collective2

Quote from late apex:

Excellent question, OddTrader. I've looked into that a while ago. Actually, no trades have gone missing there. All equity curves and drawdown stats on C2 are derived on an Unrealized, i.e, continuously mark-to-market, intra-day basis. The "Closed Positions" past performance tables are just what the title says: Closed, or Realized basis. It's almost apples and oranges.

For many types of forex trading systems, employing wide stops (including Hawk-fx), a strong argument can be made that the latter (Realized / Closed) equity curves and DD stats are considerably more useful and informative. Unless, of course, you are (= one is, not you) a risk-averse institutional investor, evaluating a system's track record...
More...

The dates of some trades are hard to understand. (Actually I don't really understand the details of C2. However, I am interested only in the figures for benchmarking purpose.)

As we can see one trade dated 7/27 is between 7/29 and 7/29; 7/29 between 8/22 and 8/19; 8/22 between 8/24 and 8/23; and 7/27 between 10/12 and 8/24, the C2 software would probably allow editing of data and others. Perhaps some amazing results could be produced if a vendor works hard enough to edit data. :D
 
Quote from MAESTRO:

Collective?
More...

I think the vision of the project and its idea is good and even outstanding, basically. However, bugs (you know what I mean) and greed would easily kill the vision, I'm afraid.
 
Quote from MAESTRO:

Price vs Time automatic systems will never work on the long run. Illusion, religion, rituals etc. Dancing around the bone fire praying for rain. Collective? Another place for enjoyable masturbation. Bunch of kids ... 1 dollar store for trading systems. Would you like a candy with your purchase?
More...

LOL...So what's wrong with masturbating once in a while?? Eating candy can sometimes enhance the ejaculation. BTW, my little kid system is on there with a 265% return :D
 
Quote from OddTrader:

The dates of some trades are hard to understand. (Actually I don't really understand the C2 at all. However, I am interested only in the figures for benchmarking purpose.)

As we can see one trade dated 7/27 is between 7/29 and 7/29; 7/29 between 8/22 and 8/19; 8/22 between 8/24 and 8/23; and 7/27 between 10/12 and 8/24, the C2 software would probably allow editing of data and others. Perhaps some amazing results could be produced if a vendor works hard enough to edit data. :D
More...

In the case of Hawk-fx, if you look at the bottom of the "System Description" section, just above the track record, you'll see the following:

"Final Note: Due to some technical issues with the C2 site, we did not post any trades from September to early October 2005"

The technical issues were not unique to that vendor. That explains the gap in dates.

Also, I am positive that vendors have no way to edit their historical data at all. In fact, on their systems pages and on the forums they sometimes resort to pointing out various problems with their C2 data. Such as the order entry platform being down, issues with ITM (Instant Trade Messenger), no OCA orders, bad fills, no fills, treating break-evens as losses, problems with BulletTrade auto-trading, etc.

I think Collective2 is promising, especially with someone as innovative and dedicated as Matthew Klein at the helm, but is obviously still going through the early-growth stages.
 
Quote from late apex:

The technical issues were not unique to that vendor. That explains the gap in dates.
More...

Then how can we fairly evaluate the drawdown figures? :confused:
 
I put my system up on Collective2. It's a very long term system so I don't really expect any subscribers, but if I get some I'd be quite happy. I just wanted it to build a public track record.

The free trial is good for the first 5 trades. Since I go weeks or months without trading, that should last me a while. And if my results are bad during that period (which is unlikely), I just won't pay to continue.
 
Quote from OddTrader:

Then how can we fairly evaluate the drawdown figures? :confused:
More...

I am not sure that we can.

One alternative would be to ask the vendor of interest for a more complete track record and analyze it yourself, in Excel. Of course, that approach suffers from at least 2 drawbacks: it's time-consuming and it misses out on the whole notion of independent / third-party / real-time verification of each trade by C2.

On the plus side, this approach has at least 3 major advantages I can think of. First, it allows us to come up with "Realized basis" equity curve and drawdown stats, and other performance metrics, not shown on C2. (See either Thomas Stridsman's "Trading Systems That Work" or Lars Kestner's "Quantitative Trading Strategies" for some nice Excel templates.)

Second, once the raw trade data is in Excel, we can equalize each system of interest on C2 with respect to key decision variables such as risk, leverage / margin and compounding. All C2 system vendors start off with an identical $100,000 model account, but then have full control over "heat", or those variables, by specifying the size of each trade and its stop loss, if any. Simply put, they are all over the map, from pretty conservative to pedal-to-the-metal all the time... Without such equalization, comparing the expectancy and risk/reward profile of different systems and choosing "the best" (for a given system buyer) is going to be highly misleading, despite the seemingly plentiful statistics shown on C2.

Finally, you'd be in a position to apply acrary's framework -- yes! -- for optimally combining multiple positive-expectancy systems...
 
Quote from late apex:

I am not sure that we can.
More...

It appears that the data problem for entry date sequence during the same period cannot be found from another vendor's trade details history. :confused:

Why is that the vendor didn't request required corrections from C2 on a timely basis?
 
You must log in or register to reply here.
Back
Top