Forecasting The Effectiveness Of Non-State Pension Fund Investment Strategies: The Case Of Georgia

Abstract

This paper evaluates the effectiveness of Georgia’s non-state pension fund using monthly data for 2022–2023 and develops short-to-medium-term forecasts to 2030. We assemble indicators on pension contributions, investment returns, and the exchange rate, and define a fund-effectiveness measure to capture the system’s ability to meet obligations. Ordinary Least Squares (OLS) regression is used to estimate the relationship between effectiveness and its determinants; we verify time-series properties with Augmented Dickey–Fuller tests, assess Granger causality, and construct an ARIMA model to forecast dynamics under current policies. The results indicate a positive and statistically significant association between contribution inflows and effectiveness, while the link between investment returns and effectiveness is weak or negative during the sample period, consistent with portfolio transition, limited diversification, and market volatility. Exchange-rate movements appear positively associated with effectiveness, reflecting the valuation of foreign assets, but also imply additional risk to beneficiaries’ purchasing power. Forecasts suggest that, absent reforms, effectiveness is unlikely to rise materially by 2030. We discuss implications for policy design and governance. Priority actions include: strengthening risk management and diversification; aligning asset duration with liabilities; instituting foreign exchange risk hedging where appropriate; publishing clearer performance and risk disclosures; and enhancing default portfolio design and member choice architecture. The findings inform the calibration of contribution policy and investment guidelines for Georgia’s non-state pension system.