Check out the TrendSensor Blog ....

 MarketBrief

The Best Trading
Systems On 
Collective2

  

 

This update: Jan 22, 2008

View Archive

Like beauty, "the best" of anything is very much in the eye of the beholder.  For this analysis I have tried to approach the process from the perspective of a fund manager looking for a specialist trader to trade a portion of the managers fund, or an experienced, analytical broker looking for Collective2 systems to recommend to clients.

So good, consistent results with excellent risk management is the ideal, and while I might add systems under 6 months old to my "watchlist", I'm mainly interested in the more mature, proven systems and traders.

I'll get into the full details of criteria and scoring later (towards the bottom of this page).  At this stage let me just summarize the process and results:

  1. On Jan 21, 2008, I ran a query on the new version of "The Grid" at Collective2, applying filters where I could for the criteria I have.  I excluded options systems since I don't feel adequately qualified or experienced in this area to assess them.  So if you're interested in trading options, this analysis won't help.
  2. This generated a list of 40 systems, which I copied into Excel for further analysis, which included applying further criteria and calculating "The Muzza Score".
  3. In the table below the eight systems highlighted with a green background at the top are the only ones that meet all criteria, out of the 40 considered. The table shows the top 20.
  4. The two systems with their Age column entry highlighted in yellow (Team Aphid Bird and Gold Swing Survivor Futures ptf $25000) are the only two "mature" systems that meet all criteria.
  5. Have Fun is a mature system (over 180 days old) that only fails to be on the final list due to "Market Experience" at 90% (up from 77% in the last update), and would be of particular interest to our fund manager or broker.  Have Fun would be of particular interest as it has a good Muzza Score.  Have Fun's creator only trades the best opportunities and is comfortable being out of the market until one comes along.  This is a sign of a good trader, but means it takes time to demonstrate significant market experience within the C2 environment.  Have Fun will qualify under the "Market Experience" criteria at the next update, but I already view it as being an "honorary member" of the "best list".

Changes since the last update:

Method:

  1. I have changed slightly the way I filter for, and weight, APD (average profit per dollar of drawdown).  The effect of this is:
    • Increase the chances of short-term systems making it to the "best list",
    • Increase slightly the weighting of APD in the overall Muzza Score, and
    • Increase Muzza Score values overall (meaning the values this time can't be compared with the Nov 2007 ones, but the rankings can).

  2. The C2 rankings I include for comparison purposes are now based on the following page: "System Finder" (in left side navigation)/"Very Conservative" (in Step 1). Also, exclude options systems in Step 2 (uncheck the box) and check all boxes in Step 3.  This appears to give the closest match between Collective2 rankings and my own analysis - at least for the top few systems.
  3. Collective2 has introduced a new version of "The Grid" which has resulted in a change in process, but this shouldn't have materially affected the results.
  4. I've introduced the "Stars Of Collective2" logo:   N
    To qualify for this logo, a system must have been on at least two consecutive "best lists" (the actual number of times is shown by the number "N", to the right of the logo). It must also be a mature system on Collective2 (at least 180 days old). 

    To retain the logo a Star system can only fail to meet a criteria for a maximum of one update.  If the failed criteria is "maximum drawdown under 20%", then it ceases to be a Star immediately as the system creator cannot subsequently reduce the maximum drawdown experienced.  Team Aphid Bird currently fails to meet the "% winning trades at least 40%" criteria (it's at 38.5%), and will need to lift back up above 40% before the next update (due late March 2008) to remain a Star Of Collective2.

Results:

  1. Since November 2007, the number of systems meeting all the criteria for being on the Best Systems list has grown from five to eight.
  2. Of the five qualifying systems in November, two have dropped of the list: DataGlobe FX II, and Holly FTSE.  Meanwhile three systems that didn't qualify in November are now on the Best Systems list:  TrendSensor Golden Index (from rank 10 to rank 3), TrendSensor JUMBO (from rank 9 to 5), and GoldenStocks Short Only (rank 6 on both occasions). In addition, two systems that are new to the list are now qualifying ones:  EXLENCE - that has made its entry at number 2, and Lilac (entering at rank 8).

  3. Team Aphid Bird and Gold Swing Survivor Futures ptf $25000 are now rated as Stars Of Collective2, and are the first systems in this new "best-of-the-best" category.  Jubilee has also been on the Best List both times and may well become the next Star of Collective2 at the next update, due in late March 2008.  Finally, Have Fun continues to be an impressive system that merits a special mention and honorary inclusion in the Best Systems list.



Summary: As a broker or fund manager, here are the "mature" systems (at least 180 days on C2) I would be interested in:

System:

Rated a Star Of Collective2?

Team Aphid Bird

Important Update: Feb 2008
Team Aphid no longer trade the Team Aphid Bird system.  It remains to be seen whether its replacement (Team Aphid Daily) also becomes a star of Collective2.

 2

Gold Swing Survivor Futures ptf $25000

 2

Have Fun

   No

   

And here are the newer systems I'd add to my "watchlist":

You can compare these systems on one page, at my Analyst Page at Collective2.

