Colombia: Selected Issues Paper

International Monetary Fund. Western Hemisphere Dept.

doi:10.5089/9781498302128.002.A003

Jump to Content Jump to Main Navigation

Colombia: Selected Issues Paper

Author: International Monetary Fund. Western Hemisphere Dept.

Publication Date:: 23 Jun 2014

eISBN:: 9781498302128

ISBN:: 9781498302128

Language:: English

Keywords:: ISCR; CR; Colombia; firm; medium term fiscal framework; revenue; intensive company; manufacturing firm; REER appreciation; firms' profits; REER depreciation; Exports; Real exchange rates; Imports; Global

Download PDF (2 MB)

This Selected Issues paper analyzes spillover risks for Colombia. It highlights that external shocks could spill over to the Colombian economy through the country’s important and growing trade and financial linkages with the rest of the world. Colombia would be most exposed to a decline in oil prices, which could have a sizable adverse impact on the balance of payments, the fiscal accounts and growth. Growth shocks in key trading partners could also have a negative impact, particularly in the United States, which is Colombia’s main trading partner. Colombia’s fiscal rule and adjustment in the context of resource wealth is also analyzed.

Abstract

Recent Performance of Colombia’s Manufacturing Sector: Exchange Rate or Structural?¹

Colombia’s manufacturing sector has performed relatively poorly in recent years despite the country’s strong overall economic performance. Several possible explanations have emerged to explain the sluggish growth in this sector. They include real exchange rate appreciation, weak external demand, structural changes, and other competitiveness issues including poor infrastructure and high labor costs. A firm-level analysis shows some structural changes (related to trade disruption with Venezuela and increased trade competition from China) may partially explain the recent weakness of the manufacturing sector. At the same time, we find no strong evidence that real appreciation has negatively affected the performance of the manufacturing sector.

A. Introduction

1. Colombia’s manufacturing sector has performed relatively poorly since 2008. Real GDP grew, on average, by 4 percent between 2008 and 2013 despite the global financial turmoil of 2008–09, due to favorable commodity prices and Colombia’s strong economic policy framework. At the same time, Colombia’s manufacturing sector output grew only by 0.1 percent on average. As a result, the share of manufacturing in percent of real GDP has declined from 14 percent at the end of 2007 to 11 percent in 2013. The contraction of the manufacturing sector relative to the overall economic activity was more pronounced during the cyclical downturns. The manufacturing sector contracted by four percent during the global financial turmoil, even though the growth of the overall economy remained positive. Similarly, the manufacturing sector exhibited a fairly sharp contraction between the second half of 2012 and the first half of 2013, when overall economic activity slowed but kept growing at the average pace of 3 percent (y/y).

2. Several possible explanations have emerged to explain the recent weakness of the manufacturing sector. The main policy concern has been the overvaluation of the real effective exchange rate which may have hampered the competitiveness of the manufacturing sector; Colombia’s peso has appreciated by 18 percent in real terms against the U.S. dollar between 2008 and 2012. In this light, the recent weakening of the peso is welcomed by policymakers as it could restore the competitiveness of the manufacturing sector and boost its output. Alternative explanations include weak external demand for Colombia’s manufacturing products as well as structural changes induced by international trade. In particular, trade disruption with Venezuela and increased imports from countries such as China and Mexico may have contributed to the contraction of Colombian manufacturing output. Finally, other competitiveness issues, including high labor costs and poor infrastructure, have long been regarded as key obstacles for the growth of the manufacturing sector.

3. This paper examines the effects of real appreciation, external demand, and structural changes related to international trade on the recent performance of the Colombian manufacturing sector. In particular, the paper investigates the real effective exchange rate and external demand specific to the manufacturing sector, using the trade-partner weights taken from the United Nation’s COMTRADE database. Using the same database, the paper also looks into the structural changes in Colombia’s trading partners for exports and imports of manufacturing goods. This paper does not investigate other competitive issues such as high labor costs and poor infrastructure, since these issues have existed for quite some time. However, it is possible that these pre-existing competitiveness issues have made the negative effects of real appreciation and structural changes more acute than they otherwise would have been. The paper also does not look into the effects of technological changes that may have affected the performance of certain manufacturing industries.

4. Real effective exchange rate appreciation has been fairly modest for manufacturing exports. Even though the U.S. is the dominant market for Colombia’s petroleum products, the share of the U.S. in Colombia’s manufacturing exports has been less significant. In addition, many of Colombia’s trading partners for manufacturing exports have experienced real appreciation of their currencies against the U.S. dollar. Accordingly, the loss of competitiveness associated with the appreciation of the real exchange rate does not appear to be so severe vis-à-vis most of the major manufacturing trading partners, with the exceptions for the U.S. and Mexico.

5. There have been some significant structural changes in the international trade of manufacturing products. In particular, exports of manufacturing goods to Venezuela declined sharply after 2008 following deterioration in their trade relations. At the same time, imports of manufacturing goods from China and Mexico increased significantly. Increased manufacturing imports from China have not been driven by the real effective exchange rate, as the Colombian peso has depreciated against the Chinese currency during this period. Rather, it seems to be driven by increased access to Chinese manufacturing products. In the case of Mexico, the real appreciation against the Mexican peso seems to have played a role in increasing imports from Mexico.

