A Re-Assessment of the Importance of Accounting Data to the Corporate Bond Market: What Do Large Block Trades Know?
- Author(s): English, Malachy Edward
- Advisor(s): Dechow, Patricia
- et al.
This dissertation evaluates the importance of the uncapped Enhanced TRACE dataset that has previously been rarely used within academic literature. In doing so I answer two important questions, one relevant to academic researchers, and one relevant to financial regulators. First, how economically significant are the differences between the Enhanced TRACE dataset and the Historic TRACE dataset that has been used previously? Secondly, are these differences a result of informed trading, and thus potentially in need of the protection provided by the caps currently imposed by TRACE? I find striking differences between the two datasets. Hidden volume in the periods preceding earnings announcements occurs frequently and is large in size, often exceeding 30% of total volume in the period. Despite this, I find little evidence to suggest that this trading volume is driven by informed investors. The hidden volume shows little ability to anticipate the news in earnings announcements and appears to be somewhat randomly distributed throughout time. My research suggests that researchers should move away from the Historic TRACE dataset and instead utilize the new Enhanced TRACE dataset when examining corporate bond markets. In addition, my research, suggests that large block trades typically are not informed. This provides preliminary evidence supporting the view, held by many market participants, that regulators should remove the currently imposed TRACE dissemination caps. My research supports the claims of these market participants that the caps simply inhibit investors from accurately assessing the quality of trade execution they have received from broker-dealers.