As a regular user of Collective2, I thought it would be fitting to begin my series of reviews here. It’s difficult to start a review series in a void, so I will be comparing aspects of Collective2 (or C2, as it is often abbreviated) to some of the other sites listed in The Autotrading Universe; but don’t worry–I intend to review, eventually, as many of those sites as possible to provide further context. N.B. this review will be updated as Collective2 continues to evolve.
What is Collective2?
To quote from the site itself:
Collective2 lets you choose from a library of over 20,080 trading systems, and then lets you automatically trade those systems in your brokerage account.
Let’s parse that a bit. As previously discussed, a trading system is a set of rules by which trades are made, and can be made either manually by a person, or automated through a set of algorithms and executed by machine. I’m a bit curious as to where the count of 20,800 comes from, as my own searches through C2’s “Grid” show a total of between 700 and 900 systems, depending on the time of year.
Who creates these trading systems?
Details are scarce. According to C2’s Frequently Asked Questions, system developers do not need to register with any regulatory body to publish trading signals through Collective2, and thus the potential subscriber to a system depends on the system developer to reveal information about himself, either in his profile, or by sending the developer a message through Collective2’s messaging system. If the developer has chosen to fill out the profile page, it will typically look like this, from the creator of the Topaz NQ 100M system:
(I chose Dr. Koch for this example because he is the subject of C2’s case study).
In any case, caveat emptor. Unless the developer is a financial professional registered with a securities regulator, it’s difficult to know what the developer’s background or qualifications are.
How can these system automatically be traded?
Collective2 uses technology provided by a company called Trade Integration to enable its customers to autotrade their accounts. Users subscribe to a system, then set up autotrading for that system by providing Collective2 with their login credentials to a partner broker. Collective2’s autotrade setup allows the user to determine how to scale their own allocation vs. the model account (e.g. trade at 25% of the position size that the model account uses), set a limit on the size of each trade, and override positions opened by the system by implementing custom stops, limits, and closing orders. (Screenshots to be posted at a later date).
How Much Can Be Earned?
Here’s what Collective2 says on the matter:
If your system performs well at C2, you will make money. How much? That obviously depends on your system, but we can give you some benchmarks. The highest-grossing system on C2 has earned over a half-million dollars. Other systems approach the $100,000 dollars in revenue level. And quite a few good systems cluster around the $20,000 to $30,000 level. While these are the “stars” of course, many systems on the C2 Platform earn money for their developers. In fact, 26% of C2 systems that are over one month old and have positive trading performance have earned at least some subscriber revenue. Almost half of these have earned over $1,000.
A few caveats. In my years of using Collective2, first as a subscriber, then as a vendor, I’ve noticed that there is a clear dichotomy in earnings. Several high-flying (and ultra-high-risk) futures and forex systems come out of the gate with mind-blowing returns in the first months, and naive subscribers flock to those systems in anticipation of lottery-like gains. These systems almost always burn out in their first six months, but generate quite a bit of revenue. The other type of system generates slow but steady returns, but often takes years to acquire enough subscribers to earn significant revenue. If your system returns less than 30% CAGR, your system will be ignored. Stock systems are also far less popular than futures and forex systems, and will consequently generate less revenue. Also, please remember that Collective2 takes a 30% commission on all revenue that your system generates, on top of the semi-annual $98 listing fee charged to vendors per system.
Trades Own System
Collective2 implemented the “Trades Own System” (TOS) program in early 2012. Quoting from Collective2’s description:
The Trades-Own-System (TOS) Program is a special program established by Collective2 which allows trading-system creators to “put their money where their mouth is.” To achieve TOS Certification, a system creator must trade his own system using the same exact AutoTrade technology that subscribers use. Because we can monitor trades made by the system creator using AutoTrade technology, we can verify that the system developer actually trades his own system, and at what quantity.
