Statistical evidence of tax fraud on the carbon allowances market

Marius-Cristian Frunza (1, 2) , Dominique Guegan (1, 3) , Antonin Lassoudière (2)

1 CES - Centre d'économie de la Sorbonne
2 Sagacarbon
3 PSE - Paris School of Economics

The aim of this paper is to show evidence and to quantify with forensic econometric methods the impact of the Value Added Tax fraud on European carbon allowances markets. This fraud mainly occurred at the beginning of between the end of 2008 and the beginning of 2009. In this paper, we explore the financial mechanisms of the fraud and the impact on the market behavior as well as the reflexion on its econometric features. In a previous work, we showed that the European carbon market is strongly influenced by fundamentals factors as oil, energy, gas, coal and equity prices. Therefore, we calibrated Arbitrage Pricing Theory-like models and showed that they have a good forecast capacity. Those models enabled us to quantify the impact of each factor on the market. In this study, we focused more precisely on the benchmark contract for European carbon emissions prices over 2008 and 2009. We observed that during the first semester of 2009, there is a significant drop in our model performances and robustness and that the part of market volatility explained by fundamentals reduced. Therefore, we identified the period where the market was driven by VAT fraud movements and we were able to measure the value of this fraud. Soon after governments passed a law that cut the possibility of fraud occurrence the performance of the model improved rapidly. We estimate the impact of the VAT extortion on the carbon market at 1.3 billion euros.

Le but de cet article est de quantifier la fraude à la TVA sur les allocations de carbone sur les marchés européens. Cette fraude a surtout été enregistrée entre 2008 et 2009. Dans cet article, nous explorons les mécanismes financiers de la fraude et l'impact sur le comportement du marché. Nous avons pu identifier la période où le marché a été alimentée par les mouvements de fraude à la TVA et mesurer la valeur de cette fraude. Après l'intervention des gouvernements pour uniformiser la réglementation de la TVA sur ces marchés, nous constatons que l'évolution des prix changent et leur prévision redevient possible.

Mots clés: Carbon; EUA; energy; Arbitrage Pricing Theory; Switching regimes; hidden Markov Chain Model; forecast.

Mots clés: Carbone; Energie; modèle APT; modèle de Markov; prévision.

Domaines: Economies et finances; Etudes de l'environnement; Gestion et management; Méthodes et statistiques; Probabilités [math.PR]; Statistiques [math.ST]; Théorie [stat.TH]

Marius-Cristian Frunza, Dominique Guegan, Antonin Lassoudière. Statistical evidence of tax fraud on the carbon allowances market. 2010. ⟨halshs-00523458⟩