When Static Risk Labels Break Down: Regime-Sensitive Mismatch in Pension Fund Classification - Evidence from Türkiye, 2005–2025
EFİ 2026 Uluslararası Ekonomi Finans ve İşletme Kongresi , Kayseri, Türkiye, 14 - 15 Mayıs 2026, cilt.1, ss.73, (Özet Bildiri)
- Yayın Türü: Bildiri / Özet Bildiri
- Cilt numarası: 1
- Basıldığı Şehir: Kayseri
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.73
- Kocaeli Üniversitesi Adresli: Evet
Özet
Official risk labels do more than inform investors: they act as a choice architecture that steers retail savers toward a permitted universe of products. Whether such labels track the risk–return patterns funds actually deliver is therefore a question of investor protection rather than technical classification. This study tests this label–reality gap in the Turkish private pension system (BES), where official 1–6 risk classes determine fund eligibility, using 395 funds observed over 2005–2025. Because the period spans two decades of pronounced monetary-policy regime change, it offers an unusually sharp setting in which to ask whether label reliability is stable or regime-dependent. Period boundaries are set within a structural-break framework: Bai–Perron multiple structural-break analysis on the monthly CBRT policy-rate series confirms the 2018 and 2023 breaks, yielding five sub-periods. On the risk dimension, volatility-based empirical classes match official classes for 30.6% of funds—roughly twice the permutation benchmark (15.8%, p<0.001), with a Spearman correlation of 0.567—indicating that labels track risk but with coarse granularity. On the return dimension, alignment is strongly regime-sensitive: a fund-by-period panel mismatch logit with fund-clustered standard errors shows investor-adverse mismatch 14.5 and 16.3 percentage points higher in the Heterodox and HeterodoxIntensification periods than in the Orthodox baseline (p<0.01), peaking in 2021–2022. Year-level ordered logit estimates reveal that the risk–return ordering remains positive in 2023, turns neutral in 2024, and inverts in 2025—high-rate disinflation reversed the return hierarchy with a roughly one-year lag. Static risk labels thus carry meaningful information about risk yet can systematically distort the return ordering implied to investors during regime transitions, a finding that generalizes to any system built on static, ex-ante label architecture and that motivates regime-sensitive review mechanisms.