6. A firm-level analysis is conducted using a large number of manufacturing companies. The database of the manufacturing firms, prepared by the Colombian Superintendence of Corporations (Superintendencia de Sociedades), includes a large number of private manufacturing companies that satisfy certain criteria in terms of their asset levels.² The database is annual and covers the period from 2000 to 2012 for roughly 4,850 manufacturing companies.

7. There is no strong evidence that real exchange rate appreciation negatively affected manufacturing firms’ profitability during this period. In contrast, on average, real exchange rate appreciation (export-weighted at 2-digit industry level) has contributed positively to manufacturing firms’ profitability. The positive correlation between real appreciation and profit growth may be caused by the presence of imported intermediate inputs, to the extent that the export-weighted real effective exchange rate used in the analysis is correlated with the real effective exchange rate for imported intermediate goods. At the same time, the regression results show that export intensive companies suffered greater losses with real exchange rate appreciation compared to the companies that exported less.

8. Structural changes related to international trade may have contributed to the poor performance of Colombia’s manufacturers in recent years. The results of the firm-level analysis show the structural changes caused by trade disruption with Venezuela and greater competition from China seem to have reduced the average profitability of the firms that belong to the affected industries. The results also show that external demand (constructed by using export-weighted real GDP growth of trading partners at 2-digit industry level) did not seem to have played a key role.

B. Real Effective Exchange Rate and External Demand

9. Colombia’s strong policy framework combined with the boom in the commodity sector has strengthened Colombia’s exchange rate. Colombia has enjoyed a strong inflow of foreign direct investment over the past several years. These inflows, averaging close to 4 percent of GDP between 2006 and 2012, have allowed Colombia to maintain a balance of payments surplus and to accumulate international reserves. At the same time, Colombia’s real effective exchange rate, especially against the U.S. dollar, has appreciated significantly, by over 30 percent between 2006 and 2012. Appreciation of the Colombian peso has raised a concern that the exchange rate may have been a key factor for the relatively weak performance of the manufacturing sector in recent years.³

10. Colombia’s real exchange rate remained fairly competitive against most major trading partners of manufacturing goods except for the U.S. and Mexico. Despite the real appreciation of the Colombian peso against the U.S. dollar, the appreciation of the bilateral real exchange rate since 2008 against most major trading partners has been fairly modest, with the exception of Mexico. As a result, the real effective exchange rate weighted by export trading partners of manufacturing goods remained relatively stable since 2008 (the spike in 2009 was caused by Venezuela).⁴ The real effective exchange rate weighted by import trading partner of manufactured goods exhibits slightly sharper appreciation since 2008 due to the larger weight of the United States.

Manufacturing Goods Trading Partners

(Average for 2007, 2009 and 2012)

Sources: COMTRADE; and IMF staff calculations.

Manufacturing Goods Trading Partners

(Average for 2007, 2009 and 2012)

Export	Percent share	Import	Percent share
Venezuela	29.9	United States	38.4
United States	19.5	Argentina	9.9
Ecuador	10.6	Brazil	7.7
Peru	6.4	Mexico	7.2
Mexico	3.8	China	6.5
China	3.7	Canada	4.4
Brazil	3.2	Chile	4.4
Belgium	3.0	Ecuador	4.2
Chile	2.6	Germany	3.2
Germany	2.6	Peru	3.0
Italy	2.5	Venezuela	2.4
Japan	2.5	Japan	1.6
Panama	2.1	France	1.5
Spain	1.9	Spain	1.2
Netherlands	1.8	Italy	1.2
Others	4.1	Others	3.1

Sources: COMTRADE; and IMF staff calculations.

Manufacturing Goods Trading Partners

(Average for 2007, 2009 and 2012)

Export	Percent share	Import	Percent share
Venezuela	29.9	United States	38.4
United States	19.5	Argentina	9.9
Ecuador	10.6	Brazil	7.7
Peru	6.4	Mexico	7.2
Mexico	3.8	China	6.5
China	3.7	Canada	4.4
Brazil	3.2	Chile	4.4
Belgium	3.0	Ecuador	4.2
Chile	2.6	Germany	3.2
Germany	2.6	Peru	3.0
Italy	2.5	Venezuela	2.4
Japan	2.5	Japan	1.6
Panama	2.1	France	1.5
Spain	1.9	Spain	1.2
Netherlands	1.8	Italy	1.2
Others	4.1	Others	3.1

Sources: COMTRADE; and IMF staff calculations.

11. External demand for Colombia’s manufacturing goods fell sharply in 2009 and dropped modestly again in 2013. Average real GDP growth weighted by main trading partners of manufacturing exports declined sharply in 2009. Although it recovered and reached 4 percent in 2011 and 2012, the growth rate is still less than the average growth of 6 percent observed between 2004 and 2008 (the weak growth of 2002 and 2003 was caused by Venezuela). The external demand growth fell again to 2 percent in 2013.

