Value assessment of companies by using an enterprise value assessment system based on their public transfer specification
Introduction
Enterprise value assessment, also known as company valuation or corporate valuation, is an saccurate assessment of the intrinsic value of a company. Enterprise value assessment provides not only references for the internal financial management of the enterprise, but also an important basis for external investors to investigate the company. Research believed that company values have different aspects, namely income, growth, and risk (Copeland & Copeland, 1999). In practice, companies report a variety of financial indicators which relate to these three aspects. Despite information contained in complicated financial indicators, the intrinsic value of a company is also related to that company's technical level, human resources, reputation, and other non-financial information. Finally, the reaction of the company's intrinsic value in the market is called the stock price.
The National Equities Exchange and Quotations (NEEQ) in China is an over-the-counter market aiming to provide an investment platform between startup companies and registered investors. There were already over 10,038 companies listed on NEEQ up to June 2019. As an important part of the Chinese capital market, NEEQ is mainly targeted at small or startup companies which are unlisted in mainboard markets. Note that some differences between startup companies on NEEQ and listed companies in mainboard markets raise challenges for the traditional company valuation method in the following three aspects. Firstly, different from the listed mainboard market, low liquidity on NEEQ means that companies have no reliable stock price, or, in other words, stock price is not a reasonable estimation of a company's intrinsic value. Secondly, due to the short period of time that companies listed on NEEQ have been operating, and their infrequent transactions, financial data about a specific company is sparse, which increases the difficulty in estimating future growth trends based solely upon past financial information. Moreover, the scales of startup companies’ assets on NEEQ are generally small and this brings higher risk and uncertainty to value assessment.
For listed companies, the enterprise valuation model based on financial indicators has been widely used. The three basic valuation methods, the cost method, the market method, and the income method, have been discussed and applied by enterprise themselves and academic institutions (Kooler et al., 2010). In the scenario, where startup companies are unable to make a profit or provide reasonable financial indicators, these methods fail to give a reasonable value assessment due to lack of financial data. What is more, traditional methods ignore the value reflected by non- information, which is even more important for startup companies. In fact, non-financial information not only demonstrates the operation of all aspects in the past, but also implies the company's future planning and development. It is one significant compensation for the lack of financial indicators, which largely saffects investors’ judgment of a company's value.
As a result, constructing a comprehensive value assessment system for the startup companies and determining a reasonable valuation range for them are central topics that investors, startup companies and financial researchers are truly concerned about. Taking both financial factors and non-financial factors into account, this study proposes a new enterprise value assessment system (EVAS) and combines text-mining technology with the traditional relative valuation method to improve the assessment accuracy. The public transfer statement, a type of announcement issued by the company at the time of listing, systematically and comprehensively demonstrates the operation of the company, which is an important basis for assessing the company's value. Therefore, we use the content of this statement as the source for the text.
Our contribution can be summarized in two aspects:
- 1)
A text-based Enterprise Value Assessment System (EVAS) combining financial information and non-financial information is proposed. Compared to traditional value assessment models which mostly rely on financial indicators, our system leverages text technology to quantify non-financial information.
- 2)
We conduct detailed experiments on the value assessment task of companies listed on NEEQ. Evaluations of both our classification and regression prove our system's effectiveness in relation to startup companies.
The remaining part of our study is structured as follows: Section 2 reviews related work about value assessment, usage of non-financial information and some applications of text technologies. Section 3 interprets our proposed value assessment system in theory. Section 4 lists in detail the text techniques we used to quantify indicators in our proposed system. Section 5 shows our experiment design and evaluation result. Section 6 concludes our work and highlights areas for future direction.
