GDP-Linked Bonds: A Proposal Worth Looking Into

Abstract

GDP-linked bonds (GLBs) have been the subject of recurrent debates among economists and financial market participants but no major debt office has so far decided to launch them. Issuers remain skeptical about the practical feasibility or desirability of these securities, citing potentially high risk premia and various technical issues, including GDP data integrity and revisions, moral hazard and adverse selection. Against this backdrop, we discuss the key features of these securities and we argue that technical difficulties can be resolved, especially if standardized terms are agreed at the international level and statistical offices agree to publish dedicated GDP series. From the standpoint of investors, GLBs would provide direct exposure to GDP growth of a country or basket of countries without running into the more complex valuation issues that characterize equities. Long-dated GLBs also have the interesting feature that their duration may offset the impact of changes in GDP growth expectations. For issuers, regular issuance of long-dated GLBs would gradually raise the share of liabilities indexed to GDP. In the event of a major economic downturn, the debt-to-GDP ratio would rise more moderately than if all debt consisted of conventional bonds. In the limit case of GLBs accounting for the whole stock of public debt and without a price floor, the debt-to-GDP ratio would not change at all apart from the cyclical change in the budget balance.