The fight against illicit financial flows (IFFs), in particular against those originating from trade misinvoicing in commodity exports, has become especially relevant for resource-rich countries focused on strengthening their domestic resource mobilization (DRM) strategies. In an effort to promote evidence-based and effective policymaking, studies on IFFs are facing diverse challenges, particularly over the conceptual framework and empirical measurement techniques. In this article, Alessandra Rojas, Master in Development Studies from the Graduate Institute of International and Development Studies, discusses the main learnings from her research on the export of Peruvian copper concentrate.
The case of Peru
As a recognized player in the global mining industry, Peru represents an interesting case for the study of illicit flows. In 2018, Peru ranked second in the global production and export of copper, silver and zinc, and it was the leading producer of gold, zinc, lead and tin in Latin America (Ministerio de Energía y Minas, 2018). Overall, the sector contributes up to 13.9% to the country’s GDP and up to 62% to the total value of Peruvian exports (Ministerio de Energía y Minas, 2018).
The study aimed to assess the existence of trade mispricing in the export of Peruvian copper concentrate for the years 2003 to 2017. Following current advances in measurement techniques, I conducted a quantitative analysis of abnormal pricing by applying the price-filter methodology to compare declared transaction-level prices from customs records against the LME daily price series for copper, selected as the free market benchmark price. The analysis was further supported by qualitative data that I collected from interviews with key stakeholders of the Peruvian mining sector, which informed the selection of the relevant price filters, as well as helped at understanding the market complexities that can be reflected in the data. The price filters selected that determined the acceptable magnitude of price deviations around the LME benchmark price were based on the following four criteria as derived by qualitative research: copper content, payable byproducts, penalties and commercial terms.
The research found asymmetries between declared transaction prices and expected export prices across the data, which varied heavily depending on the selected price filter. I conducted the analysis using four different price filters to show the impact of the assumptions in the final estimations:
- For a price filter of -50%, the estimated magnitude of undervalued exports would reach US$ 81 billion, which equals to 98% of the total exported value for the period 2003-2017.
- For a price filter of -70%, the undervalued exports reached US$ 61 billion, which equals to 75% of the total exported value.
- In contrast, when considering a price filter of -80%, the estimated magnitude of undervalued exports now reached US$ 11 billion, representing a 13% of the total exported value.
- Lastly, a price filter of -90% resulted in undervalued exports of US$ 67 million, which equals to merely 0.1% of the total export value for the analyzed period.
Although the selection of such high filters might seem counterintuitive, the effect that the above mentioned criteria have on the final export price of the concentrates must be fully weighted and understood. For example, when considering the copper content, it is crucial to note that the LME benchmark price is based on refined copper, whereas per definition, concentrates contain different grades of copper, usually between 18% to 34%, which also varies from mine to mine. Therefore, the copper grade used to calculate the final transaction price, has a negative impact on the observed transaction price that could deviate up to 80% from the LME benchmark price for refined copper. Such considerations are taken to estimate the effect the four different criteria have on the declared transaction price and thus build the suggested price filter.
Depending on the selected price filters, results might suggest that a significant proportion of Peruvian copper concentrate exports are abnormally undervalued. In turn, I analyzed the export transactions for the year 2017, disaggregating the data by exporting firm and average copper grade and thus comparing transaction prices across exporting destination within firms. The results showed that trade asymmetries are driven by a combination of factors not all of which can be attributed to price manipulation, as explained by different hypotheses informed by the qualitative interviews. This in-depth assessment of results showed the added value of further data disaggregation: initial estimates of trade mispricing are not to be taken as final straightforward evidence of illicit financial flows, but instead as valuable indications of asymmetries that need to be further researched.
Additionally, the study highlighted the following recommendations:
- More country-level studies that use disaggregated, micro-level data on specific commodities are needed: Aggregated analyses of illicit flows at both an international or national level might result in inaccurate estimates of the magnitude of illicit flows and of its channels. Previous studies that included estimates for Peruvian commodity exports did not provide indications of the affected commodities or a detailed analysis of the source of the outflows. Additionally, testing the research approach across diverse countries’ datasets and research contexts allows methodological improvements.
- Statistical analyses must be backed by context specificity: This is especially relevant when using the price filter methodology, since results can vary strongly depending on the selection of the price filters, as shown in the estimates of this study. Therefore, an understanding of local commercial practices, recording procedures and trade partner relations, gained through qualitative data collection, is crucial for discussing observed asymmetries to avoid automatically attributing all statistical discrepancies to illicit flows.
- Empirical measurement techniques for estimating trade mispricing in commodity exports must account for the commodity characteristics: When working with a commodity such as copper concentrate, where its transaction price is affected by multiple components, the methodology used must be adapted to consider such data complexities. The composition of prices must be well understood, otherwise a direct comparison between benchmark and transaction-level prices will not be based on the same premise resulting in an over- or underestimation of results.
Further research is needed to combine the empirical estimates with political economy analyses. This will allow to additionally identify critical regulatory loopholes that also influence the availability and complexity of existing data, providing a more comprehensive study. Developing strong and reliable measurements can help resource-rich countries like Peru correctly identify current gaps and enable for more focused responses in the fight against illicit financial flows.
Link to the article available soon.
Ministerio de Energía y Minas (2018). Anuario Minero 2018.