Comment and Discussion

Comment by Ken Kavajecz: The key result of the paper is that the same factors leading to a country's stock market development are also leading to the migration of stock markets overseas. This is a very important result that I have not seen anywhere else, and the authors draw a number of policy implications from it. Their recommendation to emerging market countries is to concentrate on promoting fundamentals rather than setting up and promoting stock market development since the stock market is going to migrate overseas anyway. Instead, the authors argue that emerging market countries should facilitate the integration of their trading systems with the more developed markets. In my discussion, I address five points.

First, endogeneity and causality are difficult to determine. This is a criticism of the literature, not of the paper per se. I took these two lines from the paper: "Financial markets tend to develop as income per capita grows," and "the liquidity of the stock market has been found to be a useful predictor of future economic growth." We really cannot say that the financial market is driving gross domestic product (GDP) or that GDP is driving the financial market. We have to be realistic and acknowledge that they are driving each other. Again, this is by no means an attack on the paper; rather, it is a comment on the entire set of papers in this literature.

Second, are the results driven partially by the sample period or by the particular countries involved? To what extent are the results driven by the bull market of the 1990s? It would be easy to issue an American depository receipt (ADR) when people are willing to buy anything. I wonder whether these results would be mitigated somewhat if the authors investigated different periods?

Do the results change when you eliminate the United States and the United Kingdom from the sample? It is easy to think about foreign trading (an ADR, for example) and domestic trading when you are talking about a small emerging market country. But I am not sure what it means to have foreign trading for the developed countries like the United States. Furthermore, the data set must have many firms from the United States, United Kingdom, Germany, and so forth. I am curious to see how the results would change, if at all, if the United States and United Kingdom were eliminated from the sample.

Third, I would expand the set of explanatory variables. I agree with the use of every one of their variables. Since much of this research is exploratory, I would encourage the authors to investigate alternative variables further. The index of economic freedom ranks countries based on a number of specific criteria, for example, development of the banking sector, fiscal policy, and taxes. Alternative law and order variables, such as political risk or governmental bureaucracy, may generate different results. ¹ Measures of accounting standards are also likely to be important. All of these would be good variables to explore.

I was not surprised that the trading cost variable failed to be significant, given that the vast majority of it appears to be commissions and fees. I do not think of commissions and fees as having anything to do with a stock market or a stock exchange. Commissions and fees are largely payments to brokers. That is not the business of an exchange. The business of an exchange, as I see it, is to provide liquidity. How well an exchange does its job is best measured by the bid-ask spread, quoted depth, and price impact variables. Those sorts of things should be included in a trading cost variable for a stock market or stock exchange.

Fourth, I do not see how we can necessarily say that emerging market countries should abandon their domestic exchanges and let trading migrate overseas. First of all, I want to make a distinction between markets and exchanges. Most of the paper discusses markets. The authors are talking about...