Answers are grounded in this paper and cite their sources.
India's foreign-exchange reserves rose from 7.6 billion US dollars in 1991, about four months of imports, to 700.1 billion in 2025, 8.85 months of imports and 17.7% of GDP. I trace the standard adequacy ratios with World Bank data (vintage 8 October 2026) and then ask what level is optimal when reserves are insurance. Holding a unit of reserves costs δ a year; with probability π a year brings a stress outflow X (as a share of GDP), and each unit of outflow not covered costs k. I prove that the expected cost is convex in reserves and is minimised where the probability that the stress outflow exceeds reserves equals δ/(πk), so that optimal reserves are a quantile of the outflow distribution. I also prove a closed form under an exponential tail and a bound on the cost of holding more than the optimum. I calibrate the distribution on annual reserve losses in 116 low- and middle-income economies over 1992–2025: 683 of 3,609 country-years saw losses above 1% of GDP. With δ equal to the September 2026 spread of India's 10-year yield over the US 10-year yield (2.07 points) and k between 1 and 2, the rule gives optimal reserves of 6.6% to 8.8% of GDP. India's 17.7% exceeds all but 1.3% of observed stress-year losses. The excess insurance costs at most 0.23% of GDP a year by the bound, and 0.09% to 0.15% when computed directly. India's reserves are high by this rule, but the cost of holding them is small.
No takes yet. Share an insight, caveat, or question.
Manoj Kumar Aguru (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: