Sunday, August 21, 2011

Collective2 review - not a real scam

First, the good news:

Performance of Project Athena is 17% in 2 months! But if we would weight in the fees, the real performance is between 9% and 15%, depending on your broker and how much you trade.




Started$10,000
Buy Power$11,738
Cash$11,738
Equity$0
Cumulative $$1,738
Total System Equity$11,738
Margined$0
Open P/L$0 Refresh System Stats


And some more technical stats:


Trades131
# Profitable67 (51.1%)
Avg trade duration7.9 hours
Annual return (compounded)133.5%
Average win$84
Average loss$60
Profit factor1.4:1
Max peak-to-valley drawdown (historical)5.39%
drawdown periodJuly 18, 2011 to July 28, 2011
Correlation w/ S&P-0.211



The bad news are that I have managed to piss off C2 owner with my 'constructive feedback' on C2 forum.

I will list here my last threads with him, I will paste only my mails but I will just present succinctly his replies, I have a feeling he will take it personal if I paste them here.

Here we go:

C2
C2 asked me to stop posting on and/or using C2, and they are happy to issue you a full refund for my system. Followed by 'Your continued desire to torment me borders on the personal and vindictive.'

 Me
Hi Matthew,

It is not my desire to torment you, I am sorry if that what you think.

I listed some issues (some facts and some personal opinions based on the facts), you might not like them, not that I expected that.

For example is it true or not that there is no way to evaluate the platform without having to pay for it? True/False? AFAIK this fact is true. And my opinion is that the implication is this: system developers will do less testing, and subscribers will end up paying for this.

You can disagree with my opinions and you can change the facts! It is up to you!

I payed for 6 months, if you plan to give back the money, just add a "money back guaranteed" for all developers and I might exercise this feature. If you feel my posts are disrespectful (which I do not try, I just try to express facts, and my opinion based on the implications of those facts) then feel free to ban my account.

Fell free to respond on forum and show what facts I am misunderstanding, or why my logic is broken in the mater of opinions. 

C2
Something in the lines of "Just stop, I have no interest to debate with you. "

Me
Well, then please change the EULA (e.g. you are not allowed to post negative opinions or critics on the forum ...) and I will stop.
Or ban my account if you find that my comments do meet the bar for banning an account, or delete them.
I don't ask for special privileges and I like the freedom of expression, as long as I don't insult and I try to state facts and use my (poor) reasoning and opinions based on my experience in software engineering.
I call my comments ... constructive feedback.
C2
I lengthily email, basically "Leave me and my site alone". Also he said I am causing him pain with my comments, and he does not not care the reasons for which I harass and torture him, like constructive criticism, stating facts, protecting customers, and many more. Also he send me full refund and asked me to not use the site or use it in silence. He also said that he made some business tradeoffs and while maybe not the best ones, I should not complain and go somewhere else if I want.

While now it is clear C2 is not a scam, it is clear that (like many other business) it does not have it's customers interest at heart. But at least other companies try to mask it better. How much does C2 care about both subscribers and system developers can be seen from C2 forum.

C2 owner:
I always tell system developers who get upset when people unsubscribe from their free trial period, "Relax. You're not running online jewelry store http://bluenile.com. You're not selling diamonds through the mail." Your marginal cost of offering a free trial is... well, it's zero. Okay... perhaps your point is right, that it costs some of your time to answer questions from your free-trial customers. Fine. Then the marginal cost of providing a free trial is not quite zero. But it's darn close. 

My private reply:
I am stupefied about the attitude! ""Relax. You're not running online jewelry store http://bluenile.com." is not the kind of thing I want to hear from a business owner. I want to hear how the product will improve, or at least to be presented with viable workarounds. It is not about providing a free system or not, it is about listening BOTH sides and provide a fair product.
I have not seen any response about my requests, nor agree, nor disagree, nor a "we are put this in our feature request list" ...

