C2 stats?

I still think there is defiantly something wrong with C2 stats and charting.



Example:

http://www.collective2.com/cgi-perl/system38125396# (a futures system):

Using Typical Broker & Autotrade commission plan ($6.98/trade) C2 says this system lost 4.9% in Feb and lost 3.3% in Mar.

If you change to MB Trading ($2.95/trade) then C2 says this system lost 11.8% in Feb and now made a gain of 5.3% in Mar!



Try changing the commission plans and chart types and you will see the Monthly Returns seem to change randomly.



See: http://img232.imageshack.us/img232/9060/c2stats.png

Warren:



I cannot replicate the screen shots you are pointing to. When, on my screen, I compare MB Trading and "typical broker commissions" the MB Trading results are more favorable.



Try this: Shut your browser, re-launch your browser, and try again.



Matthew

I just signed in again and this time C2 is saying Feb was -3.9% and Mar was -2.7% with MB trading commissions. The system has not traded to make these changes.

See updated clip: http://img5.imageshack.us/img5/6526/c2stats2.png

I changed the graph to Personalized and un-checked the commissions.

Now Feb was -11% and Mar was +5.4%

See clip: http://img580.imageshack.us/img580/3846/c2stats3.png

Crazy.

Warren:



You did trade in February and March.



And the results with MB Trading commissions during those months are indeed -3.9% and -2.7%. That is completely correct.



“Personalized” results are different from this and depend entirely on the personalized parameters you entered. The “personalized” charts are just estimates and are calculated on a closed-trade basis only, not a marked-to-market basis. For systems that keep trades open for a long time, this can introduce error, so I suggest you focus on the broker-by-broker commission-included results (such as the Mb Trading results you question), which are entirely accurate.



If you need further detail about how these are calculated, hold your mouse over the right-most column in the trade list to see how each trade’s commission and fees are calculated.



Matthew

I meant that I hadn’t traded TODAY but the results still keep changing TODAY. All trades were closed last night. Trades were not open for a long time.

As I said, I UN-checked the commission and fees (commission and fees set to zero) in the personalized chart then Feb DECREASED to -11% and Mar INCREASED to +5.4%! Why?





I meant that I hadn’t traded TODAY but the results still keep changing TODAY. All trades were closed last night. Non of the trades were ‘open for a long time’.

As I said, I UN-checked the commission and fees (commission and fees set to zero) in the personalized chart then Feb DECREASED to minus11%! Why? With NO commissions or fees?

There is definitely some problem. You must see the same problems I have described and taken clips of for proof above. I don’t understand why you can’t see there is a problem. If the commissions are reduced to zero, why did the loss for Feb increase? Why did the Feb loss INCREASE from -4.9% to -11.8 when I changed to a LESS expensive futures broker (‘Typical’ to ‘MB’)?







Warren:



I’m going to decline going into detailed discussion about the Personalized Charts. Some people find them helpful; others don’t like them. Your mileage may vary. Other than that, I will say only that they’re kind of like interpreting dreams – very interesting to only one person in the room.



Regarding the stats: Your equity (before commission) looks like this:

20110228 $18,900

20110301 $17,015

20110302 $16,915

20110303 $17,375

20110304 $16,920

20110306 $17,500

20110307 $17,530

20110308 $18,490



The monthly percentages you see, Warren, are based on these numbers, plus or minus adjustments for brokerage commissions, which depend on the broker you choose.



Matthew

I checked the stats again just now and they all seem to be OK now - choosing a more expensive broker or increasing commissions increses losses and vica versa. And both Feb and Mar return % change in the same direction. I know there was no problem - but thanks for fixing it.