Example: Captura DTS System
http://www.collective2.com/cgi-perl/systems.mpl?want=publicdetails&systemid=26720715&session=424587118246917723519843181251427
Realism Factor = 93.9%
March Hypothetical Monthly Results = 0.3%
With these numbers, I would expect March to be near break-even. Nope:
Total March P/L = -$5511.
In fact, until March, the RF was 100% for this system, then it dropped sharply in March. Due to more realistic fills in previous months, the cumulative RF remains relatively high, which is deceiving, because the RF for March is clearly not high. I think a better way to calculate RF would be on a monthly basis, to show how ‘realistic’ the system is doing now.
Total March P/L = -$5511.
What are you looking at???
I added up the P/L for every March trade.
I would assume you also have to consider open trades at the start and the end of the month.
I have. There were no open trades at the end of February (last one closed on 2/27), and there were no open trades at the end of March (last one closed on 3/31). I added up the P/Ls for the thirteeen trades made in March, which is -$5511, yet C2’s “Hypothetical Monthly Results” says March was profitable at +0.3%. This discrepancy is rather troubling, as it calls into question the HMR of every system in C2 (yes, I have read their disclaimer, but come on). Again, the only explanation I can come up with is that the realism factor for this system tanked in March, and the actual fills were nowhere near the posted fills.
Max:
Your math is incorrect. What you are missing is that monthly results include both open and closed positions, calculated at a marked-to-market basis. For the system in question, there were several trades that were still open at the end of February which were subsequently closed in March.
Thus, adding together the P/Ls of only those trades that were both opened and closed within the single month of March will not give you full March results.
You need to get a hobby…
Now I understand why you are ignored by 78%. I’ll gladly increase that right now…
Mathew-
Thanks. It sounds like you’re saying that C2 calculates the HMR for a given month based on all positions opened (not necessarily closed) in that month. Is this definition correct?
Mathew-
Thanks. It sounds like you’re saying that C2 calculates the HMR for a given month based on all positions opened (not necessarily closed) in that month. Is this definition correct?
No. I am saying that monthly returns are calculated based on the change in value in the system’s overall equity, from the beginning of the month to the end of the month. The system’s overall equity includes closed positions plus open positions valued with marked-to-market accounting, as per GAAP.
Okay. I see HMR is more complex than I assumed.
Can you comment on Realism Factor? It seems to be cumulative. That is, it could be 100% for months; then, one month the system realism goes awry, and the factor still stays relatively high, because it is weighted toward the cumulative. If this is the case, it would be nice to see RF broken out by month, as an additional way to monitor abrupt changes.
Finally, where’s that cheap, reliable, all-instruments-in one GEN3 broker?!
Best,
Max
You are rather fragile aren’t you? But considering this thread, that is no loss…