 And some of my replies to forum posts have started to disappear, like this one in response to someone losing $200 because of a C2 issue:
And it will get ugly one day! 
C2 lacks big time in the quality assurance sector. 
1) the testing systems are only for developers and are invisible to subscribers. It should be the case that a developer could build a testing system, and (optionally) let subscribers use it to simulate and test paper trades. 
2) system developers have to PAY to have a testing system, so it is understandable why most developers are not aware of the issues, because they need to pay extra just to evaluate the platform. 
3) I have seen over and over small issues like this! 
4) C2 does not maintain a public list of "known issues". I worked for many years in software quality assurance, and believe me, EVERY software has its list of known issues which will probably never get fixed because it is not economically feasible. And if C2 can make money for 8 years with a host of internal known issues, they will, like any other company. C2 should disclose them and their workarounds, YOU can lose a fortune! 
5) Ultimately, C2 will make money just because system developers will pay the monthly fee, and C2 will resist with current quality for a long time, but will you, the subscriber, resist? 
Take time to read the forums. I have listed many other opinions. Both system developers and subscribers have to be very careful while using this site. It runs $400 millions weekly, and the possibility for large screw-ups are endless!
My assessment of C2 platform

Not a scam, but based on C2 posts it seems C2 is in Maintenance mode with little or no code development. I am wondering if there is any active development team. Read - don't expect much bug fixing, nor expect improvements. Whenever there is an issue they seem to fix it in the database, but the underlying issue/bug seems to remain there happy to hit someone else. So scan the forums, and whatever issues you see, keep in mind they might bite you too!

Second, C2 owners did not seem to disagree much with my assessment, but this is the way site works and they don't plan to change that. It is their business model, and they know well that the improvements asked by many users would be good for both system developers and subscribers, but it will hurt their bottom line. For example a highly asked feature is to evaluate the platform at no cost, this would make both system developers and subscribers be aware of any issues and limitations before they do put any real money. But at a close examination you will see that predicting stock market is one of the hardest thing out there. Most of the traders actually lose money. Most of the C2 systems are actually very poor. So after 3 months of free evaluation, most system developers will quit and the remaining ones will start with many lessons learned, but this would mean at least 2 - 3 times less subscription fees from system developers.

It does not mean that you should stop using C2. If trading is a hobby (like mine) or if you want to get serious about it, then just putting some money into it would be the best thing.

But if you do not have the meanings or you live in a third world country with huge trading fees, then C2 might be just perfect for you.

If you are a system developer:

  • Don't jump in too soon, make sure your strategy does work. Back test it well. For example before posting results on C2, I posted results on twitter for 1-2 months. Only after that experiment was successful I joined C2.
  • Don't screw your subscribers, make sure the strategy scales for the number of subscribers you have. C2 treads a system as infinitely scale-able, which we know it is not true! Don't use securities with large spreads or low liquidity!


If you are a subscriber:

  • C2 is not a scam, but individual system developers could be scamming you. And C2 acts as a legit umbrella for them and it does enables them though complex rules (some business trade offs, some governmental rules) - read the forum to understand what are those corner cases where you could lose money but C2 lists it as a winner. 
  • Also another way C2 acts as an umbrella, although not obvious and not too often is though their slow infrastructure. For example my system does trades at open of market and at close of market. But C2 used to have a 2-15 minute lag at market open, so my trade would be executed later than what I asked. If the trade was much worst than what I wanted, I would fill a request to fix it. If the trade is actually better than what the developer wanted, do you think they will request a fix from C2? One day when I asked C2 to fix a trade which seemed for them that would actually decrease my performance, they asked me if I really want that!
  • Another way system developers can scam you is by creating (under multiple accounts) many strategies in the hope that one would just work. A sure way to do this is to have 2 strategies in the mirror (e.g. one bets IBM would go down in the next month, one would bet it would go up). One of them would be a winner, one not. It does not even have to be the same stock, they could use pairs of highly (anti-) correlated securities for that! To protect against scams:
    • Do make sure the systems you use do have about 2-3+ months at minimum
    • Look for high sharpe systems (it means it makes constant profit, and it does not have it's profit from a few trades)
    • Stay away of systems who have most of their profit from 2-3 trades.
    • Watch out for market trend, if the market was booming, any long systems should do money.
    • Protect yourself based on your risk profile (use limit orders and/or buy cheap options to protect against major loses)
  • Don't screw yourself! If you want to trade 10x a system, make sure you check first with the system developer! It might not scale that much.
C2 can successfully be used, but both system developers and subscribers have to be alert! C2 acts as a place to meet, and guaranties that one's strategy maybe could work. Nothing more! It is up to you to make sure you don't screw each other.

C2 feeds on system developers aspirations and subscribers greed, all  under "government regulations" umbrella while not building features that would truly make light into system quality and scale-ability nor subscriber quality, because they know it will hurt their bottom line!

    Friday, August 5, 2011

    Is Collective2.com an elaborated scam?

    First, the numbers: our system, Project Athena, went though a correction period and we had 5% drawndown. That was though to see real money losses for that amount. One of those bad days  I was so stressed that I was barely able to breath while I was coming home with the bus! But the good news is that now it is at 13% profit since inception, and my stress tolerance went up a lot!


