When I backtest Z12 in Zorro 1.83.2 the result is a huge loss (-153.268€ with my test settings). After replacing Z12.fac with the old version from 1.74.4 the backtest is profitable again (+67.426€). So may it be that Z12.fac is not optimized at all but somehow "buggy" in the new version?
The problem in the new Z12.fac version is mainly the XAG/USD:BB:L component I think. When I set this to .000 (as it was in the version from Zorro 1.74.4) I get +98.242€ as backtest result. But the drawdowns are about 3 times as big as with the Z12.fac from the old version, though...