Estimating Systematic Risk in a Sparsely-Traded Frontier Market: A Reproducible Beta Methodology and Sector-Level Evidence from the Mongolian Stock Exchange

Methodology working paper specifying a frontier-market-native CAPM beta estimation procedure for the Mongolian Stock Exchange (MSE, MIC: XULA), where the default Bloomberg/MSCI five-year-weekly calibration breaks down under sparse trading, stale prices, and short survivorship-corrected listing histories. The method uses a three-year rolling weekly (Friday-close ISO-week) window, the TOP-20 capitalisation-weighted index as market proxy, and a 60-trade-day / 30-weekly-observation eligibility floor, applied to all 161 currently-listed MSE common shares. Findings: all seven sector portfolios show beta below unity (ticker-weighted mean 0.49); R-squared exceeds 0.20 for only one ticker; fourteen negative-beta estimates are statistically indistinguishable from zero; and Vasicek shrinkage toward a sector-conditional prior moves median beta by approximately 12 percentage points. Beta estimates are combined with the Damodaran (2023) country equity risk premium for CAPM cost-of-equity, cross-validated across OLS, Vasicek, Dimson sum-beta, and sector-portfolio methods. All data is from public MSE feeds; the procedure is in quarterly production at Metricon.