The Application of Computational Methods in Practice from the perspective of a Law Firm

by Dr. Michael Marquardt

Gleiss Lutz

Lawyers in private practice primarily use digital information for research, in particular for speedy access to court decisions and codes, official statements issued by the legislator and the federal and the states ministries for justice and finance and other legal publications. Datasets, which organize such information for example by underlining the key sentences in respect of specific legal issues, constitute important tools of knowledge management. Argumentative and other logical structures, which may be taken from decisions of courts or state authorities and summarized in essence and in the abstract, also form part of such datasets. Visualized codes, regulations and decrees function as study materials and training tools.

Further, knowledge management systems usually compile text blocks comprising model arguments and fragments of court appeals and other formal legal statements, sometimes even more importantly, prototype contracts or model agreements with standardized terms and conditions. Internal interactive digital platforms and virtual rooms used by lawyers working together from different locations help to improve and update the datasets. Sometimes clients ask for access raising additional issues for cyber security.

Visualized transaction schemes and corresponding legal issues often support client presentations and, in particular, help decisions makers who are not lawyers to identify relevant information fast without misunderstandings. In other cases, servicing clients requires data to be provided for due diligences or the exchange of information with state authorities, which in turn and under certain legal prerequisites may require access to the electronic books of enterprises. Digital tax audits are to become possible. Electronic data delivered to clients often contain complete work products like electronic draft tax balances and tax returns to be filed. In addition, clients have started requesting electronic tools so that they can complete such products themselves. We expect electronic law suits and correspondence in addition to physical oral proceedings soon to come, though no lawyer should sign a legal opinion or statement written by a computer uncontrolled.

Computational methods, which may substitute legal decisions or recommendations in total, do not exist and will probably (and hopefully) not come into existence, at least to the extent that the application of inexact legal concepts, the interpretation of vague legal terms and the exercise of legal discretion have to take facts and circumstances into account which are beyond mathematical logic.

However, sophisticated legaltech software does analyze and compare fact patterns and the legal consequences deriving therefrom, and also recommend favorable legal consequences in a specific scenario. Such recommendations are subject to a formal subsumption process which, due to technical constraints, currently requires unambiguous facts and clear terms and conditions in the applicable legal provisions. In particular in the field of tax law where facts and legal consequences often consist of figures and/or are reflected by bookings processes, computational methods have been developed, which assist to comply with legal obligations and to make decisions by showing legal consequences. Such assistance may implicitly include legal recommendations, also in the area of tax planning.

Financial institutions use software to identify dividend and interest income and to calculate taxes to be withheld on behalf and on account of their customers, and to provide customer data according to the rules of the international exchange of tax relevant information. Tax planning is facilitated in the following examples: As a general rule but subject to certain exemptions, the direct or indirect transfer of at least 95% of the shares in a corporation which owns real property in Germany, triggers German real estate transfer tax. Appropriate software may analyze, whether a sales transaction or restructuring measure within a group of companies would be subject to tax or not. On the basis of fixed parameters, this software may also calculate the value of the property as the tax base and the total amount of tax. So called “interest barrier rules” for enterprises define the legal prerequisites for the tax deductibility of interest expenses, which depends on the amount of net interest paid during a tax period and, where that exceeds a certain threshold, on the taxpayer’s EBITDA. Exemptions from the restrictions set forth by those rules are linked to the debt to equity ratio of affiliated enterprises, or, in the case of corporations, also depend on the ratio of the net interest expenses and the interest paid to specific shareholders. Computational methods may assist in financial decisions by providing options for fully tax deductible interest expenses.

Maybe already in the near future, computational methods may in addition prepare complex legal due diligence reports, legal opinions and other statements by allocating the decisive legal consequences to the relevant facts. Or, insurance companies will demand such statements to be controlled by computers to grant insurance protection. Legal technology will change the legal profession.