Purpose We test the informational efficiency of Venezuelan USD sovereign bond yields when the black market exchange-rate premium (BMERP) changes. Design We use a non-parametric, asymmetric, Granger causality test to test our hypothesis. Findings We find that the bond market with less than or equal to 5 years of maturity seems to be efficient when good news is released on the BMERP. However, this market is not informationally efficient, and when combined with unbiased bad news regarding the BMERP, arbitrage opportunities are created. Originality/value Capital controls that restrict free exchange-rate mechanisms create arbitrage opportunities with negative news as opposed to positive news.