Econometrics

Descriptive statistics

The table below demonstrates the descriptive statistics for the data of US current account and exchange rates vis-a-vis Canada, Sri Lanka and Mexico. Exchange rates of these particular countries have been chosen because they have shown a good correlation with US current account data. Statistics for the variables were obtained from time series data for the period of January 1998 to February 2008.

[pic]

The US current account had been in negative all through the period and averaged at -40930,84 million dollars. During the period US trade balance hit the lowest negative of -67606 million dollars and could not rise beyond -10290 millions. The standard deviation of the data is quite large as relative to individual observations. This is in turn the indication of the reasonably high fluctuation in the data set.

The average amount of Canadian dollars per US dollar had been 1,36 units whereas much bigger figures of Chinese Yuan and Sri Lanka Rupee corresponded to a unit of US dollar. Among three exchange rates the highest rate of both appreciation and depreciation against USD was observed in Rupee of Sri Lanka. This trend could be easily noticed from a high standard deviation (14.64) of the currency data which had been the highest of the three.

The skewness and kurtosis are the attributes of the shape of the histogram that could be drawn from the given data sets[1]. Skewness measures the symmetry or asymmetry of the data distribution whereas kurtosis shows how peaked or flat the data histogram is. A negative skewness figure would mean that the shape of the data histogram tends to left while vice versa is true if skewness is positive. Below we can observe the histograms of three exchange rates all with negative skewness figures having longer and heavier tails on the left.

Histograms for above variables and their skewnesses

[pic]

The current account figures have the skewness close to zero – a characteristic that belongs to...