How do lead lag relations in cross listings move?

We investigate the lead lag relations in portfolio returns, volatility, and spread. We test two comprehensive hypotheses on multi lag spillovers in returns, volatility, and spread using two sets of portfolios. The first set of portfolios are matched American depository receipts (ADRs) and their corresponding home shares from 26 developed and emerging countries/exchanges around the globe; the second set includes large, mid, and small-cap portfolios of ADRs. Our results based on bi-variate vector auto regression (VAR) methodology indicate bidirectional causality for ADR and their corresponding home shares portfolios for almost all our sample countries/exchanges. We compute daily returns from closing prices and volatility and spreads using daily high and low prices. Our results on large, mid, and small-cap portfolio returns find small-cap portfolios lead returns in both large and mid-cap portfolios. We further investigate these portfolio returns, volatility, and spread.

We develop and test two empirical hypotheses, size spillover and cross market spillover. So far, we have limited but encouraging results to report. Contrary to the existing evidence in the literature, our empirical results related to large, mid, and small-cap portfolios find small-cap returns leading large and mid-cap returns. We continue to probe into this strikingly anomalous findings. Results from bivariate VAR tests between ADR and matched home shares portfolios indicate scattered serial auto correlation but highly significant multi-lag cross dependence for all three tested variables- return, spread, and volatility. Granger causality tests indicate strong bidirectional causality between ADR New York Stock Exchange (NYSE) and home shares (primary listing exchange) portfolios for return, volatility, and spread for most countries/exchanges.