Solving the Yield Puzzle: Building Lithuania’s Term Structure from the Fragmented Data

Assessing government bond yields in smaller, less liquid markets such as Lithuania presents unique challenges due to limited secondary market activity, sparse trading data, and unreliable existing benchmarks like Bloomberg estimates. This study addresses these limitations by integrating data from primary issuances and secondary market quotes to construct a more accurate and consistent view of Lithuania’s yield curve. Employing the Nelson-Siegel-Svensson1 (NSS) model, we estimate a synthetic 10-year benchmark yield that offers improved comparability with larger Eurozone economies. Our analysis shows that while yield spreads between primary and secondary markets are generally minimal, they widened during periods of geopolitical and economic stress—particularly following outbreak of the Russia–Ukraine war and the COVID-19 pandemic—reflecting heightened risk perceptions and liquidity constraints. Notably, primary market yields temporarily rose above secondary quotes after the Russian invasion of Ukraine, signalling price discovery issues. Additionally, while a spread was observed between international issuance yields and their corresponding synthetic benchmarks, this was not the case for domestic bonds. We also find that central bank interventions, such as reverse auctions conducted by the Bank of Lithuania, played a stabilizing role during times of elevated uncertainty. These findings underscore the value of combining diverse data sources with robust yield modelling to produce reliable benchmarks and better understand dynamics in illiquid sovereign debt markets.