The Impact Of The Financial Crisis On Trade Economics Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

4.0 Introduction

This chapter will give an insight on the data, the hypotheses and methodology chosen for the purpose of this study. All data have been taken from the World Bank's, World Development Indicator catalogue. The methodology chapter is structured as follows: section 4.1 describes our sample, section 4.2 explores the models and variables used, section 4.3 examines the variables and their priori expectation, section 4.4 provide an explanation of the panel data, section 4.5 discusses panel regression techniques used, section 4.6 explains the Hausman specification test and section 4.7 concludes this chapter.

4.1 Sample countries.

One of the objectives of this study is to compare the effect of the recent financial crisis on trade and trade finance between Asian countries and African countries. In this prospect, 10 countries where chosen, of which 5 are African countries and the rest are Asian countries. These countries are: Bolivia, Kenya, Ghana, Sierra Leone, South Africa, India, China, Russian Federation, Malaysia and Indonesia. Obviously, for analysis more countries would have been ideal, but due to the limitation in the availability of certain variables, only these ten countries were chosen for the years 2001 up till 2012.

4.2 Model Specification

For the purpose of this study, we will be using three models. The first one will test the impact of the financial crisis on trade finance; the other two models will respectively test the impact of the fall in trade finance on the exportation and importation of our sample.

4.2.1 Testing the impact of the financial crisis on trade finance

= The financial crisis significantly affected trade finance.

= The financial crisis did not significantly affect trade finance.

The model used as follows:

TF = (Financial crisis, CreditPG, GDP)

For the purpose of estimation, the above equation is arranged as follows:


i = country

t = year

TF = External short term debt

Financial crisis = Dummy variable [1 for post crisis (2007-2011), 0 for pre crisis (2002-2006)]

CreditPG = Credit to the Private sector (% GDP)

Log GDP = log of Gross domestic product

U = error term

4.2.2 Testing the impact of Trade Flows and Trade Finance using Merchandise Exports and Imports

Model 2.1: Testing Trade Finance and Export

: Trade Finance affected exports.

: Trade Finance did not affect exports.

The model which is used to test for the hypothesis is given below:


By reorganizing the above model into an econometric equation for a better estimation of the 'Y' dependent variable with the 'X' explanatory variables, we obtain the following equation:


Where the variables are listed as follows:

i = country

t = year

Xit= Merchandise Exports volumes

EXCHit = Exchange Rate

LOG TFit =Log of External short term debt

LOG GDPit = Log of Gross Domestic Product

INFLit=Rate of Inflation

CRISISit= Dummy Variable [1 for post crisis (2007-2011), 0 for pre-crisis (2002-2006)]

Uit = Disturbance Term.

Model 2.2: Testing Trade Finance and Imports

: Trade Finance affected imports.

: Trade Finance did not affect imports.

The model which is used to test for the hypothesis is given below:


By reorganizing the above model into an econometric equation, we obtain the following equation:


Where the variables are listed as follows:

i = country

t = year

Mit= Merchandise Imports volumes

LOG TFit = Log of External short term debt

LOG GDPit = Log of Gross Domestic Product

EXCHit = Exchange Rate

INFLit=Rate of Inflation

CRISISit= Dummy Variable [1 for post crisis (2007-2011), 0 for pre-crisis (2002-2006)]

Uit = Disturbance Term.

4.3 Variables and priori expectations.

External short term debt

The External short term debt constitutes of the debt in USD, with maturity of one year or less. In both models, the logarithm of this variable is chosen to measure the level of trade finance. In the first model, the logarithm of the External Short term debt is defined as our independent variable. In the second model, it is used as an explanatory variable, with the expectation of a positive coefficient with respect to trade, because with a higher availability of trade related finance, traders would find it easier to trade their products internationally. However, during financial shake up, due to counterparty risk, trade finance activities have plummeted-thus showing the positive relationship between those two factors.

Credit to the private sector (% GDP)

The variable credit to the private sector constitutes of all the funds provided to the private sector, in terms of "loans, purchases of non-equity securities, and trade credits and other accounts receivable that establish a claim for repayment", and is calculated as a percentage of Gross Domestic Product. Using this ratio as a measure financial development, the higher the ratio is, the higher the contribution of GDP towards financing. Thus, we expect a positive relationship between credit to the private sector (as a percentage of GDP) and trade finance. Moreover, as corroborated by Ellingsen and Vlachos (2009), the financial development in country may determine the effect of financial crisis on trade financing; thereby, the higher is the ratio the higher the effect of the financial crisis will be on trade financing.

Merchandise Exports Volumes

Merchandise Exports show the volumes of goods which are sold to different countries across the globe. The global merchandise exports fall by 32% mid of 2008-where the crisis has caused major havoc in the trading sector. However, due to unwillingness for banks to provide for cross-border lending, this has resulted in a shortage of funds for trade financing. As a result of which, there was a significant decline in the exports. Hence, there is a positive coefficient between trade finance and merchandise exports.

Merchandise Imports Volumes

Merchandise imports are the volumes of goods which are bought by countries from other trading nations. The various trade financing tools were put at the disposal of importers to facilitate payment to the suppliers. Due to the high dependency on those tools, the import volumes have witnessed a downturn trend since the crisis. This was because most banks were reluctant to provide trade financing services to the importers given the high degree of default risk. Henceforth, trade finance and imports are positively correlated.

