Uncertainty measure in international trade
Banque de France
General Description: Part of a project studying how to measure uncertainty in international trade. Your contribution: Write a R program that - 1/ Collects and clean international trade data. - 2/ Use that data to compute measures of uncertainty in international trade that can be updated every month. The measure of uncertainty is in the spirit of Bricongne & Gigout (2020) https://drive.google.com/file/d/15cyQR-VNpg4TCm6nz3WHXxT2LGyngTlY/view?usp=sharing Step 1/ Write a sub-program that downloads monthly trade from comext (the European Commission External trade database) : https://ec.europa.eu/eurostat/estat-navtree-portlet-prod/BulkDownloadListing?sort=1&dir=comext%2FCOMEXT_DATA%2FPRODUCTS Your program should download the data, unzip it and clean it. Your program should offer the ability to download past data from 2017 onward and the ability to download the latest month only. The cleaning steps are outlined at this address http://www.isabellemejean.com/BasicCleaning.do (you may ignore anything related to firm level variables since this level of disaggregation is not available in comext). Finally convert the 8-digit level product codes (nc8) to 6-digit codes (hs6) and aggregate the data at this level. Step 2/ Write a sub-program that takes the output from step 1, applies various transformations to the data then computes various measures of dispersions meant to capture the uncertainty of foreign demand in international trade. Once again, your program should allow the option to execute those operations on the historical data or on the latest month of data only. - 2a/ compute the year-on-year growth rate of the level of trade in euro for each month. Taking into account the so called “zero trade flows” is a key challenge in international trade. Therefore including an option to compute different types of growth rate will be a plus (eg. Delta log, midpoint growth rate, delta log(1+x), etc). Implementing one or two in the programs will suffice as long as you include provisions to easily add others in the future. - 2b/ write a sub-program that takes those growth rates and applies various residualization procedures. This program will need to include at least the following one: o v_ijpt = gamma_ipt + gamma_jpt + upsilon_ijpt o Where the left-hand side variable is the growth rate of trade of product “p” between exporter “I” and importer “j” for month “t”. The “gammas” are respectively exporter-product-month and importer-product-month fixed effects. o The program will need to collect the residual upsilon. You may use any estimation command available on CRAN (fixest or lfe for instance). You will use a linear estimator but provisions to add other estimators in the future (eg. Poisson) need to be included. - 2c/ The program will then compute various moments of the distribution of upsilon. The minimum requirement is the standard deviation of upsilon_ijpt for each importer(i)-hs4(k)-month(t) . Adding options to compute other moments of the distribution (skewness, interquartile range, etc.) would be valuable. At this point, you will have written a program that compute a measure of the uncertainty of the demand originating from each import market (j-k pairs). Step 3/ Various events affected international trade since 2017. Using the data that you generated in step 2, produce a few (no more than 4) stylized facts (graphics or tables) that you think would be interesting. Did some of those events generate trade uncertainty? If any, what sector or country were most affected? The reward will be an interview for an internship in Banque de France. Candidates must be students in France (without being necessarily French though).
Une interview pour un stage rémunéré à la Banque de France (durée de plusieurs mois, à fixer), pour les personnes étudiant en France.
0 €
The winners of this challenge will receive a certificate of participation attesting to their skills.
From 05/10/2021 until 28/02/2022.
Bazinga! composed of :
- Thomas VALADE
Bahamut composed of :
- Mathieu Véron
- BAPTISTE NEPVEUX
Bahamut composed of :
- BAPTISTE NEPVEUX
DASEE students composed of :
- Taj Muhammad Ahmadzai
- Ivan Marin
- PERLA NAHOUM