Blog about my experience on C2

C2
1

Hi,



I started a blog about my experience on C2, you can find it here: http://day-trading-project-athena.blogspot.com



Hope it will be useful for others too.

2

Please post backtest statistics and equity curve from year 2007 to present.

3

not that simple because of technical issues, I have this equity curve from a (random) predictor, but it is not the best we have. It is for 1 LONG + 1 SHORT / day strategy, for the last 48 months



What kind of data do you want from backtest? Any documentation?



Notice: does not include fees!



1.8 means that if you started last N month with 10000 dollars, now you would have 18000 (the gain is 8000)



1.08061 - money made in last 1 month

1.08001 - money made in last 2 months

1.07546 - money made in last 3 months

1.14055 - and so on …

1.15527

1.2203

1.19259

1.21911

1.21366

1.31874

1.26752

1.34472

1.3758

1.50591

1.48463

1.50095

1.45421

1.53596

1.49138

1.40638

1.40881

1.32997

1.43463

1.54308

1.65457

1.71232

1.61832

1.63636

1.5589

1.46293

1.76139

2.06064

1.91752

2.24234

2.35693

2.62586

2.93077

2.80481

2.46883

2.48439

2.68081

2.73695

2.76747

2.9783

2.98288

3.16142

3.17029

3.32154

4

I will post more once the data is ready. thanks for suggestion!

5

Thank you for the data. I charted it on Excel and noticed it performed well though the 2008 crash. How probable is it?



I am interested in the statistics shown on C2 strategy pages, such as: #trades, %profitable, AvgWin, AvgLoss, etc. The purpose is to baseline expectations going forward.



The 48 month equity data shown has a max drawdown of about 15%. Is that a design objective with some probability of occurance?



Can you describe the market characteristics that the strategy performs best in.

6

"Thank you for the data. I charted it on Excel and noticed it performed well though the 2008 crash. How probable is it? "



When the strategy used is 1 SHORT + 1 LONG / day, does it really matter if the market is crashing or booming? IMHO the direction of the market matter less for this strategy, as long as you can indeed short the stocks, and you assume the unlimited risks of shorting.



Regarding dragdown: the worst I have seen in backtesting was 50% loss in a year. Dragdown was never a factor in algorithm (I stopped working a few years ago at the algorithm, but I don’t remember to ever consider it). I only cared about two things: in the long term the algorithm does well when I run 100+ simulations, and to perform “equally” regardless of where the market goes, even if it might not beat S&P500, that is why I enforced the algorithm to do N SHORTs and N LONGs every day (but we do get better results in backtesting if we remove this enforcement and let the algorithm to decide how many SHORTs and how many LONGs to do - even in 2008, but you will need to take my word for it).