C. Structural Changes Related to International Trade

12. Manufacturing exports have declined since 2008. Manufacturing sector seems to be exporting less and producing more for domestic consumption today compared to several years ago. Manufacturing exports in percent of GDP has declined from around 40 percent in 2008 to 26 percent in 2013. Similarly, manufacturing exports in percent of total exports have declined from close to 40 percent in mid-2000s to roughly 20 percent in 2013.

13. Significant portion of the decline in manufacturing exports in recent years seems to be related to Venezuela. According to COMTRADE database, manufacturing exports (excluding food, beverages and tobacco products) to Venezuela declined from US$4 billion (35 percent of the total) in 2007 to US$1 billion (13 percent of the total) in 2012.⁵ The sharp decline largely reflects the deterioration in Colombia’s trade relationship with Venezuela since 2008. The industries that were particularly affected by the trade disruption with Venezuela include electrical machinery, road vehicles, textiles, and medical and pharmaceutical products.⁶ Since then, however, part of the decline has been offset by an increase in exports to other countries in the region, including Brazil, Ecuador and Peru. Exports to the U.S. have declined from US$1.7 billion in 2007 to US$1 billion in 2012, possibly due to the real appreciation of Colombia’s peso vis-à-vis the U.S. dollar. While significant, the magnitude of decline is small in comparison with the drop in exports caused by the trade disruption with Venezuela.

14. Manufacturing imports have increased significantly since 2009, especially from China and Mexico. According to the COMTRADE database, manufacturing imports from China more than doubled since 2007, from US$3 billion (8 percent of the total) in 2007 to roughly US$8 billion (18 percent of the total) in 2012. Increased imports from China are associated especially with machinery and transport equipment, but also with chemical products, rubber and plastic products, metallic and non-metallic mineral products and textiles. Imports from Mexico also increased during the same period, from US$3 billion in 2007 to US$5 billion in 2012. The increase is mostly associated with machinery and equipment. It is noteworthy that, despite the significant appreciation of the bilateral real exchange rate, the increase in imports from the US was smaller than the increase in imports from China.⁷

15. Increased imports from China appear unrelated to the real exchange rate dynamics. To investigate if the recent increase in imports from China is driven by the real exchange rate appreciation, the bilateral real effective exchange rate vis-à-vis the Chinese currency is plotted against total imports from China. The figure highlights that the recent surge in imports did not accompany steady real appreciation of the Colombian peso vis-à-vis the Chinese currency. Therefore, higher imports are more likely to be associated with the increased access by Colombia to Chinese manufacturing products rather than the exchange rate.

16. Increased imports from Mexico seem to be partly driven by the real exchange rate appreciation. Unlike the case of China, the bilateral real effective exchange rate vis-à-vis the Mexican peso has steadily appreciated since 2008, except in 2013. Accordingly, it seems likely that increased manufacturing imports from Mexico are at least partially driven by exchange rate dynamics, rather than increased access by Colombia to Mexican manufacturing products. The renewal of the Free Trade Agreement with Mexico in 2011 may also have contributed to the higher volume of imports from Mexico after 2011.

D. International Comparison

17. Colombia’s experience does not seem particularly unusual in the international context. In recent years, many countries in the region, including Brazil, Chile, Mexico, Paraguay, Uruguay and Venezuela, have experienced a reduction in the share of the manufacturing sector. Some Asian countries, such as Indonesia, Malaysia, and the Philippines, have also experienced a decline in the share of manufacturing, although Asian countries generally have a higher share of manufacturing in comparison with Latin American countries. In contrast, some Asian countries such as Korea, Taiwan, and Vietnam have experienced very strong growth in the manufacturing sector relative to real GDP.

E. Firm-level Analysis: Model

18. A simple model is developed to shed light on the key relationship between profit growth and REER at the firm level. Consider a simple short-run model where firms have only one input of production: intermediate inputs, m_t. For simplicity, it is assumed that firms sell the pre-determined fraction s of the output f (m_t) to the domestic market at the price $p_{t}^{d}$ and export the rest (1 - s) to the world market at the price $p_{t}^{w}$ and the nominal exchange rate, E_t. The profit maximizing firms decide the optimal level of output by choosing $m_{t}^{*}$ that would solve the following problem:

\begin{matrix} \max_{m_{t}} p_{t}^{d} \cdot s \cdot f (m_{t}) + p_{t}^{w} \cdot E_{t} \cdot (1 - s) \cdot f (m_{t}) - c (e_{t}) \cdot m_{t} \\ s . t . p_{t}^{d} = φ (d_{t}, e_{t}) and p_{t}^{w} = λ (d_{t}^{w}) & (1) \end{matrix}

The production function has the standard property (i.e., f ′(m_t) > 0 and f ″(m_t) > 0). The domestic price $p_{t}^{d}$ positively depends on domestic demand d_t as well as real exchange rate e_t, where $e_{t} = E_{t} \cdot p_{t}^{w} / p_{t}^{d}$ . Real appreciation (lower e_t) would drive down the price of imported substitutes in domestic currency, pressuring the domestic price to fall (i.e., $φ_{e}^{'} (\cdot) > 0$ ). The world market price $p_{t}^{w}$ positively depends on the global demand $d_{t}^{w}$ . It is further assumed that all intermediate inputs are produced abroad and their costs depend on real exchange rate. Real depreciation (higher e_t) would increase the costs of intermediate inputs (i.e., c′(e_t) > 0). Profit maximizing firms would choose the optimal level of m_t to equate the marginal increase in revenues from production with the marginal cost.

