Using a panel cointegration framework, the article explores the two‐way link between FDI and growth for a panel of 23 developing countries. In addition, it investigates the impact of liberalization on the dynamics of the FDI and GDP relationship. A long‐run cointegrating relationship is found between FDI and GDP after allowing for heterogeneous country effects. The cointegrating vectors reveal a bidirectional causality between GDP and FDI for more open economies. For relatively closed economies, long‐run causality appears unidirectional and runs from GDP to FDI, implying that growth and FDI are not mutually reinforcing under restrictive trade and investment regimes.

Data revisions routinely introduced by the World Bank can lead to significant revisions in empiricalresults. We show this by re-estimating our aggregate indicator for predicting the 1997 Asian crisis utilizing the 1999 and 2004 updates of the 1996 World Bank data and comparing these results to those we obtained (this Journal, 2000) for predicting the same event using the original, unrevised, 1996 World Bank data. Since most data-gathering organizations routinely revise their data, this may represent a much greater problem for policy makers than might be recognized.Researchers utilizing data for conducting empirical studies and hypothesis testing may face a problem much more serious than generally recognized because of data revisions. The World Bank, the International Monetary Fund, the United Nations, and other international and national, public and private, data-gathering organizations routinely revise their data, not only for the previous period of observation but also for several decades. Often, these data revisions are significant and can lead to very different results of empirical estimations and hypothesis tests of the same event utilizing data from before and after the revisions. Coefficients that were significant can become insignificant, and vice-versa, and conclusions previously accepted can be reversed.Errors in variable have been explored econometrically (eg. Greene, 2002) and are one source of problems with data revisions. If we consider the revisions simply as errors, they can cause spurious results, or be correlated with explanatory variables. In addition to these statistical problems of error in variables, the revision method itself can add serial correlation or unit roots (Patterson and Heravi, 1991; Santangelo, 2004) independent of the original data.Oskar Morgenstern (1963) questioned the feasibility of any international comparison due to these data problems. He cited problems from initial data collection, sampling problems, and interpolation errors. David Runkle (1998) finds that

No abstract

Contingent claims separate revenue and cost into two different time periods. Revenue comes in the initial origination process, while the cost comes upon completion of the contract in the event of default. With banks increasing contingent claims in recent years, a higher taxable income leads to a shift in a bank's balance sheet toward tax‐free income and tax‐shielding liabilities. This provides a valuable case‐study of corporate finance theories of tax management. This paper builds a model to illustrate the income features of contingent claims. Call Reports from 1990–1996 are examined, and show significant evidence of increases in leverage associated with contingent claims.

No abstract