Section snippets
Related work
The existing research on value assessment from a financial viewpoint mainly explores the corporate value from various financial indicators. This research follows three directions: 1) comparing the effects of different models to improve their applicability (Pan, 2016; Leilei, 2016; Ru 2017); 2) comprehensively considering the valuation results of different valuation models to improve the accuracy of the valuation (Duojiao & Yujun, 2010); 3) improving the valuation model by introducing more
Value chain theory
Michaele Porter believes that the goal of company management is to maximize its value and that a company creates value through its value chain (Porter, 1989). According to value chain theory, the value creation process of an enterprise consists of a series of activities that are different but related to each other, namely basic activities and auxiliary activities. The basic activities focus on the process of directly creating value, including designing, manufacturing, sales and operation. On
Valuation model combining financial and non-financial information
In this section, our quantification method and valuation model are introduced. Since our data sources are mainly unstructured text, the first step of our work is to build a price-range labeled vocabulary to extract features for indicators in EVAS. The vocabulary collects words for each indicator and each price range, which will be discussed in Section 4.1. Secondly, we discuss how we use the constructed vocabulary and proposed indicator system to compose feature vectors in Section 4.2. Thirdly,
Company selection and data source
Before discussing our company selection, we need to introduce the “private placement” activity on NEEQ. Private placement is one of the main approaches for companies to obtain financial investment. During private placement, companies sell shares to some specific institutional investors in exchange for financial investments. Involved investors need to estimate the share price of the targeted company but they do not want to use the stock price because the stock price on NEEQ cannot reflect
Conclusion and future work
With regard to the characteristics of companies listed on NEEQ, a comprehensive enterprise value assessment system is proposed to consider the non-financial aspects of overall company activities. Taking text in the public transfer statement as an important basis for value assessment, we construct and augment the business word vocabulary corresponding to each indicator. In order to improve evaluation of the special factor K in the P/B relative valuation method, two neural network models are
CRediT authorship contribution statement
Win-Bin Huang: Conceptualization, Supervision, Writing - review & editing. Junting Liu: Data curation, Methodology, Formal analysis, Writing - original draft. Haodong Bai: Formal analysis, Visualization, Writing - review & editing. Pengyi Zhang: Supervision, Writing - review & editing.
References (34)
- et al.
Corporate diversification, ownership structure, and firm value: The Singapore evidence
International Review of Financial Analysis
(2000) - et al.
Media-aware quantitative trading based on public Web information
Decision Support Systems
(2014) - et al.
Learning short-text semantic similarity with word embeddings and external knowledge sources
Knowledge-Based Systems
(2019) CEO attributes, compensation, and firm value: evidence from a structural estimation
Journal of Financial Economics
(2018)- et al.
How do family ownership, control and management affect firm value?
Journal of financial Economics
(2006) - et al.
Cognitive-inspired domain adaptation of sentiment lexicons
Information Processing & Management
(2019) The effect of the set of comparable firms on the accuracy of the price-earnings valuation method
Journal of Accounting Research
(1992)Random forests
Machine Learning
(2001)Classification and regression trees
(2017)Affective computing and sentiment analysis
IEEE Intelligent Systems
(2016)
Managing corporate FX risk: a value-maximizing approach
Financial Management
An improved valuation model for technology companies
International Journal of Financial Studies
How good corporate reputations create corporate value
Corporate Reputation Review
Improving regressors using boosting techniques
In ICML
LIBLINEAR: A library for large linear classification
Journal of Machine Learning Research
Greedy function approximation: a gradient boosting machine
Annals of Statistics
Corporate valuation : an easy guide to measuring value
Computer Programming Language.
Cited by (10)
Extraction and classification of risk-related sentences from securities reports
2022, International Journal of Information Management Data InsightsCitation Excerpt :Bao & Datta (2014) and Dyer et al. (2017) provided a method for analyzing companies' textual risk disclosures to discover risk factors (Bao & Datta, 2014; Dyer et al., 2017). Huang et al. (2020) and Chang et al. (2021) assessed corporate value evaluation using text mining, and their results confirmed that there was an improvement in the accuracy of value assessments for companies making investment recommendations (Chang et al., 2021; Huang et al., 2020). Kurian et al. (2020) used machine learning and keyword analysis to analyze incidents and attempted to mitigate the risk of oil sands operations (Kurian et al., 2020).
A machine learning, bias-free approach for predicting business success using Crunchbase data
2021, Information Processing and ManagementCitation Excerpt :The alignment of the personal goals of the founder and the growth of the company was also studied by Cooper. More recently, [23] built a framework for assessing enterprise value using factors like number of patents, R&D employees, and share of owners among management. Enterprise valuation is an important metric that could be used to determine the success of a company.
Social Media Marketing and Financial Forecasting
2020, Information Processing and ManagementEvaluation metrics and dimensional reduction for binary classification algorithms: A case study on bankruptcy prediction
2022, Knowledge Engineering ReviewExploring investor-business-market interplay for business success prediction
2023, Journal of Big DataDo Financial Support Policies Catalyse the Development of New Consumption Field?—Evidence from China’s New Consumer Enterprises
2022, Sustainability (Switzerland)