Journal of Applied Economic Sciences
Volume IV/ Issue 1(7)/ Spring 2009
Table 2. Data sources and frequencies as they were retrieved
|
Variables for Stock Market Crises Index |
Source |
Frequency and Period |
|
Country's Main Stock Market Index |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Independent Variables in Stock Market Model |
Source |
Frequency and Period |
|
5 YR Credit Default Swaps in USD |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Price/Earnings Ratio of Index |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Stock Returns |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Real Interest Rates |
IMF IFS line 60 and 64.X |
Monthly, 2000:10 - 2008:8 |
|
Variables for Currency Market Crises Index |
Source |
Frequency and Period |
|
Exchange Rate National Currency per USD |
IMF IFS line AE |
Monthly, 2000:10 - 2008:8 |
|
Total Reserves Minus Gold |
IMF IFS line 1L.D |
Monthly, 2000:10 - 2008:8 |
|
Independent Variables in Currency Market Model |
Source |
Frequency and Period |
|
5 YR Credit Default Swaps in USD |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Gross Domestic Product |
IMF IFS line 99B |
Annual, 2000 - 2008 |
|
Current Account |
IMF IFS line 78ALD |
Annual, 2000 - 2008 |
|
Unemployment Rate |
IMF IFS line 67R |
Annual, 2000 - 2008 |
|
Terms of Trade |
Kaminsky et al (1998) |
Quarterly, 2000:9 - 2008:6 |
|
Country's Main Stock Market Returns |
Bloomberg |
Monthly, 2000:10 - 2008:8 |
|
Real Interest Rates |
IMF IFS line 60 and 64.X |
Monthly, 2000:10 - 2008:8 |
|
Change in Consumer Prices |
IMF IFS line 64.X |
Monthly, 2000:10 - 2008:8 |
We used linear interpolation to convert the annually and quarterly data observations into
monthly frequencies. The variables converted were: GDP growth, current account and unemployment
rate. However, in cases where data with lower frequency was available for some countries on these
series, we retrieved the lower frequency data. For the monthly series, the last day of the month is
used.
By correlation coefficients, we checked whether CDS is showing information that is already
contained in another variable. Table 3 below shows correlation matrices for both markets. The
correlation coefficient that measures the relationship of CDS to other variables falls in the range of -
0.063 (GDP growth) and 0.548 (Inflation). Any coefficient significantly different from 1 and -1 shows
that we cannot fit a linear relationship between those variables. Therefore, we argue that information
contained in CDS is distinct from the one contained in other variables and hence, it can improve
results.
Table 3. Correlation matrices of variables used in stock market and currency market models
Stock Market
|
One Month Return CDS Real Interest Rate Price/Earnings Ratio |
Return 1 -0.01 |
CDS 1 0.3433 0.0825 |
Real Int Rate 1 -0.0821 |
P/E Ratio ________________1 |
|
Currency Market | ||||
|
CDS |
Terms |
CurrAccount |
Unemploy. Inflation Real Int Rate GDP Growth Return | |
|
CDS |
1 | |||
|
Terms of Trade |
0.0442 |
1 | ||
|
Current Account |
-0.0487 |
0.0426 |
1 | |
|
Unemployment Rate |
0.1371 |
0.3065 |
-0.1246 |
1 |
|
Inflation |
0.548 |
0.1443 |
0.0569 |
-0.0367 1 |
|
Real Interest Rate |
0.1129 |
0.0884 |
-0.1359 |
0.0699 -0.0454 1 |
|
GDP Growt h |
-0.0639 |
0.0444 |
0.1508 |
-0.1015 0.023 0.0666 1 |
|
One Month Return |
0.0049 |
0.0148 |
0.1609 |
-0.0249 0.0377 -0.0614 0.1201__________1 |
131