Setting the Margin slider for Z strategies

I'm trying to find out how to set the Margin slider to a proper value for my account. Tried searching the forums, but found no conclusive answer.

Is running a test with different margin settings trying to match my starting capital a good way? For example, running Z12 with margin/risk to 100/10 shows a required capital of $11032. Given that my starting capital was just that, are then 100/10 my recommended settings?

Thanks!

Edit:
I see that a similar question was just asked earlier. http://www.opserver.de/ubb7/ubbthreads.php?ubb=showflat&Number=424411

Yes, set the Margin slider so that the required capital in a test matches your starting capital.

My understanding is that "Capital required" in the performance report comes from the historical "worst case scenario" that would have occurred, had you been trading during the backtest period. Here is what the manual says:

Do you think it's safe to assume that a trader could start with the minimum "capital required" safely, based on the unlikelihood of that "worst case" being exceeded? I'm asking only to understand your line of thinking -- because I had previously assumed that figure was a bare minimum that we should exceed in real trading. However, for nanolot trading "it sure would be nice" to follow that figure religiously as a rule-of-thumb.

Also, what do you mean by the phrase "maximum drawdown normalized to 36 months"? Where does the 36 months come from? Are you saying that in a 10-year backtest, you are only looking at the max drawdown of any consecutive 36-month period? I just want to understand.

THANKS

As to whether it's safe to assume that the worst case is unlikely, in my experience it's not safe. I often see that a worst case is not only likely, it does really happen .

The 36 months are just a convention used in Finance. Parameters that depend on the test period, like drawdown, are usually normalized to 3 years for calculating performance figures.

Thanks, jcl!

And for Zorro. I've been researching system trading for quite a while and it is by far the most honest and eye-opening texts/software I've seen about the subject.

Hi jcl. Been meaning to ask this for a while but I keep forgetting... Since "normalized" has a number of definitions, what does it mean here?

For example, let's say a 5 year Test results in 47 drawdowns ranging from 3000 to 7000 over 5 to 17 bars. How are the foregoing "normalized" to 36 months (formula?), and what number is produced as the DDmax?

Thanks.

DDmax is determined from the simulation, not normalized. It is described in the manual.

The meaning of normalizing a drawdown is not in the manual, but is described in this paper:

http://www.cs.rpi.edu/~magdon/ps/journal/drawdown_journal.pdf

Thanks for the weekend reply!

Sorry - meant "normalized drawdown", not DDmax, as referenced in AR & CR.

Thanks for the link.

OK, I readily confess that paper is over my (current) head. But it is searchable and "normal" is nowhere within... Is there a "Readers Digest" version of how normalized drawdown is computed and how it compares to DDmax?

Thanks.

The paper calculates three drawdown expectancy functions: for a system with positive (u > 0), neutral (u == 0), and negative profit expectancy (u < 0). The essential formula is (7). You can see in fig. 1 how the Q functions, which are proportional to the drawdown expectancy, develop over time.

We're using the second curve in fig. 1, which is a square root function, for normalizing the simulation DDmax to a 3 years period. This means when a one-year simulation produces a drawdown X, the normalized drawdown would be Y = X * sqrt(3).

Thanks jcl!

If I understand correctly, you're using the breakeven case from the paper for the normalization (regardless of whether the simualation is profitable or a loser)?

And, to generalize the equation above, if y is the years in the simulation, DDnorm = DDmax * sqrt(3/y)? Is this for all simulation lengths, or is DDnorm = DDmax for y>3?

Thanks.