Saving–Investment Comovement in India, 1975–2025: Feldstein–Horioka Estimates, Cointegration, and a Small-Open-Economy Result on What the Coefficient Identifies

Using World Bank WDI data for India (gross savings and gross capital formation, % of GDP, 1975–2025, source update 8 October 2026), I estimate the Feldstein–Horioka (FH) saving-retention coefficient, test for cointegration, and fit an error-correction model. The full-sample coefficient is 0.70 (HAC s.e. 0.06). For 1991–2025 it is 1.08 (s.e. 0.11), and the hypothesis of unity is not rejected. The Engle–Granger test does not reject no-cointegration (p = 0.24), while the error-correction loading is −0.45 (s.e. 0.10). I then derive in closed form the FH coefficient implied by a small open economy with perfect capital mobility and permanent-income consumption: β = θρ(1+r−ρ)/(1−ρ+rθρ), which rises from 0 to 1 as shock persistence ρ rises. A high Indian coefficient therefore does not, by itself, measure low capital mobility. The WDI saving–investment gap also differs from the WDI current-account balance by 3.3 percentage points of GDP on average (correlation 0.08), which limits any accounting interpretation.