are three different places to review our company. 1) Trustpilot 2) Product-page on the Marketplace 3) A post on social media

  Do you mean that it’s the same guys that reviewing on all these 3 places?

  • On trustpilot we have 33 reviews.
  • On our product pages we display an image of some reviews of social media tweets. I believe there are 8 reviews there, but there are around 30 posts about ProRealAlgos on twitter.
  • And finally, there are reviews on our PRC Marketplace product pages.

  You can probably find 1-2 persons who have made both a trustpilot review and a post on social media. But hey there are over 60 reviews in total, so to say that the same guys are reviewing is just not true. And I know for the fact that the purchasers of our algos on PRC Marketplace have not made a review on either trustpilot or social media.

   

  In conclusion

  If you ask your customers to leave a review or like a product (only if they actually like it), many will. If I’d say “Hey, there are people who believe all of this is fake. Could you go into the PRC forum and just write your opinion” that would happen too. Wouldn’t mean that it is fake.

   

   

  

  ProRealAlgos

  Participant

  Topics: 6

  Replies: 74

  Been thanked: 28 times
  05/26/2021 at 1:32 PM #170501
  PeterSt

  Hello Marco,

  I like to at least apologise for making you write such an enormous response. Of course this is for your defense, but I think it would have been sufficient to tell that I was BSing you(r company). I was. 😉

  Not to make it a discussion as such, but in order to put my remarks in a better context, hopefully :

  ProRealAlgos wrote:

  Let me ask what exactly do you think is “fake”.

  That would be the out of context remark, only referring to the topic title. With context added it is just plainly impossible to show back test results from many years back, while the forward testing does not exist. So what you present is fake as such. It is all together what you show in order to make things realistic, and personal me does not dig it. And so we continue with my more explicit remarks, now elaborated a little :

   

  ProRealAlgos wrote:

  And what would that indicate? Fake reviews?  I don’t understand. We claimed our Trustpilot profile in February 2021 and after doing so we sent out an e-mail to all our customers to please review our business. Of course there will be a peak of reviews after that point. But as mentioned there are reviews ranging over a couple of months.

  “Construed” (between large quotes). You are correct that not so much is wrong with that, but it will not represent reality. AFAIK people (like me) are not invited to write positive reviews. Ok – oops – what did I say there.
  See the emphasis in your quote (which is mine). It is not this in itself … it adds up.

   

  ProRealAlgos wrote:

  Today (May 26th 11:42 UTC+02) we have the following number of likes on our four products 32, 36, 35, 35. And what is fake with this?

  Statistics depict that you can’t be the only one with only your products who receive so many Likes. If you don’t mind then I combine it with what people tell about your product e.g. in this thread and this does not come across as all over positive (but please skip my previous post :-)).
  That the number of likes are equal for all of them (yes, a deviation of one or two is means equal), well, is smelly. But you explained it ! (for me in aftermath, obviously) :

   

  ProRealAlgos wrote:

  The marketplace opened just two days ago and we asked our customers to go into the marketplace and give us a like for the algos they enjoy.

  So, same story. You are faking things, just because it should not work like that (my book – sorry). It is supposed to be smart, but for me it really is not so much. Btw, it seems that you forgot to ask 33 people to write a review. 🙂

  Etc. Etc.
  In my view, you may be a little bitty manipulating the lot. The internet s*cks anyway, and you help a little with that.

  ProRealAlgos wrote:

  Do you mean that it’s the same guys that reviewing on all these 3 places?

  No, I don’t mean that. I gave you a screenshot as a hint.

  As I said, all adds up. I hope for you others dig it. I just gave my opinion.
  Btw, I tried to stay neutral with respect to the Market Place here (although I used it as an argument). IOW, I could have posted this in the main topic and join the others. Instead I posted it in the topic over here. The stuff about the reviews is not exactly mine (you may have missed that) and all I did was confirming it.

  <hr />

  Marco, this is all not that important. You just can’t pretend that an algo works by means of any backtest you provide yourself. In my book this is thus fake. If you then, next, commercially add with “construed” testimonials (never mind you asked possibly selected others to write them), your algos will work out even less for me.

