Mert Önen

Financial repression has historically accompanied periods of high fiscal needs. Yet there is no high-frequency measure of repression for emerging markets (EMs), where weaker institutions and lower debt tolerance make it especially prevalent. To that end, I examine the Domestic Bond Premium (DBP): the yield spread between EM local currency (LC) government bonds and AAA-rated supranational bonds in the same currency and tenor. Using data from 4,017 supranational bonds across 11 EM currencies, I document that the DBP is frequently negative and remains stable during periods of financial stress when government LC yield spreads over US Treasuries widen sharply. A simple portfolio choice model decomposes the DBP into default, liquidity, and financial repression components. The empirical evidence supports the model predictions by showing that default risk and liquidity alone cannot account for the low DBPs, pointing to repression as a key force bringing government borrowing costs below the supranational benchmarks.

This article examines sovereign bond markets to assess the current state of Original Sin, the inability of a country to borrow (abroad) in its own currency. We present a synthesis of different strands of the literature using a new, tailored dataset. We find that major emerging market economies (EMEs) have made progress towards overcoming original sin by issuing more government bonds in local currency while promoting foreign participation in domestic bond markets; this went hand in hand with rising exposure to EME currencies among foreign investors. In panel regressions, we show that country-specific variables played a role alongside global push factors. However, progress has been slow and uneven, with a key role for institutional development. Progress is most evident among major EMEs, and stronger for sovereigns than for other issuers. Reducing reliance on foreign currency borrowing implies a greater role for investors whose sensitivity to currency risk can make capital flows more volatile—reintroducing the problem in a different guise, as original sin redux.