Selling Risk Retention Bonds Signals Low Quality Collateral

Sean Flynn Jr., Andra Ghent, and Alexei Tchistyi (Flynn et al., 2020) document that choosing to comply with Dodd-Frank’s risk retention rule by retaining a deal’s most junior tranches correlates with lower spreads on the deal’s investment-grade bonds, which they interpret to indicate that this choice of risk retention mechanism signals high collateral quality. By con-trast, we argue that since in practice securitizers actually sell these types of “retained” risk retention bonds, standard signaling models would predict such selling to indicate lower quality collateral. In support of our prediction, we find that this risk retention choice is associated with lower overall deal profitability and worse ex-post collateral loan performance. We reconcile our evidence with Flynn et al., 2020 by demonstrating that their results are not robust to what we believe to be reasonable modifications to their specification. Thus, the evidence suggests that choosing to satisfy Dodd-Frank by “retaining” and selling the most junior tranches actually signals lowquality collateral.