19. The effect of real deprecation on firms’ profits is given by the following equation:

\begin{matrix} \partial π_{t} / \partial e_{t} = \underset{(+) import substitution}{\underset{︸}{φ_{e}^{'} (d_{t}, e_{t}) \cdot s \cdot f (m_{t}^{*})}} + \underset{(+) higher export revenues}{\underset{︸}{(p_{t}^{d} + φ_{e}^{'} (d_{t}, e_{t}) \cdot e_{t}) \cdot (1 - s) \cdot f (m_{t}^{*})}} - \underset{(+) higher costs}{\underset{︸}{c' (e_{t}) \cdot m_{t}^{*}}} & (2) \end{matrix}

The first term explains the rise in profits associated with import substitution. Real depreciation would raise the price of imported substitutes, making domestic products more profitable in the domestic market. The second term explains the rise in profits associated with higher export revenue through real depreciation. The last term explains the higher costs of imported intermediate inputs.

For firms that only produce for domestic market (i.e., s = 1), the effect of real depreciation is given by the following equation:

\begin{matrix} \partial π_{t} / \partial e_{t} = \underset{(+) import substitution}{\underset{︸}{φ_{e}^{'} (d_{t}, e_{t}) \cdot f (m_{t}^{*})}} - \underset{(+) higher costs}{\underset{︸}{c' (e_{t}) \cdot m_{t}}} & (3) \end{matrix}

For these firms, the effect of real deprecation is negative if the pass-through of real exchange rate to domestic price is low (i.e., $φ_{e}^{'} (\cdot)$ is small).

F. Firm-level Analysis: Empirical Evidence

20. A firm-level database is used to investigate the fundamental relationship between firms’ performance and certain key economic variables. Colombia’s Superintendence of Corporations collects a large amount of data on financial and income statements from private corporations that satisfy certain threshold asset levels, and make some of this information publicly available.⁸ The database used for this analysis is an annual unbalanced panel data from 2000 to 2012 for about 4,850 manufacturing firms. The database does not cover large publicly listed companies that are generally most export intensive.⁹ In the subsequent analysis, the firm-level data was complemented with various industry and aggregate time series data.

21. The following regression specification is used to investigate the factors affecting the profitability of manufacturing firms:

\begin{matrix} {\begin{matrix} p r o f i t G r o w t h \end{matrix}}_{i, k, t} \\ = & α + β_{1} {R E E R}_{k, t} + β_{2} E x t D_{k, t} + β_{3} {D o m D}_{t} + β_{4} R_{t} + {S t r u c t u r a l t r a d e d u m m i e s}_{k, t} \\ + & u_{i, k, t} \end{matrix}

The regressions include firm-level fixed effects.

Prof it Growth_i,k,t : The real growth of profits of firm i that belongs to industry k in year t. The operational profits are deflated by manufacturing producer price index (PPI).
REER_k,t : The growth of export-weighted real effective exchange rate (REER) of industry k in year t. Industry k is classified at the 2-digit level of ISIC (International Standard Industrial Classification).¹⁰
ExtD_k,t : Export-weighted real GDP growth of trading partners for industry k in year t.
DomD_t : Colombia’s real GDP growth in year t.
R_t: Colombia’s interest rate in year t.¹¹
Structural trade dummies_k,t : Structural trade dummies include the trade dummy with Venezuela (which takes the value of 1 if industry k faced reduced exports to Venezuela after 2008), and the trade dummy with China (which takes the value of 1 if industry k faced increased competition from China after 2008).¹²

In the second regression, U.S. real GDP growth is added. In the third regression, we also included the growth of import-weighted REER for each industry in order to investigate the import substitution effects through REER. The observations with very large real growth of profits (300 percent in absolute terms) are excluded from the regressions, as those large movements are unlikely to be explained by the simple specification given above.¹³

22. The coefficient on REER is significant, but has an unexpected sign (Table 1). If real appreciation has contributed to the contraction of the manufacturing industry through reduced competitiveness, one would expect a positive coefficient on REER (i.e., the depreciation promotes profitability growth). However, the negative sign in the regressions indicates that real appreciation increases manufacturing firms’ profit growth on average. In fact, real profits grew by 0.3 percent on average in response to the appreciation of REER by 1 percent. This relationship may be caused by the presence of imported intermediate inputs, to the extent that export-weighted REER (at ISIC 2-digit level) is correlated with the REER for intermediate inputs.¹⁴

Table 1.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (A)

Notes:Standard errors in parentheses.*** p<0.01, ** p<0.05, * p<0.1All regressions include firm-level fixed effects. The observations with very large real growth of profits (above 300 percent in absolute terms) are excluded.

