The Cost of Prudential Regulation: Evidence from Operational Expenditure per Capita in Pension Funds

Abstract

This paper evaluates whether the 2019 prudential regulatory event is associated with persistent changes in pension-fund cost dynamics. The analysis uses only estimates already reported in the thesis results archive and combines ARIMA intervention models, polynomial time-trend regressions, Chow structural-break tests, and time-series decompositions. For dpc, the ARIMA intervention term is positive and significant in the control group (xreg = 370.0129; p = 0.0041) but negative and non-significant in treatment (xreg = -156.3812; p = 0.6023). Chow tests reveal strong breaks in both groups (Control: F = 3.218528, p < 0.0001; Treatment: F = 2.641218, p < 0.0001). Complementary expenditure regressions show a large baseline treatment coefficient (64.8547; p < 0.001), but no significant event or event-treatment effect. Decomposition results indicate that post-2019 change is expressed primarily through greater volatility around trend rather than a clean level shift. Taken together, the evidence supports the existence of meaningful post-2019 cost reconfiguration, while suggesting caution in attributing a differential treated-group increase to the intervention alone.