A constraint-based approach for analysing financial market operations

Abstract

The article describes a framework for modelling and verification of constraint rules on operations with financial instruments. These constraints are applied on certain attributes of domains of financial objects. A methodology and implementation of automatic constraint analysis in two steps is presented. The first step involves preparation of constraints on specified domains and creation of formulas defining them. The other step consists in waiting for real time transactions and responding to them by alerting the user on newly occurred constraint violations. Computation reduction method is proposed. A satisfaction coefficient is calculated that aids the end user in taking consecutive actions on their portfolio. INTRODUCTIONModern financial markets are characterized by a rich variety of offered financial instruments and include many participants with competitive goals, which are achieved in highly dynamic market conditions. The financial instruments are often pooled in hierarchical structures (classification groups) like portfolios, sub-portfolios by country, currency, instrument type. The market participants must conform to regulatory rules specifying the distribution of the assets under their control in the separate classification groups. In many cases, the rules for asset allocation in the groups are alternative to each other and require decision making. Due to the high complexity of the regulatory rules, the market participants often make mistakes while operating on the highly dynamic markets and violate the rules about the distribution of the limited financial resources. Thus they impose high risk on the organizations or people whose money they are dealing with. This makes automatization of financial operation analysis in real time a priority. The market participant must make a decision about buying or selling of a certain position before the market trend changes. The dealer has to quickly simulate an operation on a position and make sure all the constraint rules are satisfied before executing the deal on the market. To ensure better experience for market participants and quickly identify constraint violations, formal specification of regulatory rules is required. These regulatory rules are different for each country and are based on local laws.The problems reviewed in the article are related to constraint programming [1]. The authors formally specify constraints using expressions and use constraint logic programming over finite domains. The expressions are processed by a production system program containing facts and rules. The proposed approach allows "What if..." simulations and flexibility in taking alternative decisions. This guarantees continuity and validity of market participants' actions. The analysis is performed in two steps. During the initial check of the constraints, facts representing domains are asserted and synthetic rules are generated from constraint specification by a special parsing rule. After firing, the new rules yield the required result. The described step in a real-life system involv...

Automated Reasoning and Learning for Automated Payroll Management

“…Jin, Tsang, and Li (2009) demonstrates how many central economic problems, such as bargaining, financial investment and supply-chain management can be treated as optimisation problems with certain constraints associated. Nikolov, Nikolov, and Antonov (2013) focusses on operations with financial instruments that model dynamic financial markets. Back and Back (1995) uses tax constraints to verify and guide planning but uses hardcoded parts of legislation as an expert system.…”