Table 1.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (A)

Dependent variable: Profit Growth
	(1)	(2)	(3)
REER (export-weighted)	−0.339***	−0.342***
	(0.0597)	(0.0637)
External Demand	0.0513	0.0673	−0.228
	(0.240)	(0.266)	(0.259)
Domestic Demand	2.731***	2.737***	3.368***
	(0.442)	(0.444)	(0.452)
R	1.100***	1.081***	0.773***
	(0.306)	(0.334)	(0.287)
U.S. growth		−0.0665
		(0.484)
Dummy (Venezuela)	−12.45***	−12.53***	−13.80***
	(2.401)	(2.458)	(2.376)
Dummy (China)	−6.760***	−6.855***	−7.571***
	(1.902)	(2.024)	(1.884)
REER (import-weighted)			−0.436***
			(0.0811)
Constant	−6.771**	−6.605**	−6.846**
	(3.106)	(3.333)	(3.120)
Observations	29,699	29,699	29,699
R-squared	0.147	0.147	0.147

Table 1.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (A)

Dependent variable: Profit Growth
	(1)	(2)	(3)
REER (export-weighted)	−0.339***	−0.342***
	(0.0597)	(0.0637)
External Demand	0.0513	0.0673	−0.228
	(0.240)	(0.266)	(0.259)
Domestic Demand	2.731***	2.737***	3.368***
	(0.442)	(0.444)	(0.452)
R	1.100***	1.081***	0.773***
	(0.306)	(0.334)	(0.287)
U.S. growth		−0.0665
		(0.484)
Dummy (Venezuela)	−12.45***	−12.53***	−13.80***
	(2.401)	(2.458)	(2.376)
Dummy (China)	−6.760***	−6.855***	−7.571***
	(1.902)	(2.024)	(1.884)
REER (import-weighted)			−0.436***
			(0.0811)
Constant	−6.771**	−6.605**	−6.846**
	(3.106)	(3.333)	(3.120)
Observations	29,699	29,699	29,699
R-squared	0.147	0.147	0.147

23. The effect of domestic demand is positive and significant. The results of the first regression show that on average 1 percent increase in Colombia’s real GDP increases profit growth by 2.7 percent. The results also show that the effect of external demand is generally positive, but weak as the coefficients are not significant. The weak correlation may be caused by the construction of the external demand index as the weights are not variable over time.¹⁵ Surprisingly, the effect of the domestic interest rate is positive and significant. Given that the results may be driven by certain omitted variables, the growth rate of U.S. real GDP is included in the second regression. However, the coefficient on the interest rate remained positive and significant.

24. Structural changes in trade seem to have affected firms’ profit growth. The structural trade dummies both have negative and significant coefficients. Based on the first regression, the profit growth after 2008 was on average 12 percent lower for those industries that were affected by trade disruption with Venezuela. Similarly, the profit growth after 2008 was, on average 8 percent lower for those industries that faced higher competition from Chinese imports.

25. There is no strong evidence that import-weighted REER reduced manufacturing firms’ profits. In the third regression, import-weighted REER was used in the regression. If the profitability of manufacturing firms has fallen as a result of competition with cheaper imported substitutes (due to real appreciation), the coefficient on this variable should be positive. However, the coefficient is negative and significant. Therefore, we cannot conclude that REER appreciation has reduced manufacturing firms’ profitability via import substitution channel.

26. External revenues responded positively to REER depreciation and external demand. The same regressions were run by using real external revenue growth as a dependent variable (Table 2).¹⁶ The results show the expected positive relationship between REER and external revenue growth: real external revenue grew by 0.3–0.5 percent in response to 1 percent REER depreciation. Similarly, real external revenue grew by roughly 3 percent in response to 1 percent growth in external demand. To investigate why REER depreciation affects profit growth negatively despite its positive impact on external revenues, a simple regression was run just using external revenue growth and domestic demand growth (Table 3). The results show that external revenue growth affects profit growth positively, and the coefficient is significant, but the magnitude is small. As a result, the positive impact of REER depreciation through higher external revenue may be dwarfed by negative effects through other channels, such as the higher cost of imported material inputs.

Table 2.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (B)

Notes:Standard errors in parentheses.*** p<0.01, ** p<0.05, * p<0.1All regressions include firm-level fixed effects. The observations with very large real growth of external revenues (above 1000 percent) are excluded.