Or, just view the three trading systems I personally recommend .... 

Updates:  I'll aim to update this analysis on a regular basis and include a comment here on new additions to the list of best systems, and deletions from the list.  The next update is planned for late Mar 2008.

The Best Of Collective2 (Excludes Options systems): Based On Data Extracted Jan 21, 2008.

Muzza Rank Rank Change C2 Rank System Stars Of C2? Uses Stop-loss? Age (Tgt>=60) Mature=>=180 No. of Trades Avg Trade (Days) Market Experience (Days) Mkt Exp Weight (Tgt>=100) Win% Profit Factor Weight (Tgt>=100) Ann. Return% Max Draw Down Ann Ret % : Max DD Weight (Tgt>=25%) Sharpe Ratio Realism Factor (Tgt>=80) APD Factor
(Tgt >= 100%)		 The Muzza Score
1 + 3 1

Team Aphid Bird

Yes Yes 361 338 0.896 303 300% 38.50% 107% 52.0% 4.80% 54.17% 3.607 361% 400% 19.01
2 New 2

EXLENCE

  Yes 119 60 6.000 360 300% 53.30% 207% 211.0% 8.60% 122.67% 6.641 85% 400% 14.71
3 + 7 3

TrendSensor Golden Index

  Yes 107 28 3.000 84 120% 57.10% 320% 123.2% 5.00% 123.20% 3.843 93% 400% 13.21
4 - 2 5

Gold Swing Survivor Futures Ptf $25000

Yes Yes 186 87 3.300 287 300% 51.70% 187% 191.4% 10.10% 94.75% 3.961 88% 327% 7.55
5 + 4 6

TrendSensor
>> J U M B O

  Yes 107 65 3.000 195 279% 49.20% 187% 134.1% 7.40% 90.61% 4.294 93% 266% 7.35
6 0 -

GoldenStocks Short Only

     84 183 9.800 1793 300% 69.40% 167% 70.7% 7.10% 49.79% 4.312 85% 301% 4.10
7 - 6 -

Jubilee

   Yes 139 113 1.400 158 226% 53.80% 133% 98.5% 9.80% 50.26% 2.998 95% 222% 1.64
8 New -

Lilac

   Yes 105 62 2.300 143 204% 40.30% 113% 122.8% 13.80% 44.49% 2.253 100% 137% 0.44
9 + 9 -

Have Fun

  Yes 277 35 1.800 63 90% 82.90% 313% 54.2% 3.50% 77.43% 3.181 99% 359% 6.38
10 New 4 MAIN QQQQ    ? 61 7 4.700 33 47% 85.70% 293% 214.5% 10.80% 99.31% 4.784 100% 400% 2.52
11 + 16 - Zombie Breakout    No 195 114 7.700 878 300% 72.80% 127% 70.7% 10.50% 33.67% 2.85 96% 155% 0.91
12 + 11 - Scalper FOREX      87 81 0.054 4 6% 86.60% 500% 71.9% 2.20% 163.41% 8.827 100% 179% 0.81
13 New - Gold Survivor Energy Portfolio      492 571 0.092 52 75% 46.40% 107% 74.0% 6.20% 59.68% 3.524 99% 124% 0.40
14 New - Gold Survivor DayTrader      487 677 0.100 68 97% 46.50% 100% 79.2% 9.00% 44.00% 2.657 100% 84% 0.20
15 New - Blackbird II      148 107 0.613 66 94% 34.60% 100% 38.7% 6.30% 30.71% 2.392 90% 78% 0.14
16 New - Leveraged Russell 2000      90 25 1.700 43 61% 52.00% 113% 131.5% 13.50% 48.70% 1.933 85% 149% 0.14
17 New - Turning Points      435 149 4.400 656 300% 53.70% 100% 90.8% 16.70% 27.19% 1.911 83% 93% 0.10
18 New - SPY 1 Trade Per Day      130 49 0.133 7 9% 77.60% 140% 35.8% 4.40% 40.68% 2.993 100% 145% 0.06
19 New - Gold Survivor $10000 Portfolio      186 196 0.075 15 21% 46.40% 100% 202.8% 15.70% 64.59% 3.114 96% 90% 0.03
20 New - Sleeper Cell - 1st Phase      93 41 0.546 22 32% 51.20% 107% 84.9% 16.60% 25.57% 2.057 84% 106% 0.01

Criteria:

  1. Max Drawdown not more than 20%
  2. Realism Factor of 80% or higher
  3. Age on C2 of 60 days or higher (180 days or higher to qualify as a "mature" system)
  4. APD (average profit per dollar of drawdown, after adjustment) of 0.30 or higher
  5. Profit Factor (total dollars gained to total dollars lost to-date) of 1.5 or higher
  6. Market Experience of 70 days or more (number of trades x average trade duration, expressed in days)
  7. Annual Return (%) : Max Drawdown ratio of 5:1 or higher
  8. Percent of winning trades of 40% or higher
  9. Uses automated stop-losses that are set at trade entry (this is checked with the system vendors of systems that meet all the above criteria).