In order to achieve and maintain TOS Certification, a system creator must follow a majority of his own trades in a live brokerage account. By multiplying the “scaling percentage” used by the system developer by the the size of the Collective2 Model Account he runs, we can calculate the Nominal Dollars at Risk by the vendor.
TOS was implemented in response to the concerns often expressed by users on sites that provide signals: Do the system vendors have skin in the game? What incentive do system vendors have to manage money carefully, instead of shooting for the moon, achieving huge returns and drawing in users, only to collapse thereafter? TOS is meant to provide added assurance to users that a system vendor is truly invested in the success of his system, proven by his trading of the system using actual dollars. TOS doesn’t appear to be particular popular, and users do not seem to reward vendors disproportionately for acquiring TOS status. That said, most TOS systems are unprofitable, so the jury’s out as to whether it will provide added value or even survive in the long term. (See the “My Misgivings” section below to see some updates).
Collective2 is testing a feature called System Constraints. I’m only mentioning this, and not reviewing this feature, as I have not implemented it myself. It seems like a good idea, but like the Trades Own System qualification, I think that providing an incentive for system developers to use this will spur it along. That said, the incentives for TOS were withdrawn after one year… I’m going to just quote the explanation below:
Beta Test Notice: The System Constraints feature is currently undergoing beta testing. During this period, despite what is written below, we will allow system developers to change or remove their system constraints any time they like, and changes will become effective immediately. Also, during this period, we will not publicly reveal on your system page which system constraints, if any, are in place. The purpose of this testing period is to make sure that C2 correctly evaluates and handles your system constraints. Please be aware this feature is still in beta.
Set Trading Constraints for [Your System]
You can increase a potential subscriber’s comfort with your trading system by agreeing up front to follow certain trading constraints. Once you specify trading constraints, subscribers know you will be required to follow them. C2 will prevent orders from being placed which would violate the constraints, and/or may close any positions that violate constraints.
If any constraints are violated, C2 will close any position its software deems fit; it may not close the position you would have wished to close. Remember, the idea here is to avoid ever reaching the point where this happens.
Can I change my mind?
Yes, but it is a bit of a process.
Once you enter constraints below, and once you have a subscriber, you can only change your constraints in the following way:
- You may enter a Constraint Change Request, which will be sent to all subscribers for approval.
- Once all subscribers approve of the change, or once 7 days pass — whichever comes first — your changes will become effective.
Remember, the purpose of this feature is to allow conservative system developers to promise that their strategy will be run in a conservative fashion. This will attract subscribers who seek such strategies. This is not appropriate for all systems, system creators, or trading styles.
- No single position may take more than % of system capital (Suggestion: 7%)
- Sum of all open positions may take no more than % of system capital (Suggestion: 30%)
- Trades must have a stop loss capping loss at this % of system equity (C2 will create one if you don’t) (Suggestion: 3%)
Collective2 has supposedly “soft-launched” (i.e. without any announcement) an international version of its site that will allow futures and forex system developers to be paid per trade (0.5 pips per round trade for forex, $0.50 per futures contract bought or sold). At least for forex, this is in line with its competitors. This payment scheme is only in force when non-USA subscribers trade the system; American citizens will continue to pay the flat monthly fee set up by the developer; indeed, only non-US citizens are even able to view the site. After a test period, system developers will need to pay a separate listing fee to offer their systems on this international site. As a US citizen, I am unable to evaluate the international site.
The proprietor of the site, Matthew Klein, is a published author (Con Ed and Switchback) and has a quirky and self-deprecating sense of humor. Occasionally he sends out a newsletter to C2 users reviewing popular systems on his site (see here, here, and here), interviewing popular system developers (see here and here), or advice on how to choose a trading system (see here, here, and here), all representative of his jocular writing style).
I began using Collective2 in 2009 as a subscriber. I subscribed to several stock, forex, and futures systems and ended up losing money for a variety of reasons, which are common to subscribers:
- The system didn’t have enough of a track record for me to accurately gauge its risk. It’s important to see how a system recovers from adversity before subscribing. See if the system can recover from drawdowns, and how long it takes. Wait to see how the system performs in bull and bear markets.
