Knowledge that Transforms

Using a large cross-section of 22 countries, we analyze whether macro risks are spanned by the yield curve. Our tests show that macro information, both first and second moments, provides additional explanatory power for bond excess returns beyond yield factors, contrary to the spanned model implications. However, when considering in-sample fit and term premium predictions, distinguishing between spanned and unspanned term structure models makes no difference. These findings are robust across the cross-section of countries. We find the strongest out-of-sample predictive power for second moments of macro information for long-term emerging market bonds.

ABSTRACT Primary dealers are sophisticated investors appointed by sovereign issuers to buy, promote, and distribute sovereign debt that develop a deep knowledge of sovereign debt markets. This study examines domestic primary dealers’ sovereign debt classification, which is presumed to reflect their superior information sets on expected sovereign yields. We hypothesize that when this classification is disclosed in financial statements, peer banks adjust their asset allocation accordingly. We first document the predictive ability of domestic primary dealers’ sovereign debt classification for future sovereign yields. Next, using a sample of 6,437 bank-year observations over the 2012–2019 period and after controlling for publicly available information and other determinants of banks’ asset allocation decisions, we show that peer banks divest financial instruments and increase loans when domestic primary dealers disclose more sovereign debt at amortised cost. These effects are more pronounced among peer banks facing greater informational disadvantages.

By decomposing close-to-close mid-quote returns of ETFs into their overnight and intraday components, we find that the overnight return is significantly positive, whereas the intraday return is negative. This overnight–intraday return differential is ubiquitous across ETFs tracking different asset classes or assets located in different time zones. This phenomenon cannot be explained by differences in overnight and intraday risks, macroeconomic announcements, or information asymmetry. Instead, our analysis reveals that the return pattern is primarily driven by demand shocks from retail investors and limited supply from arbitrageurs. These results indicate that the convenience of buying ETFs during intraday trading hours carries a hidden cost to investors.