Login
Request a Demo

Economic Substance: A Machine Learning Perspective on the Multi-Factorial Analysis

Economic substance is part of a series of U.S. anti-abuse doctrines designed to combat tax shelters which comply with the letter of the rules contained in the Internal Revenue Code (IRC) but violate the spirit of the Code. At its core, the economic substance doctrine stands for the principle that the tax authorities are not obligated to respect the tax effects that flow from transactions that lack economic substance.

Partly in response to uncertainty surrounding the application of the test at common law, the economic substance doctrine was codified with the addition of § 7701(o) of the IRC. However, economic substance remains a frequently litigated issue with persisting grey areas that necessitate interpretation.

In this article, we examine three recent economic substance cases to illustrate how Blue J Legal’s machine learning technology can help practitioners validate their tax positions. These cases are: Austin v. Commissioner (T.C., 2017); Gregg v. Department of Revenue (Or. Tax, 2017); and DTDV, LLC v. Commissioner (T.C., 2018).

Codification of Economic Substance

While the judicial roots of the economic substance doctrine can be traced back to the seminal Supreme Court decision in Gregory v. Helvering, 293 U.S. 465 (1935), the doctrine was only codified in 2010 with the addition of § 7701(o) of the IRC. § 7701(o) provides that a transaction shall be treated as having economic substance only if:

  1. the transaction changes in a meaningful way (apart from Federal income tax effects) the taxpayer’s economic position, and 

  2. the taxpayer has a substantial purpose (apart from Federal income tax effects) for entering into such transaction. 

The statute also introduces a substantial penalty for underpayment that ranges from 20% to 40% (§ 6662(b)(6);(i)(1-2)).

Prior to the introduction of § 7701(o), there was a substantial degree of judicial divergence in applying the economic substance test. Some circuit courts applied the two-pronged test, while some relied on either one of the prongs, and others resisted a rigid two-step analysis altogether. Although the statute attempted to clarify the application of the test, the doctrine’s application has remained inconsistent.

Persistent Ambiguity in Application of Test

To begin, the statute provides no guidance on what constitutes a “meaningful” economic change or a “substantial” business purpose. Moreover, § 7701(o)(5)(C) specifically provides that codification does not alter the determination of the doctrine’s applicability at common law. In Notice 2010-62, the IRS has also confirmed that it would continue to rely on relevant common law cases in applying the test set out in § 7701(o).

For practitioners, this means that the uncertainty within this legal area continues to persist despite statutory enactment. This trend is supported by the fact that the doctrine continues to be frequently litigated; Blue J Legal’s database reveals that judgment on economic substance has been rendered in over 60 decisions since 2011. The complexity can be illustrated in three recent decisions where the Court elected to frame the economic substance analysis using the common law test instead of the statutory formulation.

In Austin, which involved the creation of multiple entities and the sale and repurchase of stock, the Court applied the doctrine by characterizing the taxpayer’s activity as four separate transactions. Although the Court conducted its analysis using the common law test articulated in Frank Lyon Co. v. United States, 435 U.S. 561 (1978) and Rice’s Toyota World Inc. v. Commissioner, 81 T.C. 194 (1983), the Court nonetheless arrived at conclusions in line with Blue J Legal’s machine learning and algorithmic predictions in all four instances.

Similarly, the Court in Gregg sought to answer the question of whether two individuals’ venture was considered a business, and whether that venture lacked true economic substance. Instead of applying the conjunctive test articulated in §7701(o), the Court applied the Frank Lyon test at common law. The Court held that the venture lacked economic substance based on findings that the primary investment motivation was tax savings, and that the transaction lacked any reasonable possibility of profit. Although the Court conducted the analysis using the common-law factors, Blue J’s machine learning algorithm was able to predict results consistent with the Court’s application by using principles that underlie the statutory analysis.

Most recently, in DTDV, which concerned an examination of the legitimacy of a partnership interest, the Court engaged in the same economic substance doctrine analysis as above, without any regard to the statutory formulation. The Court went through a six-part fact-driven analysis using the common-law factors and ultimately held that the economic substance doctrine did not justify disregarding the transaction. The machine learning algorithm was also able to handle the common-law application and validate the Court’s decision.

Dynamic Multi-Factor Prediction using Machine Learning

To be comprehensive, the two-part test needs to be expanded to a five-factor analysis that informs the origin, structure, economic impacts, and non-profit effects of the transaction, as well as the taxpayer’s risk level. Blue J Legal’s machine learning analysis incorporates up to two dozen fact-driven considerations commonly cited in leading cases and applied by the IRS. Each of these considerations are weighted differently depending on the unique circumstances of the case.

By leveraging machine learning technology, our multi-factor algorithm is able to predict case outcomes accurately and the analysis remains robust in the face of inconsistent judicial application of the test. Machine learning is particularly well-suited to complex multi-factor analysis because it processes raw factual data while remaining impartial to varying formulations of a legal test that capture the same principles.

If you’re interested in learning more about how machine learning can help practitioners navigate the complexity of the economic substance analysis to validate a tax prediction, watch our Economic Substance: Influential Factors Webinar on demand.

Thank you to Monica Layarda, Legal Research Analyst for assisting in the publication of this article.

Stay ahead of the curve.

Join our community of innovative professionals

Subscribe to Newsletter