Back to: Economics, Finance, and Accounting
Black Box Model
A black box model describes the relationship between the inputs and outputs of a system. This model is used in different contexts and has different meanings. As it is often used in science, computing and engineering, a black box model is a device that describes the functional relationships between system inputs and outputs.
In business, a black box model is a financial model where a computerized program is designed to change various investment data into strategies that are useful for investments. The “black” in the black box model refers to the lack of access to the internal workings or parameters of functions of the model. The white box model is the opposite of a black box model, in the sense that its inner components are accessible and can be inspected.
A Little More on What is Black Box Model
Before the black box model was widely adopted in the financial markets, it was peculiar to the science, computing and engineering fields where the relationships between system inputs and outputs are described. In the financial markets, the black box model is associated with the decision-making process of an investment. This model could be an algorithm, a transistor or even a human memory or brain. The use of the black box model in the financial markets has however raised questions on the systematic risks that this model contributes to the market, given the tendency of investors to hide the actual risk of investments under the use of computer programs and technology.
The black box model is also used as a model consumer behavior theory to describe the stimulus-response pattern of consumers.
The Black Box Model Over the Years
The use of the black box model in the financial markets is largely dependent on the market conditions and the market cycle. During periods of high volatility in the market, black box models can cause more dangers and the ultimate destruction of the market. Examples of how black box strategies cause destruction are the flash crash of 2015, the portfolio insurance episode of 1987, the long-term capital management implosion of 1998, among others. Given that black box strategies carry inherent risks, a number of concerns have been raised against their use.
However, technological advancement, machine learning, data science and other related fields have led to the sophistication of the black box models. Presently, institution investment managers and hedge funds still use these strategies when handling complicated investments.
Reference for “Black Box Model”
Academics research on “Black Box Model”
Capacity, leadership, and organizational performance: Testing the black box model of public management, Andrews, R., & Boyne, G. A. (2010). Capacity, leadership, and organizational performance: Testing the black box model of public management. Public Administration Review, 70(3), 443-454. According to recent “black box” models of public management, managerial capacity is a critical component for achieving service delivery improvement. In particular, black box models assume that the impact of management systems is maximized through integration with effective leadership. This assumption is tested by analyzing the effects of managerial capacity and organizational leadership on the performance of English local governments, while controlling for a range of other variables, including organizational size, resources, external constraints, and past performance. Empirical results show that capacity possesses a positive statistical association with local government performance and that leadership can enhance the impact of effective management systems.
Stock trading rule discovery with an evolutionary trend following model, Hu, Y., Feng, B., Zhang, X., Ngai, E. W. T., & Liu, M. (2015). Stock trading rule discovery with an evolutionary trend following model. Expert Systems with Applications, 42(1), 212-222. Evolutionary learning is one of the most popular techniques for designing quantitative investment (QI) products. Trend following (TF) strategies, owing to their briefness and efficiency, are widely accepted by investors. Surprisingly, to the best of our knowledge, no related research has investigated TF investment strategies within an evolutionary learning model. This paper proposes a hybrid long-term and short-term evolutionary trend following algorithm (eTrend) that combines TF investment strategies with the eXtended Classifier Systems (XCS). The proposed eTrend algorithm has two advantages: (1) the combination of stock investment strategies (i.e., TF) and evolutionary learning (i.e., XCS) can significantly improve computation effectiveness and model practicability, and (2) XCS can automatically adapt to market directions and uncover reasonable and understandable trading rules for further analysis, which can help avoid the irrational trading behaviors of common investors. To evaluate eTrend, experiments are carried out using the daily trading data stream of three famous indexes in the Shanghai Stock Exchange. Experimental results indicate that eTrend outperforms the buy-and-hold strategy with high Sortino ratio after the transaction cost. Its performance is also superior to the decision tree and artificial neural network trading models. Furthermore, as the concept drift phenomenon is common in the stock market, an exploratory concept drift analysis is conducted on the trading rules discovered in bear and bull market phases. The analysis revealed interesting and rational results. In conclusion, this paper presents convincing evidence that the proposed hybrid trend following model can indeed generate effective trading guidance for investors.
Financial fraud detection model: based on random forest, Liu, C., Chan, Y., Alam Kazmi, S. H., & Fu, H. (2015). Financial fraud detection model: based on random forest. International journal of economics and finance, 7(7). Business’s accelerated globalization has weakened regulatory capacity of the law and scholars have been paid attention to fraud detection in recent years. In this study, we introduced Random Forest (RF) for financial fraud technique detection and detailed features selection, variables’ importance measurement, partial correlation analysis and Multidimensional analysis. The results show that a combination of eight variables has the highest accuracy. The ratio of debt to equity (DEQUTY) is the most important variable in the model. Moreover, we applied four statistic methodologies, including parametric and non-parametric models to construct detection models and concluded that Random Forest has the highest accuracy and the non-parametric models have higher accuracy than non-parametric models. However, Random Forest can improve the detection efficiency significantly and have an important practical implication.
Designing personalized intelligent financial decision support systems, Palma-dos-Reis, A. (1999). Designing personalized intelligent financial decision support systems. Decision Support Systems, 26(1), 31-47. The variety of investment methods and the complexity of investment decisions have increased steadily in the last two decades. This growth has created a need for comprehensive and expandable financial decision support systems (DSS) that embody major approaches toward investment decisions. The question is whether such systems should take into account the investor’s unique requirements and personal characteristics. The answer to this question is critical to the development of personal intelligent financial agents. In this paper, we present the design of such a system, the creation of a prototype, and its use in an exploratory investigation of the impact of investors’ individual characteristics on their use of models in making investment decisions. More specifically, we report on how the gender of investors, and their attitudes towards risk, relate to their choice of investment models, and provide evidence for the possibility of personalizing DSS.
A model of portfolio optimization using time adapting genetic network programming, Chen, Y., Mabu, S., & Hirasawa, K. (2010). A model of portfolio optimization using time adapting genetic network programming. Computers & operations research, 37(10), 1697-1707. This paper describes a decision-making model of dynamic portfolio optimization for adapting to the change of stock prices based on an evolutionary computation method named genetic network programming (GNP). The proposed model, making use of the information from technical indices and candlestick chart, is trained to generate portfolio investment advice. Experimental results on the Japanese stock market show that the decision-making model using time adapting genetic network programming (TA-GNP) method outperforms other traditional models in terms of both accuracy and efficiency. A comprehensive analysis of the results is provided, and it is clarified that the TA-GNP method is effective on the portfolio optimization problem.