I am bringing to attention there are several bugs when using the personalized equity curve or at least it is not clear how it works. I do not trust it, myself. I think this feature should be fixed or temporarily disabled as it will mislead potential subscribers. The monthly % won/loss does not reset either.
I can input my same starting capital and get different equity curve for my system even when I turn off all the fees.
It is also not clear how the scaling works for futures and for stocks. I do recommend showing futures systems based on an industry typical minimum margin requirement.This will present the real time results in a non biased way in terms of vendor risk aversion/aggression.
No answers?
If you go to my system and input 100k with no fees and no commissions, the per month equity is vastly different then what is shown from default/
I feel the “personalized equity” curve is thus wrong. Please explain why they are so different or correct me if I’m wrong.
This feature should be disabled until it is corrected.
I’m, also, not sure how position size is calculated or adjusted for using different starting accounts. If share size is reduced BUT futures are not then this will give very misleading results for the virtual same trades.
I’ve explained in previous posts (sorry they got lost in the ether) the way the algorithm for personalized charts work: p/l is not calculated on a daily marked-to-market basis, but rather on a only-when-trade-is-closed basis. So for systems without frequent trades, or systems that let trades run for a long time, the approximations used in the personalized charts will be overly granular, and not too useful.
This is a result of the algorithm I needed to use to perform the calculations in a relatively fast real-time environment. Ultimately, I plan to change that algorithm to use marked-to-market results, but that will be a longer term project, and will require a bit of engineering. So, while I agree it will be nice to make the approximations more accurate even for slower trading (or less-frequent-trading systems), right now the approximation is what it is: useful in many cases, less useful in others. I do not agree that this means the feature should be removed entirely. It’s a work in progress, and in certain cases is very cool and useful; in other cases it is less so.
Thanks: that makes sense and seems reasonable. But, what about the per shares used when using a different starting size?
The reason I ask is because early on I sized at 1000 SPY shares and then sized @ 2 EMINI contracts. These are roughly the same but when I enter a 10k starting account, my equity curve appears to do better when I started with the EMINIS.
I’d prefer to see ALL equity curves based on 2 factors:
1. Minimum account size for a given starting account and max drawdown.
2. Minimum account size based on number of contracts held/traded at one time, based on an agreed upon industry standard.
Why?
1. A vendor that is more risk adverse may not get subscribers even though the system has as much edge.
2. A vendor that is more risk aggressive may lose non aggressive subscribers simply because he chose a different aggression posturing.
—
This would allow profiling systems, at least futures systems, in a vendor risk agnostic way. True, a vendor can give guidelines for the minimum starting account size if they know it (I don’t as my system has large discretionary input) but this vendor agnostic method should still be useful.
