The paper uses structured machine learning regressions for nowcasting with panel data consisting of series sampled at different frequencies. Motivated by the problem of predicting corporate earnings for a large cross-section of firms with macroeconomic, financial, and news time series sampled at different frequencies, we focus on the sparse-group LASSO regularization which can take advantage of the mixed-frequency time series panel data structures. 

Using machine learning techniques, we uncover an important number of dealers in the U.S. municipal bond market who focus on geographically adjacent states, a characteristic distinct from dealer centrality. These “specialized” dealers enjoy larger market shares in states with greater local ownership and in local bonds with more complex features. We also find that trades intermediated by these specialized dealers have significantly larger markups than those intermediated by national dealers.

This paper surveys the recent advances in machine learning method for economic forecasting. The survey covers the following topics: nowcasting, textual data, panel and tensor data, high-dimensional Granger causality tests, time series cross-validation, classification with economic losses.

George Floyd's murder caused many firms to reveal how exposed they are to racial diversity issues. We examine investor and firm behaviors after this socially significant event to provide evidence on the valuation effects of the exposure and ensuing corporate responses. We develop a text-based measure of a firm's exposure to racial diversity issues from conference call transcripts and find that, after the murder of George Floyd, firms with diversity exposure experience a stock price decrease of approximately 0.7% around the date of the conference call. We provide evidence that this effect is attributable to race-related exposure and not gender-related exposure. Initiatives taken by firms mitigate the negative market reaction.

We examine firm disclosure choice during the initial public offering (IPO) roadshow presentation to understand the informativeness of a management presentation designed to attract investors. Although firms submit a comprehensive registration filing during the IPO, managers also prepare a roadshow presentation, which is shorter and typically allows managers more autonomy to select the information released and how it is discussed. We find that IPO roadshows have significantly more positive, less negative, and less uncertain language than the SEC filing.

Drug patents are different. To improve their quality ex ante, regulators can use predictive models. Drug patents provide crucial incentives for developing life-saving medicines, but when improperly granted, they can contribute to delays in competition and limit access.

Reliably detecting insider trading is a major impediment to both research and regulatory practice. Using account-level transaction data, we propose a novel approach. Specifically, after extracting several key empirical features of typical insider trading cases from existing regulatory actions, we then employ a machine learning methodology to identify suspicious insiders across our full sample.