  It is easy enough : If your algos work out as suggested, then show your own personal results from live. All of them and not a selection. People are asking for that. I really wouldn’t care less, BUT it will be the way to have great sales. Now, it is obviously your good right to say No with any argument you can think of, but if you say No indeed, my personal idea about it will not change.

  Curious for your own personal results !
  Wishing you well … seriously.
  Peter

  PS: A weak one: I am pretty sure that I changed the “nasty” in my last sentence into not-so-nice, but like with the later added attachment, this somehow disappeared. Nasty is certainly the wrong word. Apologies for that.

  

  PeterSt

  Participant

  Topics: 73

  Replies: 2544

  Been thanked: 829 times
  05/26/2021 at 8:26 PM #170554
  ProRealAlgos

   

  “Construed” (between large quotes). You are correct that not so much is wrong with that, but it will not represent reality. AFAIK people (like me) are not invited to write positive reviews. Ok – oops – what did I say there.
  See the emphasis in your quote (which is mine). It is not this in itself … it adds up.

  Answer: Only people that are customers can ofcourse write reviews. Why would it be relevant to read a review from someone who hasn’t even tried the product? As I mentioned trustpilot enforces reviewers to prove that they’ve purchased a product from us. A e-mail was sent out to all customers to write a review, no matter their opinion about us.

   

  Statistics depict that you can’t be the only one with only your products who receive so many Likes. If you don’t mind then I combine it with what people tell about your product e.g. in this thread and this does not come across as all over positive (but please skip my previous post :-)).
  That the number of likes are equal for all of them (yes, a deviation of one or two is means equal), well, is smelly. But you explained it ! (for me in aftermath, obviously) :

  Answer: Well, first of all that’s simply not true. Indicators.net have over 80 likes on one product and AutomaticTrading have 42 likes on one product. And I would suspect that the rest of the providers haven’t asked their customers to leave a like if they like their product (or maybe the have, but the customers are unhappy)
  Secondly, there is no dis-like button in PRC so of-course only the positive (number of likes) will be shown.
  None of the negative comments from this thread are from any customer. If they were, and if they felt like it, they could easily leave a bad review on trustpilot as a verified customer.

   

  So, same story. You are faking things, just because it should not work like that (my book – sorry). It is supposed to be smart, but for me it really is not so much. Btw, it seems that you forgot to ask 33 people to write a review. 🙂

  Answer: Well, it would just be stupid not to ask any satisfied customers to give us a like if they like our product. I mean, it’s still a like from a customer. And only Verified buyers on PRC can leave a review.

   

  Marco, this is all not that important. You just can’t pretend that an algo works by means of any backtest you provide yourself. In my book this is thus fake. If you then, next, commercially add with “construed” testimonials (never mind you asked possibly selected others to write them), your algos will work out even less for me.

  Answer: We show that an algo works by a number of different means, here are three of them:
  1) Ask our customers. There is a big chat group on discord with alot of customers (link to that on our website) But you will probably think all of them are fake.
  2) Run the algos yourself LIVE (NOT A BACKTEST). We offer 21 day free trials on ALL our algos.
  3) We present ALL trades our algos take in a live-stream 24/7. If you want you can record that live-feed and check every trade for themselves and compare to our presented results.

  So regarding backtests, you believe that we manipulate the algos for that. Well.. Check with our customers.. Check with the live feed.. Check with your free demo live period.. Then decide…

   

  It is easy enough : If your algos work out as suggested, then show your own personal results from live. All of them and not a selection. People are asking for that. I really wouldn’t care less, BUT it will be the way to have great sales. Now, it is obviously your good right to say No with any argument you can think of, but if you say No indeed, my personal idea about it will not change.

  Answer: Well, actually we do. Every sunday since August 2nd 2020, we are presenting our results in Youtube live-stream. We go through all trades of all algos in ProRealTime.. You’re welcome to tune in next time.

   

  While writing this message I’ve come to the conclusion that you’ve not done your due-dilligence. All of what I’ve written today can be found on our website if you really had the intent to analyze our business fairly. I believe that you’ve set your head on this, or have an agenda and would never be willing to change your mind. So I will not waste any more time on this. Bye!

   

  2 users thanked author for this post.

  O-jay8, Monobrow

  

  ProRealAlgos

  Participant

  Topics: 6

  Replies: 74

  Been thanked: 28 times
  10/21/2023 at 1:23 PM #222734
  Gauvel

  It’s seem nobody at ProRealAlgos try to build a solid ‘portfolio’. So I have run basic correlation check, and it is alarming. I should have done it before giving away my money.

