We examine search-based peer (SBP) groups proposed by Lee, Ma, and Wang (2015) and their relationship with commonality in liquidity. Our results confirm that SBP affiliation is a significant determinant of commonality in liquidity and, unlike market- and industry-commonality, SBP-commonality has been increasing over the past 15 years. We separate retail from institutional investor queries by tracing the IP locations of Electronic Data Gathering, Analysis, and Retrieval (EDGAR) searches. Our results show that retail investors are responsible for roughly 85% of the EDGAR searches that generate SBP groups. Overall, our study provides new evidence of a significant demand-side commonality associated with SBP affiliations.

We examine how banks’ cross-border lending reacts to changes in the intensity of liquidity regulation using a new dataset on the UK’s Individual Liquidity Guidance. An increase in requirements reduces banks’ cross-border lending growth to banks and non-banks. But banks’ business models determine how they adjust: banks with higher deposit shares, such as those with UK retail operations, protect lending more; banks also preserve lending to countries they do most business, cutting elsewhere; foreign subsidiaries from countries not eligible to issue High Quality Liquid Assets (HQLA) show the strongest reduction in lending; in contrast, subsidiaries from HQLA-issuing countries cut intragroup lending.

One dominant approach to evaluate the causal effect of a treatment is through panel data analysis, whereby the behaviors of multiple units are observed over time. The information across time and units motivates two general approaches: (i) horizontal regression (i.e., unconfoundedness), which exploits time series patterns, and (ii) vertical regression (e.g., synthetic controls), which exploits cross‐sectional patterns. Conventional wisdom often considers the two approaches to be different. We establish this position to be partly false for estimation but generally true for inference. In the absence of any assumptions, we show that both approaches yield algebraically equivalent point estimates for several standard estimators. However, the source of randomness assumed by each approach leads to a distinct estimand and quantification of uncertainty even for the same point estimate. This emphasizes that researchers should carefully consider where the randomness stems from in their data, as it has direct implications for the accuracy of inference.