We test the signaling and wealth transfer hypotheses around the announcement of share repurchases using a recent and larger sample of data than previously examined while employing a methodology designed to enhance the power of our tests. Disentangling the wealth transfer and signaling hypotheses is difficult; they are not mutually exclusive and can have opposite effects for bondholders. Wealth transfers decrease bondholder wealth while positive signals increase it; the combined result obscures tests of each hypothesis. By focusing on sub-samples where signaling is more and less likely to be present we increase our ability to isolate the separate effects. In addition to traditional tests of wealth effects, we feature information inherent in the correlation of wealth changes to equity and debt. Our results are generally consistent with the positive signaling effect of stock repurchases, but also provide some support for wealth transfer. Our work also emphasizes the importance of trying to disentangle the various hypotheses. In the subset of option funding repurchases, where signaling effects are less likely, the positive correlation of wealth changes between stockholders and bondholders is completely eliminated. Bond ratings are much more likely to be upgraded in samples without executive options which is precisely where the signaling effects are expected to be concentrated. Firms with weaker shareholder rights experience greater bondholder wealth losses at the announcement of stock repurchases.