  Bellow the result for 2013-2023 with all algos except the Nikkei one ( Doesn’t want to run )

   

   

  

  Gauvel

  Participant

  Topics: 4

  Replies: 23

  Been thanked: 3 times
  10/21/2023 at 2:17 PM #222740
  PeterSt

  Hello Gauvel,

  Can you please elaborate some more ? My skills are not of the kind that I understand your pictures. For example, are these all automated systems from ProRealAlgos ? (which you all bought ?)
  And what do “correlation checks” mean ? I seem to see something like matrices where all shown instruments have been trialed on to each available system (from ProRealAlgos ?).

  In any event, thank you very much for your post; it seems important.
  Regards,
  Peter

  Edit, PS: I now see that “PRA” should stand for ProRealAlgos. The further naming I don’t recognize (like Prime and PTAF).

   

  

  PeterSt

  Participant

  Topics: 73

  Replies: 2544

  Been thanked: 829 times
  10/21/2023 at 4:19 PM #222748
  fifi743

  it’s them too see screen

  1 user thanked author for this post.

  PeterSt

  

  fifi743

  Participant

  Topics: 10

  Replies: 912

  Been thanked: 424 times
  10/21/2023 at 4:42 PM #222751
  ProRealAlgos

  Hi Gauvel,

  We are just like PeterSt interested in knowing more what the images mean and how they were produced. If it’s real live data or backtested data?

  So the images are showing some kind of “correlation”, but in what exactly? I suspect it’s correlation in results, but is it % winning trades or €€€ gains/losses. If it’s €€€ gains, which position sizes are used on each algo? Which spreads? Which timeframe is this correlation present; daily resullts, weekly, monthly? On the scale, what does a negative value mean, what does a positive value mean? What does this correlation mean in terms of Max DD? How many years of data is this analyzed on?

  I see that more than 20% of the algos are missing data in the bar chart e.g. Cuts US100 M5b, FX USD/JPY H1, PATF DOW M15 and Plus Nikkei M5. What does that mean?

  Many questions, but let’s get the talk going!

  I want to clarify that we don’t agree with the statement that our algos have a high correlation between one another. With 19 different algos it’s obvious every algo will have a correlation in single months or quarters with any of the other algos.

  The most productive way in the further discussions here would probably to look at a single algo. to get more specific and look how its results correlates to other algos. Looking forward to your comments!

  Best regards / Carl

   

  

  ProRealAlgos

  Participant

  Topics: 6

  Replies: 74

  Been thanked: 28 times
  10/21/2023 at 6:11 PM #222754
  CFD AutoTrading

  Hello,

  let us only clarify for @fifi743 –> ProRealAuto is a vendor name we own, alongside IBKR AutoTrading and CFD AutoTrading. These 3 vendor names are all associated with ProRealTime platform and all becoming to our company AutoTrading OÜ, website: autotrading.ee (in fact, the rest of vendors/usernames you can see in the image are only projects at this moment).

  Therefore, PRA and the other acronyms of algos in the @gauvel message in this thread are related with ProRealAlgos, a business not related with us.

   

  Thanks

  1 user thanked author for this post.

  PeterSt

  

  CFD AutoTrading

  Participant

  Topics: 4

  Replies: 58

  Been thanked: 21 times
  10/22/2023 at 12:50 PM #222779
  ProRealAlgos

  CFDAutoTrading is right, ProRealAuto is, even though very very very similar in name not a part of ProRealAlgos. When thinking about it, one could have easily forseen that the confusion/mixup of company names would occur when deciding to name a company so similar to a competitor. Not sure if that was the intention, but a cynical person would say that it was. So this mixup will probably continue to happen over the lifetime of our businesses, but that’s on ProRealAuto and not on us as ProRealAuto was created years and years after ours.

  

  ProRealAlgos

  Participant

  Topics: 6

  Replies: 74

  Been thanked: 28 times
  10/22/2023 at 2:20 PM #222782
  Gauvel

  Can you please elaborate some more ? My skills are not of the kind that I understand your pictures.

  Answer : It indicate correlation between loosing positions for each algos.