Gross Domestic Product (GDP)

The GDP is measured in US dollars and is defined as "the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products, at purchaser's price. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources" (World Bank). This variable is chosen to measure the level of demand in the country; the higher GDP is, the higher financial resources in terms of funds will be expected to be. A crisis stricken period is expected to have a negative impact on GDP. On the other hand, a high GDP will imply a high demand for imports and also higher export volumes therefore having a positive coefficient unless the country is highly vulnerable to the crisis.

Inflation Rate

Inflation is defined as the increase in the general price level. It is measured by the Consumer Price Index (CPI).The higher the inflation rate, the lower is the purchasing power to consumers. The effect of the financial crisis has affected the inflation rate through the channel of international price and the world consumer demand. There is indeed a positive relationship between financial crisis and inflation. On the other hand, the higher the inflation rate, the more expensive are the commodity (commodity price rises), and the lower is the export volumes. The reason behind such a result is because the exports are too expensive to compete with the manufacturers producing at low cost. There is a negative relationship between inflation rate and exports.

Exchange Rate

Exchange rate is defined as the rate of one country's currency relative to another country's currency. It has been largely used as an international comparison base mainly for commodity prices and GDP between countries. Exchange rate is highly affected due to changes of capital flows which in turn is the main targeted variable during the crisis. The merchandise exports volumes is highly affected by changes in the exchange rate such that an increase in its value resulting in an appreciation will make exports more expensive. The inverse result is expected in case of a depreciation of the currency. Thus, both the variables have a negative coefficient. In the same context, during a period of financial turmoil, exchange rate is expected to fall thus making imports appears more costly. This illustrates a positive relationship between imports and exchange rate.

Financial crisis

Financial crisis is a dummy variable, used to capture the impact the effect of financial crisis on trade financing. Alone, it is expected to have a negative impact on trade financing due to rise in the volatility in the financial market during a period of shock. This will inevitably cause reluctance among providers of trade finance as well as its users. Subsequently, following a fall in trade financing, it might entails to major trade loss. Thus, there is a positive relationship between trade finance and trade flows.

4.4 Panel Data.

Panel data, also known as longitudinal data, is a combination of time series data [1] and cross-sectional data [2] . A panel data set will contain the element of both time and space. Variables for different entities are measured over time. Importantly, the data, measured over time, should be on the same entities (for instance firms or countries). The advantage of such a data is that more issues (like relationship between variables) and more complex problems can be tackled as compared to a pure time-series data or a pure cross-sectional data, alone.

4.5 Fixed and Random Effect model

The fixed effect regression and the random effect regression are the two main techniques used in the literature to empirically analyze panel data.

The fixed effect regression model analyses the link between the dependent variable and the independent variables within entities having the same characteristics that may (or may not) affect the dependent variable. Econometrically speaking, the fixed effect model is a statistical model that represents the explanatory variables as being non-random observation. The fixed effect analysis assumes that there is something within the entity that may affect or bias the dependent or independent variables, and thus the need to control such effects. Hence, this explains the assumption of correlation between the entity's error term and the independent variable. Fixed effect regression also assumes that although the intercept may differ across entities (through the application of the dummy variable technique), each entity's intercept does not change over time; implying that it is time invariant. This enables us to assess the net effect of the independent variable on the outcome variable. This technique also allows studying the impact of the independent variable on the dependent variable, by looking at the changes in the variables over time.

The general form of a panel regression is as follows:



In order to capture individual country specific effect and anything that is left unexplained on , the disturbance term can be decomposed into an 'individual specific effect', , and the 'remainder disturbance', .


Substituting equation (2) into equation (1), we get


could be thought as constitution of all variables that affect the dependent variable cross-sectionally, but are time-invariant. This model could also be estimated by incorporating dummy variables (either multiplicatively or additionally), which is known as the Least Squares Dummy Variable approach.

"Alternatively, the random effects model, also known as the error components model, suggests that the intercepts for each cross-sectional unit arise from a common intercept α (which is the same for all cross-sectional units and over time), plus a random variable _i that varies cross-sectional but is constant overtime. i measures the random deviation of each entity's intercept term from the 'global' intercept term α" (Chris Brooks, 2008).

4.6 Hausman Specification test.

The Hausman Specification test can be used to determine which regression analysis (the fixed effect and random effect) is appropriately suited to our model. This test evaluates the significance of an estimator compared to an alternative estimator. It allows testing for the presence of correlation between the regressors and the disturbance term.

The following linear model is considered:


y is unvariate

b = vector of coefficients

X = vector of regressors

e = error term

There are two estimators for b: b0 and b1. Under the null hypothesis, H0, b0 and b1 are both consistent but, b1 is efficient, at least for the category of estimators containing b0. However, under the alternative hypothesis, H1, b0 is inconsistent and b1 is consistent. The Hausman Specification test statistic is:


where † represent the Moore-Penrose pseudo inverse. Asymptotically, the statistic follows the chi-square distribution, with the number of degrees of freedom equal to the rank of matrix Var(b0) − Var(b1).

Atleast one of the estimators will be inconsistent if the null hypothesis id rejected.

The null and alternative hypotheses are defined as follows:

H0 : λ = 0(RE specification is acceptable)

H1: λ ≠0 (RE specification is invalid: FE should be used)

A P-value ( of less than 0.05, implies that it is significant, and the fixed effect regression is appropriate for our model. If the P-value is greater than 0.05, then the alternative estimator, that is the random effect, must be used.

4.7 Conclusion

Using a sample of 10 countries (5 Asian and 5 African), we test the impact of the financial crisis on trade financing, and its effect on trade. A panel regression technique is applied for this purpose. The analysis of the result obtain is reported in the next chapter.