    Started$10,000
    Buy Power$11,385
    Cash$11,385
    Equity$0
    Cumulative $$1,385
    Total System Equity$11,385
    Margined$0
    Open P/L$0 Refresh System Stats


    Sometimes the computer gives alarms - the stock market is going to crash tomorrow! While stock market goes through a correction now, daily predictions do not seem to correlate well with actual market performance overall. But top ranked predictions are good - as seen by results.

    Second, we plan to move from Colective2.com to Covestor.com


    • On Colective2.com the system developers has to pay a fee, while on Covestor.com the system developer does not pay a fee, but it is required to actually trade.
    • On Colective2.com developers are payed a fixed commission by subscriber, while on Covestor.com, you seem to be paid % of profit.
    • Also Collective2.com seems to look more and more like an elaborated scam. I recommend first reading this forum thread. Why I think it is a scam: because there is a quality bar for systems on C2 - we actually pay to post results here, and this makes all things different. One can use the collective intelligence of the systems to make money - and this should work since we pay a fee to be here. 


    C2 claims that they don't trade based on systems, nor they offer results or aggregate of results to third parties.  But if I would be a unethical hedge fund manager, creating a web site where people would actually pay money to post systems, would be a fantastic way to make money. Businesses don't have to be ethical, just legal!

    Even it is not so, the way C2 treads subscribers vs. system developers makes the system to act like a well through scam:

    • system developers pay a fee to be on the system an post results, that makes for good quality signals than what you will find somewhere else, 
    • system developers in the beginning have to offer systems for free or extremely cheap
    • system signal became public to everyone, making every system available for reverse engineering to anyone, even if they are not registered on the site, or if they are not subscribers
    • subscribers can cancel a free, or pay for profit period before the completion
    • there is no way of knowing the quality of a subscriber and C2 does not have any features help with that (I want to know how many times a subscriber leaved a system after a certain drawn down, of before the end of a free period, or before the end of a profitable "pay for profitable" period.


    Now, if C2 is not a scam, then it is really a poor business for what a system developer is paying. I am a software engineer, and seems inadmissible to me for a 7+ year old software to have so many basic problems and so many asks and suggestions for improvement from developers, and all seem to be ignored. Just read C2's forum.

    I might be wrong on both accounts, but definitively, I will move to Covestor.com as soon as possible, it makes more business sense.

    Tuesday, July 26, 2011

    Options

    The last week was not great. Up until Wednesday (7/20) we were having minor positive gains (1%).

    But for Thursday(7/21) the prediction were looking weird. For IT and Consumer sectors, predictions looked regular. But for Finance sector they were extremely negative. Out of 48 stocks in finance, the computer recommended 38 shorts, and top 14 recommendations were strong shorts. This is in the context in which normally it recommends about 20 shorts normally. The only time in recent history when the program had similar negative predictions was for 6/1 when the market fall 2%.

    First we started to read the news to see what is happening, and we learned that in the next day or so, Europe was set to save Greece from default, or maybe not. And we learned also that the most exposed sector is Finance. That was great news for us, it meant the computer is not that stupid after all.

    So I bought options (puts) too. Lots of them. Fortunately for humankind - but unfortunately for my bank account - Europe did produced a plan that delighted the investors (at least for now).

    My options are still active, and there is still hope for some return on investment, because the predictions for tomorrow (7/26) look gloomy again (38 shorts out of 48, the only quality long signal in Financial sector is MCO at rank 5, the rest of top 15 predictions are shorts).

    Probably the computer senses that if Obama and GOP do not come up with some good debt plan, markets will penalize them heavily tomorrow (based on signal between 3 and 5%). But there could be a nice bump if Government comes up with a nice plan tomorrow. So I will buy a few calls too, (but not much), since I have invested plenty in puts.

    On C2 we are now at +8.5%, from our high of +11%

    Tuesday, July 12, 2011

    In machine we trust

    Machine is smarted than us. PERIOD.

    The end of the last week was bumpy. And it was not because of the economy nor news. Somehow the machine predicted a certain pattern for a given stock that we were not seeing, and it was supposed to be a very strong signal. But for some reason what the machine predicted didn't happen. The stock was supposed to fall, but it stayed flat for 2 days. The economic news should have shaken the stock. But nothing happen.
    The real losses stated to add up and after 2 days I could not bear the pressure and I swapped manually that stock with some another ones, predicted to fall also. A few minutes after I closed the position, a massive sell off followed. We lost because of that a lot, but the other predictions were not that bad either, and we broke now 10%+ on C2, making $1,098 in less than a month.