  For example, are these all automated systems from ProRealAlgos ? (which you all bought ?)

  Answer :  Yes all of them.

   

  And what do “correlation checks” mean ? I seem to see something like matrices where all shown instruments have been trialed on to each available system (from ProRealAlgos ?).

  Answer : Check correlation on losing trades

   

  Edit, PS: I now see that “PRA” should stand for ProRealAlgos. The further naming I don’t recognize (like Prime and PTAF).

  Answer : Prime and PTAF are algos recently added by PRA

  

  Gauvel

  Participant

  Topics: 4

  Replies: 23

  Been thanked: 3 times
  10/22/2023 at 2:23 PM #222783
  Gauvel

  Hello ProRealAlgos,

  We are just like PeterSt interested in knowing more what the images mean and how they were produced. If it’s real live data or backtested data?

  Answer : Backtested data

   

  So the images are showing some kind of “correlation”, but in what exactly? I suspect it’s correlation in results, but is it % winning trades or €€€ gains/losses.

  Answer : True, correlation on the ‘Perf Abs’ columns, €€€ on losses only.

   

  If it’s €€€ gains, which position sizes are used on each algo?

  Answer : No €€€ loses only (more meaningfull), minimum/default position size

  Which spreads?

  Answer : Default spread, minimum

  Which timeframe is this correlation present; daily resullts, weekly, monthly?

  Answer : Daily correlation

  On the scale, what does a negative value mean, what does a positive value mean?

  Answer : Negative value = Good, non correlated / Positive value = Bad, correlated, non ‘hedged’ algo

  What does this correlation mean in terms of Max DD? How many years of data is this analyzed on?

  Answer : The maximum I could put my hands on, 15oct 2013 – 15oct2023

  Question : By the way why for some algo backtesting only start in 2015 ? On those instrument volume start in 2016, it’s kinda weird.

   

  I see that more than 20% of the algos are missing data in the bar chart e.g. Cuts US100 M5b, FX USD/JPY H1, PATF DOW M15 and Plus Nikkei M5. What does that mean?

  Answer : String in the Perf Abs ( Didn’t take the time to convert it ($, Yen) )

  Many questions, but let’s get the talk going!
  I want to clarify that we don’t agree with the statement that our algos have a high correlation between one another. With 19 different algos it’s obvious every algo will have a correlation in single months or quarters with any of the other algos.

  Answer :  The current portfolio is currently in a > 7500$ drawdown ( minimum contract size, excluding PRA FX USDJPY / H1PRA PATF DOW M15 , might reduce DD by ~500$).
  Question : What are the reasons behind it ? Overfitting ? High Correlation ? Both ?

  The most productive way in the further discussions here would probably to look at a single algo. to get more specific and look how its results correlates to other algos. Looking forward to your comments!

  Answer : That’s exactly what correlation matrice checks

   

  

  Gauvel

  Participant

  Topics: 4

  Replies: 23

  Been thanked: 3 times
  10/22/2023 at 2:27 PM #222784
  CFD AutoTrading

  @prorealtimealgos: People might also think that “ProReal” comes from ProRealTime, and anyone can choose the name they prefer.

  Like you, who initially called your business ProRealTimeAlgos, and probably need to change to ProRealAlgos due to copyright or naming agreements (in fact you remain here as “prorealtimealgos”)

  Like me, who initially named the CFD AutoTrading profile @cfdauto (now @cfdautotrading). And when I named ProRealAuto, I thought that “ProReal” is not yours and that having a name starting with it makes it easier to be found and related to ProRealTime.

  Even cynical people may not see anything strange in the previous points, which are commercial and marketing decisions.

  Commercial and marketing decisions are also creating a profile called Simon as “CEO”, or making a challenge with “Marco”, or talking about a colleague “Damiano”, or the other three guys who were supposed to be part of the ProRealTime team.

  [ Edit: one phrase removed after user’s edit time expiration, at post author’s request ]

   

  Honestly, I’m more interested in the thread topic instead of continuing the stupid game between us. I’ve tried to quit many times. Now you have a new opportunity to move forward in a positive direction and focus the thread after the @gauvel comments, thanks.

   

  

  CFD AutoTrading

  Participant

  Topics: 4

  Replies: 58

  Been thanked: 21 times
• Author
  Posts