Calculating The Muzza Score:

The Muzza Score = [Mkt Exp Weight]   x   [APD Factor]   x   [Profit Factor]   x   [Ann Ret : Max DD Weight]   x   [Max DD Factor]   x   [Realism Factor]

NOTES:

  1. All factors have maximum scores set for them to prevent undue influence from "outlier" results.  These range from 100% to 800% - see below for more details for each factor.

  2. Mkt Exp Weight:  Market Experience data doesn't exist at Collective2, but I have calculated it using other C2 data, specifically: average trade duration (in days) and number of trades on C2 for this system.  Market Experience is a very important concept as it gives the total "days in the market" for a trading system (number of trades x average trade duration, expressed in days).  Let's consider an example to clarify this:  a system that has traded for a year but has only completed three trades may seem impossible to assess.  But if those trades have averaged 30 days each, then the trader has had 90 days in the market, during which he/she can botch things up, or excel.

    On each of those 90 days the trader is reviewing their trades and answering these questions:
    • Do I close the trade?
    • Do I top-up the trade?
    • Do I close some of the trade to reduce its size?
    • Do I leave the trade as it is?

      So, although there are only 3 trades, there are 90 days of market experience, which meets the minimum criteria of 70 days.  The weighting process sets 70 days at 100% and sets an upper cap at 300% (or 210 days).  At this point there is ample data to be confident in the performance analysis, and additional market experience adds little in terms of additional certainty.

  3. APD Factor: APD is included as a statistic in C2, and is a key performance measure.  I make an important adjustment I'll get to in a minute.  Here's how I think of APD (Average profit per dollar of drawdown):  It's like an ROI figure for venture capitalists investing in businesses.  They will look for returns of 35 to 50% or more to justify their high risk in investing in relatively unproven businesses.  Trading is also a risky business and we should aim to achieve 0.35 to 0.50 or more, with 0.30 as a minimum.

    As a venture capitalist I might accept a lower return (say 30-40%) from a proven management team that makes 80-90% great decisions, than from a company with an average, or unproven, management team.  This leads to the simple adjustment I make to APD: systems with a % of winning trades over 50% get their APDs adjusted up proportionately, while those with less than 50% winning trades get adjusted down.  Why?  Because a system with 90% winning trades can afford to make less gain per dollar of risk (and most likely will) and still achieve a great outcome.

    An APD of 0.3 converts to 100% after applying the % of winning trades adjustment.  I view the APD Factor as the key measure of performance in the Muzza Score calculation. The cap is set at 400% (equating to an adjusted APD of 1.2 or higher).

  4. Profit Factor = total dollars gained / total dollars lost.  A score of 1.5:1 is set at 100% and a cap is set at 500% (equates to a score of 7.5:1).

  5. Ann Ret % : Max DD Weight:  A ratio of 20:1 is set at 100% and the minimum of 5:1 equates to a 25% score or higher. The cap is 500% (equating to a ratio of 100:1).  This is also a key measure as a good score means high Annual Return with a low maximum drawdown, which is a very attractive mix.

  6. Max DD Factor: A maximum drawdown of 20% or higher is set at a 0% factor score. A maximum drawdown of 0% would achieve a 100% factor score. Other points are scaled between these two - eg: a max DD of 10% would get a 50% factor score as 10% is midway between 20% and 0%. The cap is 100%.

  7. Realism Factor:  this ranges from 0% to 100% and is taken as-is from Collective2 and used in the Muzza Score calculation.  A figure of 80% or higher is required to meet the criteria for realism.

  8. Maximum possible Muzza Score = 3 x 8 x 5 x 5 x 1 x 1 = 600, but in reality no system is likely to get close to a perfect score.

 

For Collective2 System Vendors: For those who have their systems highlighted with a link - CONGRATULATIONS! With my "fund managers hat" on I can honestly say I believe you're doing an awesome job!  Keep up the great work. You're the vendors who give Collective2 its real value.

For Investors: In my view, the highlighted systems have traded extremely well.  But what they have done, and what they will do in the future can be two very different things. I cannot give any guarantees and nor could the vendors themselves.  You need to make your own decisions based on your own research and professional advice received.  And always feel free to ask vendors at Collective2 your most difficult or important questions - just use the Private Message facility at Collective2.  Please also read the full disclaimer at the bottom of this page, and the one at Collective2.

You can view the systems highlighted on this page, altogether for ease of comparison, at my Analyst Page at Collective2.

Or, just view the three trading systems I personally recommend .... 

TrendSensor.com is a client of Collective2, but does not represent Collective2. Specifically, the analysis on this page is the work of Murray Nickel of TrendSensor.com and does not represent the views of Collective2 or its representatives.

Disclosure: I am the vendor and trader for the two TrendSensor systems at positions 3 and 5 in this list.