    When you trade with your own money, that is a totally new game! But we have no excuse, we backtested the machine and we should trust it. So lesson learned: "In machine we trust" from now on!


    Started$10,000
    Buy Power$11,098
    Cash$1
    Equity$1
    Cumulative $$1,098
    Total System Equity$11,098
    Margined$1
    Open P/L$0

    Tuesday, July 5, 2011

    Side effects of no subscribers on C2

    We started to trade with real money a week ago. Our profit margin is really small and it is easy to be eaten by bid-ask spread. Since any C2 subscriber can trade as much as s/he wants without any regard for one's actual long term profit, and because they could share the signals with friends, we removed all subscribers. We will sell our strategy (if ever) to only one subscriber, that is the only way we can ensure long term success.

    And we found what is the most important thing for a great C2 score: as soon as we removed subscribers, our score went form 900's to 300's. So C2's score can be manipulated in this way: get a 'friend' to subscribe, with a very low price ($5/quarter), and 90 free days (your credit card won't be charged for 90days) and see how your score jumps.

    0 subscribers has other side effects on your system, C2 does not execute your trades in time, we started to see again huge delays in execution, and we lost time pinging C2 to fix them (we don't even bother pinging them for delays longer than 5 minutes anyway). C2 is not suitable for daytrading, at least when you don't have subscribers because even 1 minute delays can make a huge difference.

    We witnesses differences of $20 per day between real daytrading on IB, and what C2's reports. And I mean IB is actual better than what C2 reports. Our sample is small, but consistent every day.

    The last side effect of not having subscribers, is that C2 is slow even for updating the system description, which is still pending for more than a week.

    Lack of mobile app is another issue, it is a pain to update manually C2 on your phone, so a couple of times we have not updated C2 signals in time, and while we did had good results on IB, C2's signals were not updated in time and we lost in stats.

    Saturday, June 25, 2011

    Too many fees and a new 'subscriber', guess who!

    The last two days of the week were kind of neutral. No major losses nor gains. But we have noticed the algorithm selected too many stocks which were too cheap, and if I counted correctly we paid $114 in fees. That is too much, we can reduce fees to about $4/day if we trade more expensive stocks. Since we have not seen signifiant difference in results when removing all 'cheap' stocks, we will trade only stocks whose value is above $40, effective Monday. That would decrease our weekly fees to about $20/week.

    And we have a new subscriber, guess who! Yes, we will dogfood our system using the recommended amount, putting our money where our mouth is! Our first day of trading was Friday, and we lost exactly $1. Does not match the exact results displayed on C2 because we manually trade. We wanted to use auto trade, but C2's support for IB (Interactive Brokers) is poor, requiring use of TraderStation or Trader68 to constantly run on your computer, which kind of sucks. But since IB just released their app for Android, now we can trade easily with our phones.

    And here are our latest results.


    6/24/11 9:30SELL800AMD6.986/24 15:556.89Low$56
    6/24/11 9:30BUY100A49.236/24 15:5548.47Low($78)
    6/23/11 9:30BUY100A48.206/23 15:5549.35Low$113
    6/23/11 9:30SELL700AMD6.926/23 15:557.07Low($118)

    Thursday, June 23, 2011

    First day without gains

    Yesterday, our first subscriber used auto trading, set for 50% of the recommended amount ($10k), and he had $30 profit. The bad part: today we had 4 subscribers who run $50k, and they lost $500 :(. When I have seen how much they lost, I felt like a punch in the stomach, so naturally I eat 2 chocolate bars to help with the guilt.


    Opened ETB/S#SymbolPriceClosedPriceRiskP/L
    6/22/11 9:30BUY100A49.566/22 15:5548.95Low($63)
    6/22/11 9:30SELL600MU8.156/22 15:558.20Low($43)


    Our subscribers pockets have been emptied a little. It was like a cold shower to remind us of real losses in trading. <sarcasm>But hey, our stats are doing great, our score was bumped to 935 and our system finally shows in the grid!</sarcasm>.

    And some random stuff worth noting:

    • Check this, we did ... $666 in less than 6 days. We should rename our system "The beast"!

    Model Account Status
    Started$10,000
    Buy Power$10,666
    Cash$10,666
    Equity$0
    Cumulative $$666
    Total System Equity$10,666
    Margined$0
    Open P/L$0




    • C2 fixed our first day stats, they do look into tickets and seem to be reasonable responsive.




    • Our backtesting is 30% complete. We will produce stats soon based on hypothetical results of the last 48 months.