This property of the regional factor renders it a useful measure of co-movement in flows. The factor does a good job of identifying the Mexican crisis (December 1994-April 1995) and the East Asian crisis (July 1997-January 1998). Both episodes are clearly associated with a strong attenuation of emerging market inflows. It appears that foreign investors held fast during the Mexican crisis, and actually slightly withdrew assets in the midst of the Asian crisis. During the intervening period, these investors increased their exposures to emerging market equities by almost 1% of market capitalization.
As to the long-debated relationship between flows and prices, Figure 5 provides a visually striking piece of evidence. The figure demonstrates that the de-trended regional factor and emerging market equity prices move together at low frequencies. (We later test their tendency to move together at higher frequencies.) The co-movement could, of course, be attributed to either overreaction, information shocks, or demand shocks.
However, it is worth noting that the co-movement is not likely to be attributed to the Brennan and Cao information hypothesis. They stress international investors are at an informational disadvantage vis-a-vis local investors, i.e., that international investors may face a disadvantage in knowing as much about local shocks. But Figures 4 and 5 depict common or regional components of flows across these countries. This component clearly moves (strongly) with returns.
Having estimated ftR, we project^ onto it to generate the orthogonal decomposition shown in (1). The country-specific (or idiosyncratic) component for a given country is just the residual from this regression. To gain a sense of the importance of the factor for regional and country flows, we report the R2 from the regressions. For each region, Table 2 shows the fraction of total regional variation that is explained by the regional factor, as well as an average across countries of country flows explained by the regional factor. Notice that for local regions such as Latin America, East Asia, and Emerging Europe, the regional factor explains about 20% of the variation of flows for countries within the region. bad credit loans not payday loans