یادداشتی در مورد مدل سازی بازارهای سهام جهانی با داده های غیر همزمان

We investigate information transmission in world equity markets using lengthy time series of daily data for nine developed equity markets over the 1999–2014 (post-Euro) time interval. Three nine-variable systems are examined, including stock indexes in both local currency and US dollars plus foreign exchange rates. After finding only weak evidence of cointegration in the three nine-variable systems, vector autoregression models are used to identify which price indexes drive the others. Our initial tests are consistent with earlier findings; that is, the US market appears to be the clear price leader. However, after constructing alternative models with the US data series lagged one period (making it the first market reported each day rather than the last), the results are reversed; the US is no longer the price leader. The US market is driven by nearly all other markets, while making little or no impact on the others. Thus, in earlier studies using nonsynchronous data reported by calendar date, it only appears that the US drives other stock markets, when in fact the US is merely an equal participant in information transmission around the world.