- May 2021 (3)
- April 2021 (2)
- March 2021 (5)
- February 2021 (4)
- January 2021 (4)
- December 2020 (4)
- November 2020 (5)
- October 2020 (4)
- September 2020 (4)
- August 2020 (4)
- July 2020 (1)
- June 2020 (4)
Exchange rate and GDP in the last two crises
I am going back to the "in the last two crises" commentary series. Today's topic is the exchange rate movements. I want to analyze how much the Turkish lira depreciated in real terms and to what extent this was translated into export gains and therefore GDP growth.
First, I will compare the (real) exchange rate movements during both crises net of inflation. Then, I will provide the GDP and real exchange rate movements together. Details are down below. The results are as follows:
During the 2001 crisis lira depreciated more in real terms than in case of the last crisis: the index value decreases to 70 in the third quarter after the 2001 crisis while the lowest index value witnessed after the 2008 crisis was 88 (Figure 1). In the 2001 crisis, the lira appreciated rapidly following the trough. However the uncertainty that raised in the second quarter of 2002 (the disease of PM Mr. Ecevit and the resultant possibility of early elections) as well as the Iraqi war afterwards have brought rapid depreciation of the lira. Therefore, there exist two major shocks affecting the real exchange rate. As of the first quarter of 2010 the real value of the lira has approximated to the pre-crisis level.
Until the second shock I mentioned above, the real exchange rate and GDP movements were similar in 2002. The period where the lira depreciated rapidly (where the index tended downwards) was also the period where the GDP decreased steeply. Then the two indicators moved upwards together. This harmonious movement ended following the second shock (Figure 2). In the case of the last crisis, both of the indicators move together in the same direction: GDP decreased in periods where lira depreciated in real terms and vice versa (Figure 3).
I am aware that those familiar with the issue have frown down. And they are right. First, there are tons of indicators that affect the GDP levels. My analysis focuses only on real exchange rate and GDP level. Second, the correlation between the indicators does not have to be simultaneous; there might be a certain delay. For instance, GDP might go up one quarter after the real depreciation of the lira. Third, the real exchange rate should first affect the export and import performance before having an impact on the level of GDP. As the lira depreciates it is expected that exports will increase and imports will decrease. This way production of goods and services would increase improving the GDP level.
Therefore, it is necessary to examine the correlation at least between exports and the real exchange rate. I will deal with this tomorrow. But the figures below raise questions in one's mind. It appears as if the real value of the lira did not play a major role in export and GDP movements. But attention: I did not say it did not play any role. I argue that it did not play a major role.
And here are the details: The first figure was drawn in a way where the real exchange rate movements during both of the crisis will interlock. "0" at the horizontal axis illustrates the peak level of GDP before the crisis which corresponds to the last quarter of 2000 for the 2001 crisis and the first quarter of 2008 for the global crisis. The analysis here examines the crisis over a period that covers the quarter one year before the peak quarter (-1) and eleven quarters after the peak level (11) (this is eight quarters in the case of the global crisis).
Real exchange rate index is the three-month average of the monthly indices announced by the Central Bank of Turkey. A rise in the index value implies the appreciation of the lira in real terms. GDP data is net of seasonality effects and calendar effects.
Similar to the method I used to employ, the GDP value at the point "0" is 100 and the figures for other months are adjusted to this. Similarly, the real exchange rate index values corresponding to the periods where the GDP peaked are 100 and the values for other months are adjusted accordingly.
This commentary was published in Radikal daily on 16.05.2010