- The system developer engage in martingale/pyramid trading (in which the system continues to add against a losing position in an attempt to average down and benefit from a reversal in the system’s favor). Unfortunately for me, this behavior only began after I subscribed to the system.
- The vendor did not respond to questions in a reasonable period of time, which was a warning sign. Several vendors on Collective2 have been known to simply disappear, leaving their subscribers hanging with open (and likely losing) positions.
- Several users on Collective2 engage in unethical behavior by starting a system with an account, and then if that system fails, abandoning that account and starting a new system on a new account. In this way, these users are able to conceal their track record.
- The Collective2 scoring system is opaque and does not correlate well with system performance. It tends to reward age disproportionately when rating systems. A user can always look at The Grid and select custom criteria to choose systems, but with so many different types of systems, Collective2 would add value by providing a guideline superior to its current C2 Score scheme. Alternatively, C2 should provide downloadable spreadsheets of all available systems with their performance statistics, and let users calculate their own scores.
My Misgivings About Collective2
While I have been using Collective2 both as a subscriber and a developer for four years, I have mixed feelings about the site and have begun looking for alternatives. Collective2 has several strikes against it:
- Layout. Users often post in the forums that they are new to the site and need help understanding what Collective2 is, how to use it, and how to select systems. This is a serious problem, and I often wonder how much more business Collective2 would have if it focused on improving its user-friendliness. Overall, I find the site difficult to navigate, and it has several idiosyncrasies which add to the problem. For example, the administrators thought it would be helpful to add a tool called System Guru to help analyze a user’s risk appetite and recommend systems that matched the user’s risk profile. This tool can be accessed from the home page and nowhere else, not even appearing under the “Find a System” column of the navigation menu. Other minor inconsistencies include the fact that the search box appears on the right side of the screen on the dashboard or system pages, but on the left side on the forum page. The forum system and messaging system are also rather primitive, thus inhibiting user/developer communication.
- Security. At the end of 2009, it was hacked and the credit card information was stolen. Not exactly confidence inspiring. Also, as part of the recovery, the current, new site layout was implemented, and I find it to be user-unfriendly and not welcoming for novice users.
- Choice of Brokers. One of my favorite brokerages, OptionsXpress, was removed as an autotrade broker. Since then, the only choices available to subscribers are MB Trading and Vision Financial, which do not appeal to me. Collective2 management promised for years to add Interactive Brokers to their autotrade program, but eventually was able to add it only to autotrade futures (although apparently non-US customers are able to autotrade all instruments through Interactive Brokers). Not only does such a paltry choice of brokers restrict my ability to trade through Collective2, it also limits the ability of Collective2 to attract more users. Without a business relationship with at least one of the big brokers, like TD Ameritrade or Schwab, Collective2 will always be a niche service.
- Lack of Long Term Planning. Meanwhile, starting in 2012, Collective2 began the “Trades Own System” (TOS) program mentioned above. At the time it was introduced, it included several significant incentives to join: listing fees were waived for developers enrolled in the program; Collective2 cut its commission on subscription revenue from 30% to 15%; and autotrading fees were waived. I created three forex systems and enrolled them in the TOS program, and used the anticipated saved listing fees to create three equity trading systems. On January 25, 2013, Collective2 announced that TOS would no longer waive listing fees, and its commission fee would revert to 30% of subscriber revenue.
- Customer Service. In a word, terrible. Emails to Collective2 are often ignored, trouble tickets are often ignored, and posts to Collective2’s help forum are often ignored. It’s a small team running the site, to be sure, but isn’t addressing the questions and concerns of one’s revenue-generating customers the most important part of any business?
- Stability. The API server often lags or crashes, which causes my systems to miss signals. As I’m trading my own systems with my own money, this isn’t an academic problem. What good is an algorithmic system that only executes some of its signals?
