@article { author = {Nasersadrabadi, Seyedeh Marveh and Ghaffari, Farhad and Mohammadi, Teymour and Memarnejad, Abbas}, title = {The Effects of Global Financial Crises on the Trade Patterns of Iran and its Partners: Pseudo Poisson MLE Method}, journal = {Iranian Journal of Economic Research}, volume = {28}, number = {96}, pages = {87-121}, year = {2023}, publisher = {Allameh Tabataba’i University}, issn = {1726-0728}, eissn = {2476-6445}, doi = {10.22054/ijer.2021.61543.999}, abstract = {The negative consequences of financial crises require the attention of economic policymakers and decision making centers.Therefore, considering the importance of the subject, the present study has investigated the effects of global financial crises on the trade patterns of Iran and its partners during the years 1995-2018.The variables have been estimated in the framework of the gravity model using the pseudo poisson maximum likelihood method.The findings show that the Asian financial crisis (1997) has an effective role in reducing the volume of trade but this result is the opposite in relation to the American financial crisis (2007)­; Because instead of a threat, it has become an opportunity for the movement of business flows. In this situation, it seems that the difference in the intensity and type of impact of financial crises on trade patterns can be affected by the nature of the crisis or the region where the crisis started.1.IntroductionCountries consistently grapple with economic and financial turmoil, which at specific junctures, can escalate into full-fledged financial crises (Moshiri & Nadali, 2013). These crises manifest as conditions where a significant number of financial institutions suddenly experience a substantial decline in the nominal value of their financial assets (Bonis et al., 1998). Given the historical recurrence of financial crises worldwide, it becomes imperative for economic policymakers and decision making centers to address and mitigate their adverse effects. This necessity stems from the detrimental impact that financial crises have on the real sector of economy (Kord-Zangeneh et al., 2019). Compounding the issue is the transnational nature of these crises, as they may transfer from one country to another. The transmission of financial crises occurs through various channels, including trade flows, foreign direct investment, commercial loans and financial aid (Massa & Velde, 2008). Among these channels, trade relations emerge as crucial communication pathways on a global scale, playing a pivotal role in influencing the performance of diverse economic sectors affected by financial crises.Hence, understanding international trade patterns holds greater significance than any other phenomenon in the economy, particularly in times of crisis. Furthermore, drawing insights from the experiences of other nations aids in understanding how trade flows unfold in countries that have recently weathered financial crises (Santana-Gallego & Perez-Rodriguez, 2018). Financial crises impact on the trade in two ways. First, they exert a negative influence on trade by disrupting the trade balance. Then crises transfer from one affected country to another through interconnected trade links. Consequently, the extent to which different countries engage with the global economy dictates the degree to which they are affected by the repercussions of a financial crisis.As countries are interlinked through trade flows, in the event of a shock impacting one economy, it has the potential to extend to the entire network, indirectly influencing trade relations between countries. This connection is particularly crucial because a financial crisis can transfer to other economic sectors through fluctuations in exchange rate variables, exports, imports and changes in international commodity prices (Brave & Butters, 2011).2.Materials and MethodsThe gravity model, proposed by Tinbergen (1966) to explain bilateral trade flows, is distinctive for its emphasis on reflecting international relations. In the field of international trade studies, traditional challenges arise in estimating the gravity model. Specifically, when employing the ordinary least squares method for estimating the gravity model, there is a tendency to exclude zero statistical observations. This limitation stems from the conventional method’s inability to compute a logarithm for the trade variable when trade between countries is not realized in certain years. Consequently, the omission of statistical observations in such instances renders it impossible to generate a zero logarithm. Moreover, when the model is estimated using the non-linear least squares method, there is a potential issue with the heterogeneity of variance, which can compromise the accuracy of interpretations based on the coefficients. Recognizing this challenge, Santos-Silva and Tenreyro (2006) introduced the Poisson pseudo maximum likelihood method to address the estimation of such models. A noteworthy aspect of this method is the non-elimination of zero statistical observations, ensuring unbiased and reliable estimation of variable coefficients. This is achieved by assigning equal weight to all statistical observations. Therefore, the method not only increases the number of statistical observations but also enhances the efficiency of the estimator.The main estimation approach revolves around the Poisson pseudo maximum likelihood method, as explained by theoretical foundations and existing literature that detail the connection between financial crises and international trade. This approach draws inspiration from the works of Santos-Silva and Tenreyro (2006) as well as of Santana-Gallego and Perez-Rodriguez (2018). In this respect, the research model was delineated following the model proposed by Glick and Rose (2016), as shown in Equation (1).  (1) The present study aimed to investigate trade patterns by examining the volume of bilateral trade (total export and import) between Iran and its twenty trading partners. The analysis used annual data spanning from 1995 to 2018. The study developed the proposed model, leveraging the flexibility inherent in the gravity model as well as incorporating variables such as the logarithm of the Linder economic similarity index, the logarithm of Iran’s and its trading partners’ populations, the logarithm of the nominal exchange rate, the logarithm of geographical distance and financial crises. This comprehensive formulation is defined as the generalized gravity model expressed in Equation (2).  (2)  3.Results and DiscussionDiagnostic tests are imperative before model estimation. Initially, the Chow test was employed to determine the suitable regression method. Subsequently, the Hausman test was used to decide between the methods of fixed effects or random effects.Table 1. Diagnostic testsResultProbabilityStatisticsTestNull Hypothesis Rejected0.00023.04ChowNull Hypothesis Rejected0.000437.51HausmanSource: Research findings The results outlined in Table 1 demonstrate the rejection of the null hypothesis in both the Chow and Hausman tests. To control the multilateral resistance to trade, the study estimated the coefficients of the variables by considering the country’s annual fixed effects. The process was conducted within the framework of the gravity model, employing the Poisson pseudo maximum likelihood method (see below).Table 2. Model estimation resultsProbabilityStandard deviationCoefficientsVariables***0.0000.000-0.179CRIijt 1997***0.0000.0000.135CRIijt 2007***0.0000.184-32.358LnLINijt***0.0005.9153.005LnPOPit***0.0000.1000.111LnPOPjt***0.0000.0000.017LnERijt***0.0000.000-0.400LnDISijt***0.0000.113-52.430CNumber of obs = 240R-Squared = 0.81Pseudo Log Likelihood = -171.405*** Indicates the significance of the coefficients at the level of 1 percent.Source: Research findingsThe results provided in Table 2 reveal that global financial crises exerted a significant impact on the trade volume, yet the nature of their influence on trade patterns varies. Specifically, the findings indicated that the Asian financial crisis of 1997 played a substantial role in reducing the trade volume, while the outcome was opposite in the case of the American financial crisis of 2007.The negative coefficient in the logarithm of the Linder economic similarity index indicates that the volume of trade increases as the per capita income difference decreases. Consequently, the countries with similar tastes or demand structures become the optimal markets for a country’s export goods.Conversely, the positive coefficient in the logarithm of the population of Iran and its trading partners signifies that an increase in population correlated with a rise in the trade volume. This association can be attributed to the utilization of a larger labor force, inherent in higher population figures, which positively affects the production of goods. The outcome is manifested in an increase in the trade volume.The positive coefficient in the logarithm of the nominal exchange rate indicates an increase in trade volume corresponding to an increase in this variable. This pattern emerges because foreign goods become more expensive compared to domestic ones. Consequently, both domestic and foreign consumers are inclined to substitute Iranian goods with foreign alternatives. Conversely, the negative coefficient in the logarithm of geographical distance reveals that this variable exerted a negative impact on the trade volume. In other words, the greater the distance between countries, the higher the transportation costs. As a result, distant markets become less attractive for establishing trade relations.4.ConclusionThe present study examined the effects of global financial crises on the trade patterns of Iran in relation its key trading partners. In this respect, the research used annual data from the studied countries during 1995–2018, then the coefficients of the variables were estimated within the framework of the gravity model as well as the Poisson pseudo maximum likelihood method.According to the findings, the examined countries experience the repercussions of financial crises, yet the magnitude and nature of their impact differ based on the specific characteristics of each crisis. In this context, the Asian financial crisis of 1997 played a significant role in reducing the trade volume in the countries under consideration, while the outcome was opposite in the case of the American financial crisis of 2007. Moreover, the positive coefficients of the variables specifically the logarithm of the Linder economic similarity index, the logarithm of the nominal exchange rate and the logarithm of the population of Iran and its trading partners underscore their favorable impact on the trade volume aligned with the increased trade flows in the countries. Given the negative coefficient in the logarithm of geographical distance, it is anticipated that trade with countries farther away from Iran will be comparatively lower. In fact, the majority of Iran’s trade relations are established with neighboring countries. In light of these findings, it is recommended to implement trade policies that support export-oriented domestic production in the country. This approach, in addition to generating foreign currency income, can serve as a mitigating factor against the adverse effects of financial crises.}, keywords = {financial crisis,Trade Patterns,Gravity Model,Poisson Pseudo Maximum Likelihood}, title_fa = {اثرات بحران‌های مالی جهانی بر الگوهای تجاری ایران و شرکای آن: روش شبه MLE پوآسون}, abstract_fa = {پیامدهای منفی بحران­های­­ مالی لزوم توجه سیاست­گذاران اقتصادی و مراکز تصمیم ­ساز را بر­می­انگیزد. از این رو، با توجه به اهمیت موضوع مطالعه حاضر به بررسی بحران­های مالی جهانی و تحلیل آثار آن بر الگوهای تجاری در کشور ایران و شرکای تجاری طی سال­های 2018-1995 پرداخته است. متغیرها در چارچوب مدل جاذبه و با استفاده از روش شبه حداکثر درست­نمایی توزیع پوآسون برآورد شده است.­ یافته­ ها نشان می­دهد بحران­ مالی آسیا (1997) نقش موثری در کاهش حجم تجارت دارد اما این نتیجه در ارتباط با بحران مالی آمریکا (2007) برعکس است­؛ چرا که به جای تهدید، به فرصتی مناسب در راستای تحرک جریان ­های تجاری تبدیل شده است. در این شرایط، به نظر می­رسد تفاوت در شدت و نوع تاثیرگذاری بحران­ های­­ مالی بر الگوهای تجاری می­تواند متاثر از ماهیت بحران­ یا منطقه ­ای باشد که بحران از آنجا آغاز شده است.}, keywords_fa = {بحران‌‌‌ مالی,الگوهای تجاری,مدل جاذبه,شبه حداکثر درست‌نمایی توزیع پوآسون}, url = {https://ijer.atu.ac.ir/article_13216.html}, eprint = {https://ijer.atu.ac.ir/article_13216_c96d4e24b71edd14ef6477bf92be2cf6.pdf} }