Table 2.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (B)

Dependent variable: External Revenue Growth
	(1)	(2)	(3)
REER (export-weighted)	0.289**	0.535***
	(0.144)	(0.155)
External Demand	2.919***	2.210***	3.549***
	(0.502)	(0.530)	(0.582)
Domestic Demand	−2.477**	−3.433***	−3.495***
	(1.039)	(1.063)	(1.114)
R	−0.499	0.476	−0.505
	(0.851)	(0.881)	(0.797)
U.S. growth		4.689***
		(1.119)
Dummy (Venezuela)	−23.27***	−17.49***	−21.89***
	(5.407)	(5.575)	(5.314)
Dummy (China)	−14.48***	−6.938	−13.63***
	(4.323)	(4.679)	(4.248)
REER (import-weighted)			0.588***
			(0.205)
Constant	25.97***	16.36*	28.81***
	(8.233)	(8.538)	(8.276)
Observations	10,509	10,509	10,509
R-squared	0.200	0.202	0.200

Table 2.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (B)

Dependent variable: External Revenue Growth
	(1)	(2)	(3)
REER (export-weighted)	0.289**	0.535***
	(0.144)	(0.155)
External Demand	2.919***	2.210***	3.549***
	(0.502)	(0.530)	(0.582)
Domestic Demand	−2.477**	−3.433***	−3.495***
	(1.039)	(1.063)	(1.114)
R	−0.499	0.476	−0.505
	(0.851)	(0.881)	(0.797)
U.S. growth		4.689***
		(1.119)
Dummy (Venezuela)	−23.27***	−17.49***	−21.89***
	(5.407)	(5.575)	(5.314)
Dummy (China)	−14.48***	−6.938	−13.63***
	(4.323)	(4.679)	(4.248)
REER (import-weighted)			0.588***
			(0.205)
Constant	25.97***	16.36*	28.81***
	(8.233)	(8.538)	(8.276)
Observations	10,509	10,509	10,509
R-squared	0.200	0.202	0.200

Table 3.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (C)

(Includes Only Firms with Positive External Revenues)

Table 3.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (C)

(Includes Only Firms with Positive External Revenues)

Dependent variable: Profit Growth
Domestic Demand Growth	3.737***
	(0.517)
External Revenue Growth	0.050***
	(0.00816)
Constant	−9.794***
	(2.476)
Observations	9,424
R-squared	0.196

Table 3.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (C)

(Includes Only Firms with Positive External Revenues)

Dependent variable: Profit Growth
Domestic Demand Growth	3.737***
	(0.517)
External Revenue Growth	0.050***
	(0.00816)
Constant	−9.794***
	(2.476)
Observations	9,424
R-squared	0.196

27. Companies with higher share of external revenues suffered greater losses with REER appreciation. Even though REER appreciation does not seem to have reduced the profit growth of Colombian manufacturing companies on average, it could have prevented the growth of export oriented companies. To investigate this issue, the export share (i.e., the average share of external revenues in total revenues) for each firm was interacted with REER (Table 4). The positive and significant coefficients confirm that the benefits of REER depreciation are larger for firms that are more export oriented. However, the magnitude of the coefficient is relatively small, suggesting once again that the impact of REER on firms’ profitability through this channel is rather limited.

Table 4.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (D)

Table 4.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (D)

Dependent variable: Profit Growth
	(1)	(2)	(3)
REER (export-weighted)	−0.350***	−0.353***
	(0.0598)	(0.0639)
External Demand	0.0527	0.0679	−0.224
	(0.240)	(0.266)	(0.259)
Domestic Demand	2.728***	2.734***	3.364***
	(0.442)	(0.444)	(0.452)
R	1.097***	1.079***	0.768***
	(0.306)	(0.334)	(0.287)
U.S. growth		−0.0635
		(0.484)
Dummy (Venezuela)	−12.47***	−12.54***	−13.82***
	(2.401)	(2.458)	(2.376)
Dummy (China)	−6.770***	−6.861***	−7.587***
	(1.902)	(2.024)	(1.884)
Ex Share * REER (export-weighted)	0.000926**	0.000926**
	(0.000378)	(0.000378)
REER (import-weighted)			−0.448***
			(0.0816)
Ex Share * REER (import-weighted)			0.00107
			(0.000775)
Constant	−6.737**	−6.578**	−6.795**
	(3.105)	(3.333)	(3.121)
Observations	29,699	29,699	29,699
R-squared	0.148	0.148	0.147

Table 4.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (D)

Dependent variable: Profit Growth
	(1)	(2)	(3)
REER (export-weighted)	−0.350***	−0.353***
	(0.0598)	(0.0639)
External Demand	0.0527	0.0679	−0.224
	(0.240)	(0.266)	(0.259)
Domestic Demand	2.728***	2.734***	3.364***
	(0.442)	(0.444)	(0.452)
R	1.097***	1.079***	0.768***
	(0.306)	(0.334)	(0.287)
U.S. growth		−0.0635
		(0.484)
Dummy (Venezuela)	−12.47***	−12.54***	−13.82***
	(2.401)	(2.458)	(2.376)
Dummy (China)	−6.770***	−6.861***	−7.587***
	(1.902)	(2.024)	(1.884)
Ex Share * REER (export-weighted)	0.000926**	0.000926**
	(0.000378)	(0.000378)
REER (import-weighted)			−0.448***
			(0.0816)
Ex Share * REER (import-weighted)			0.00107
			(0.000775)
Constant	−6.737**	−6.578**	−6.795**
	(3.105)	(3.333)	(3.121)
Observations	29,699	29,699	29,699
R-squared	0.148	0.148	0.147

28. The results hold for most industries, although there is some heterogeneity across industries.¹⁷ For most industries, the coefficient of REER is negative. For some industries—including textiles, publishing and printing, rubber and plastic products, fabricated metals, machinery and equipment, motor vehicles, and furniture—the coefficient on REER is negative and significant. The coefficient on REER is positive for some industries, but insignificant in almost all cases. The only industry with a positive and significant REER coefficient is manufacturers of basic metals.