- Lack of Focus. Here’s what the administrators of Collective2 say they are working on, or promise to work on: improve the Collective2 scoring system; improving the Trades Own System qualification; improving the “automated system constraints” feature (not mentioned in my review because it’s never been formally introduced on C2, and Matthew Klein says he is not yet satisfied with it); establishing relationships with more brokers; rolling out its international forex site; improving its affiliate program, etc. It’s all important, to be sure, but I’ve heard these same promises since I joined in 2009. So why is C2 management wasting time on experiments like TradeFizz? Sometimes it feels like management has lost interest in the maintenance and improvement of Collective2.
- Business Strategy. Other social trading sites are hyperactive in their marketing, advertising, and interactions with the financial media. Management of Collective2 doesn’t even formally announce when they add new brokers or features, let alone issue press releases. Does C2 want to be a player in this industry, or is this just a hobby? Perhaps they should consider selling or merging with one of the other social trading sites.
Needless to say, it’s difficult to build a business with such frequent changes at Collective2, so I am looking at Zulutrade, Tradency and Currensee to provide my forex trading signals, and Covestor, DittoTrade, or various brokerage houses’ independent newsletter autotrading technology to provide my equity trading signals. Collective2’s biggest advantage is its web services API, which is fairly easy to use, and it’s difficult to find similar functionality elsewhere. I am slowly learning C# in order to transition to one of the several available software packages to automate trading (NinjaTrader, MultiCharts, ForexConnect, etc.)
Collective2 has been in operation for several years, and the site administrators have emphasized several times that the site is “financially successful,” so hopefully it will continue to operate well into the future. Even though I have highlighted several problems above, I have been a user of Collective2 for several years, and have not yet found a complete replacement for the services provided by C2. I want to emphasize that despite my often negative tone above, I want Collective2 to succeed. That said, time is not on Collective2’s side, as new competitors are constantly emerging (I seem to add to my Algorithmic Trading and Autotrading Universe page every week), and existing competitors continually add new features that degrade Collective2’s advantages. If you’re a retail investor at one of the major brokers, Collective2 will be of limited value to you because of the lack of autotrading through your existing accounts, and if you have the chops to develop your own automated trading solution, Collective2 will be redundant.
We will explore some of these other competitors in future reviews.
Update, June 5, 2013: After a year on Collective2, I have decided to discontinue my subscription to Collective2 for my foreign exchange system, Cold Fusion FX. I will instead concentrate on the other social trading sites where I offer this model (please see INTJ Algorithmic Trading Models for more details).
Update, June 18, 2013: Another questionable design update from Collective2. The new homepage at http://www.collective2.com gives me a bit of vertigo from the color choices. Is it just me? Compare to the previous version and even the version before. I really wish they would spend more of their time on the other issues I mentioned above rather than on cosmetic updates of questionable aesthetics.
Update, June 29, 2013: Collective2 has raised its system listing price from $98 every six months to $120 every six months. To reiterate, Collective2 now charges system developers $120 every six months, plus 30% of any subscription revenue. On the subscriber side, users are charged $100/mo to use autotrading technology for stocks ($1.99/contract for futures and $0.50/mini-lot for forex) on top of any monthly subscription fee they must pay to the system developer. Meanwhile, most of Collective2’s competitors don’t charge any listing fee to system developers, and subscription fees on those sites are either volume-based, or a fraction of what Collective2 developers charge (Forex site ZipSignals recently introduced a volume-based pricing model). We’ll see which philosophy wins out in the end: the “build it and they will come” philosophy of most of C2’s competitors, or “build it and charge them to come” that C2 is pursuing.
Update, July 26, 2013: Collective2 has implemented an attractive new dashboard, and mercifully returned to the previously aesthetically pleasing homepage. I hope this is an indication of good things to come, but the speed with which competitors like Myfxbook are implementing new functionality doesn’t bode well for Collective2’s glacial progress.