Machine learning for financial transaction classification across companies using character-level word embeddings of text fields

An important initial step in accounting is mapping financial transfers to the corresponding accounts. We devised machine-learning-based systems that automate this process. They use word embeddings with character-level features to process transaction texts. When considering 473 companies independently, our approach achieved an average top-1 accuracy of 80.50%, outperforming baselines that exclude the transaction texts or rely on a lexical bag-of-words text representation. We extended the approach to generalizes across companies and even across different corporate sectors. After standardization of the account structures and careful feature engineering, a single classifier trained on 44 companies from 28 sectors achieved a test accuracy of more than 80%. When trained on 43 companies and tested on the remaining one, the system achieved an average performance of 64.62%. This rate increased to nearly 70% when considering only the largest sector.     

Effects of export product diversification on quality upgrading: an empirical study

This paper empirically examines the effects of export product diversification on overall export quality in a panel data set of 115 countries from 1970 to 2010. It uses the data sets of the overall export quality and three export diversification measures of the International Monetary Fund: the extensive margin (variation in the number of new products exported), the intensive margin (variation in export values among existing exports), and the overall (Theil) index. It finds that export quality has only been increasing with a higher variation in export values among existing exports in low- and lower-middle-income countries. It also observes that export quality has been increasing with both a higher variation in export values among existing exports and new products exported in upper-middle- and high-income countries. The results are robust to the changing measures of controls in the benchmark model, the inclusion of many other controls; i.e. various measures of globalization, country size, factor endowments, macroeconomic stance, etc., and the exclusion of outliers.     

Information and Knowledge as Antecedents of Consumer Attitudes and Intentions to Buy and Recommend Fair-Trade Products

The authors propose and empirically test a causal model to understand how the availability of fair-trade information and consumer knowledge about this issue affect consumers’ attitudes and intentions toward fair-trade products. The model is built upon the attitude-behavior paradigm and the premises of agency theory. It is tested through structural equation modeling with a sample of 292 Spanish consumers. The findings are that consumers do not have good knowledge about fair trade and that this is significantly determined by the lack of information about this in the market. It is also observed that consumers’ perceptions about the availability of fair-trade information have negative effects on their concern about this issue and that such information as is available is not effective in reducing consumer skepticism. The research represents an extension of previous fair-trade literature because the role of information and communication in improving consumer attitudes and buying intentions has rarely been explored in the case of ethical products.     

Magic mirror in my hand … How trade mirror statistics can help us detect illegal financial flows

Misreporting tricks of different sorts applied to the transfer of goods between different countries are typically exploited by criminals worldwide for money laundering ends. The main international anti‐money laundering organisations started paying attention to this phenomenon, dubbed “trade‐based money laundering” (TBML), a long time ago, but the failure to develop appropriate analytical tools has reportedly dogged preventive actions. Nonetheless, literature has widely advocated the possibility that the analysis of inconsistencies in mirrored bilateral trade data could provide some help. By building on previous contributions in the field, this work sets up a model factoring in the main structural determinants of discrepancies between mirrored data concerning Italy's 2010 to 2013 external trade at a highly detailed (6‐digit) level of goods classification for each partner country. Point estimates of freight costs are used to net each observation of the corresponding cif/fob discrepancy. The regression estimates are then deployed in order to compute TBML risk indicators at a country/4‐digit product level. Based on the indicators, rankings of countries and product lines can be compiled, which may be used for a risk‐driven search of potential illegal commercial transactions.     

Economic Impacts of the Possible China–US Trade War

This article uses a multi-country global general equilibrium (GE) model to numerically simulate the effects of possible China–US trade wars. We introduce an endogenous trade imbalance structure with trade cost into the model which helps to explore both tariff and non-tariff trade war effects. Our simulation results show that China will be significantly hurt by the China–US trade war, but negative impacts are affordable. The US can gain under unilateral sanction measures to China, but will lose if China takes retaliation measures. Comparing the effects under mutual trade war, China will lose more than the US. Introducing non-tariff barrier trade wars will intensify the negative effects, and comparatively negative effects to China are larger than to the US. Mexico’s involvement in trade war with the US will strengthen the negative effects and comparatively hurt the US more. Under non-cooperative and cooperative Nash bargaining equilibrium, the US can gain more than China in trade war negotiation, which means the US has stronger bargaining power than China. Additionally, trade wars between China and the US will hurt most countries and the world especially in GDP and manufacturing employment, but benefit their welfare and trade.     

新进入者如何在版权引进市场中有所作为

外版书引进在近现代中国出版史上具有非常重要的地位。我国版权贸易市场渐趋稳定的力量对比格局，给新进入者造成了不小的困惑：好版权抢不到，差版权没法要，夹缝中求生存异常艰难。新入者要改变这种困境就要精准定位，必须贴合本单位现有的资源基础，必须借助社会化分工并灵活创新。     

Effects of aid for trade on extensive and intensive margins of greenfield FDI

Using a panel dataset of 105 developing countries for the period 2003–15, this paper assesses the effects of Aid for Trade (AfT) on greenfield FDI flows to the aid‐recipient countries. Particularly, this paper classifies the total dollar value of greenfield FDI flows to each recipient country in terms of four different layers: the extensive and intensive margins of projects as well as the extensive and intensive margins of source countries. Applying the system GMM estimator, this paper finds that AfT not only increases the dollar value of FDI flows to the recipient countries but also helps diversify the greenfield projects and source countries. In addition, this paper finds that AfT has a greater effect for greenfield FDI from donor (developed) countries than from non‐donor (developing) countries. Among the three components of AfT, aid for trade‐related infrastructure and aid for trade policy regulations are found to have positive links with greenfield FDI, irrespective of source‐country groups, yet their effects are larger for developed source countries. In contrast, aid for building productive capacity hinders greenfield FDI flows from non‐donor countries, while it promotes greenfield FDI from donor countries. We offer some explanations for this finding.     

Exploring the Intensive and Extensive Margins of World Trade

World trade evolves at two margins. Where a bilateral trading relationship already exists it may increase through time (intensive margin). But trade may also increase if a trading bilateral relationship is newly established between countries that have not traded with each other in the past (extensive margin). We provide an empirical dissection of post–World War II growth in manufacturing world trade along these two margins. We propose a “corner-solutions version” of the gravity model to explain movements on both margins. A Tobit estimation of this model resolves the so-called “distance puzzle”. It also finds more convincing evidence than recent literature that WTO-membership enhances trade. JEL no. F12, F15