29. Policies to improve the competitiveness of the manufacturing sector should aim at productivity enhancement. A firm level analysis using a large number of manufacturing companies found that much of the weakness in manufacturing performance in recent years was unrelated to the developments of REER. At the same time, manufacturing firms have become more domestically oriented, partly due to structural changes induced by international trade. While real exchange rate depreciation would certainly benefit export oriented manufacturing firms, policies to improve the overall competitiveness of the manufacturing sector should focus on productivity enhancing measures, including labor market reforms and infrastructure improvements.

References

Carranza, J. E., A. Gonzalez, and N. Serna, 2014, “La relación entre la producción y el comercio exterior de la industria manufacturera colombiana,” Borradores de Economía, no. 806, Central Bank of Colombia.
- Search Google Scholar
- Export Citation
Carranza, J. E., A. Gonzalez, and N. Serna, 2013, “Breve Análisis de los Efectos de la Demanda Venezolana sobre el Desempeño de la Industria Colombiana,” Central Bank of Colombia.
- Search Google Scholar
- Export Citation
Carranza, J. E., A. Gonzalez, S. Moreno, and N. Serna, 2013, “Reporte sobre la coyuntura industrial (enero–mayo, 2013),” Central Bank of Colombia.
- Search Google Scholar
- Export Citation
Carranza, J. E., A. Gonzalez, S. Moreno, and N. Serna, 2013, “Analysis de la Heterogeneidad de la Industria Colombiana,” Central Bank of Colombia.
- Search Google Scholar
- Export Citation
Clavijo, S., R. A. Fandiño, and A. Vera, 2013, “Deindustrialization of Colombia: Quantitative Analysis of Determinants,” SSRN Working Paper Series No. 2362369.
- Search Google Scholar
- Export Citation
Frankel, J., 2010. “The Natural Resource Curse: A Survey,” NBER Working Paper No. 15836, (Boston: National Bureau of Economic Research).
- Search Google Scholar
- Export Citation

Prepared by Naomi Griffin, in collaboration with the Colombia’ s Superintendence of Corporations (Superintendencia de Sociedades).

The Superintendence of Companies exercises inspection, monitoring and control of commercial companies, branches of foreign companies and sole proprietorships that are not listed in stock exchange (the corporations that are listed in stock exchange are supervised by other superintendents) and whose assets exceeds 500 times the minimum monthly legal wage. The minimum monthly legal wage is set each year by the government and was 589,500 pesos (US$315) in 2013. As a result of this criterion, some corporations whose assets are close to the threshold may leave the database one year (if the assets fall below the threshold) and enter again the following year (if the assets rise above the threshold).

Economic literature has pointed out that resource rich countries often experience a phenomenon known as the “Dutch disease,” characterized by the loss of competitiveness in non-commodity producing sectors as a result of an overvalued exchange rate. For a comprehensive survey of this literature, see Frankel (2010). Clavijo, Fandiño and Vera (2014) find a support for the Dutch disease hypothesis during the 1970–2010 period using the VEC model.

The weights were calculated using United Nations COMTRADE database for 2007, 2009, and 2012, the years that COMTRADE database seems to have the most comprehensive coverage on Colombia. For this calculation, we included SITC industry code 0 (food and live animals), 5 (chemicals and related products), 6 (Manufactured goods chiefly classified by material), and 7 (machinery and transport equipment).

Even though COMTRADE database seems to have the most comprehensive coverage on Colombia for 2007, 2009 and 2012, there could be some discrepancies between the official data due to some unreported categories. In particular, the data for 2012 seems incomplete for a small number of manufacturing categories.

The research by Carranza, González and Serna (2014) shows that the manufacturing industries that had high level of exports Venezuela were, on average, relatively poor performers even before the trade disruption. In contrast, manufacturing industries that exported to other countries have demonstrated robust growth in recent years.

Based on COMTRADE database, manufacturing imports (excluding food, beverages and tobacco products) from the U.S. increased from $6.7 billion in 2007 to $8.2 billion in 2012.

See footnote 2 for more information.

The omission of these large firms may generate biases in the results, although the number of listed manufacturing companies is fairly small to date (i.e., less than 10 as of April 2014).

Export weights are calculated using COMTRADE database. Since the trade data from COMTRADE were based on SITC (Standard International Trade Classification), while the firm-level manufacturing database was based on ISIC (International Standard Industrial Classification), the industry codes were mapped at 2-digit level.

The central bank’ s policy rate (overnight lending rate) is used.

The Venezuela dummy includes apparel, leather products, electrical machinery, road vehicles, furniture and medical and pharmaceutical products. The China dummy includes textiles, chemicals and chemical products, rubber and plastic products, basic metals, metallic and non-metallic mineral products, and machinery and transport equipment.

Such observations consist roughly 9 percent of available sample observations. As a robustness check, quantile regressions—which are more robust to outliers—were run with all observations using similar specifications. The results of quantile regressions are generally similar.

Ideally, we would like to separate the effect of the imported intermediate inputs from the effect of export revenues or the effect of import substitution, as described in the previous section. Unfortunately, this was not possible due to the lack of data on imported inputs and the corresponding exchange rate.

Export destinations could change in response to the demand growth of the external markets. Unfortunately, missing observations made it difficult to construct the variable weights at the 2-digit level.

External revenues reported to the Superintendence of Corporations may include revenue unrelated to exports, and could be subject to some measurement errors.

For detailed information on the heterogeneity of performance across industries, see the research by Carranza, González, Moreno and Serna (2013) and Carranza, González and Serna (2014).

Other Publishers

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://www.elibrary.imf.org/view/journals/002/2014/167/article-A003-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Colombia: Selected Issues Paper

Author: International Monetary Fund. Western Hemisphere Dept.

Volume/Issue:: Volume 2014: Issue 167

Publisher:: International Monetary Fund

ISBN:: 9781498302128

ISSN:: 1934-7685

Pages:: 86

DOI:: https://doi.org/10.5089/9781498302128.002

Keywords
Recent Performance of Colombia’s Manufacturing Sector: Exchange Rate or Structural?1
References

View raw image

Manufacturing Sector

(In percent of real GDP)

View raw image

Growth of Real GDP and Manufacturing Sector

(y/y percent change)

View raw image

Real Effective Exchange Rate

(2000=100, + depreciation)

View raw image

Bilateral Real Exchange Rate

(2000=100, + depreciation)

View raw image

External Demand Growth for Manufacturing Sector

(Export weighted real GDP growth of trading partners)

View raw image

Manufacturing Exports

(In percent of manufacturing GDP, 4-quarter MA)

View raw image

Manufacturing Exports

(In percent of total exports)

View raw image

Manufacturing Exports by Trading Partners

(In Billions of US$)

View raw image

Selected Manufacturing Exports to Venezuela

(In Billions of US$)

View raw image

Manufacturing Imports by Trading Partners

(In Billions of US$)

View raw image

Selected Manufacturing Imports from China

(In Billions of US$)

View raw image

Total Imports from China

(In billions of US$, REER + depreciation)

View raw image

Total Imports from Mexico

(In billions of US$, REER + depreciation)

View raw image

Manufacturing Sector: Asian Countries

(In percent of real GDP)

View raw image

Manufacturing Sector: Latin American Countries

(In percent of real GDP ¹/)

Foresee

Table 4.

OLS Panel Regression Results for Manufacturing Firms, 2000–2012 (D)

Dependent variable: Profit Growth
	(1)	(2)	(3)
REER (export-weighted)	−0.350***	−0.353***
	(0.0598)	(0.0639)
External Demand	0.0527	0.0679	−0.224
	(0.240)	(0.266)	(0.259)
Domestic Demand	2.728***	2.734***	3.364***
	(0.442)	(0.444)	(0.452)
R	1.097***	1.079***	0.768***
	(0.306)	(0.334)	(0.287)
U.S. growth		−0.0635
		(0.484)
Dummy (Venezuela)	−12.47***	−12.54***	−13.82***
	(2.401)	(2.458)	(2.376)
Dummy (China)	−6.770***	−6.861***	−7.587***
	(1.902)	(2.024)	(1.884)
Ex Share * REER (export-weighted)	0.000926**	0.000926**
	(0.000378)	(0.000378)
REER (import-weighted)			−0.448***
			(0.0816)
Ex Share * REER (import-weighted)			0.00107
			(0.000775)
Constant	−6.737**	−6.578**	−6.795**
	(3.105)	(3.333)	(3.121)
Observations	29,699	29,699	29,699
R-squared	0.148	0.148	0.147

Topics

Countries

Series

About

Resources

Abstract

Recent Performance of Colombia’s Manufacturing Sector: Exchange Rate or Structural?1

A. Introduction

B. Real Effective Exchange Rate and External Demand

C. Structural Changes Related to International Trade

D. International Comparison

E. Firm-level Analysis: Model

F. Firm-level Analysis: Empirical Evidence

References

Same Series

Other IMF Content

Other Publishers

Asian Development Bank

Inter-American Development Bank

International Labour Organization

The World Bank

Share Link

Table of Contents

Headings

Figures

Metrics

Foresee

Recent Performance of Colombia’s Manufacturing Sector: Exchange Rate or Structural?¹