基於混合模型建立企業風險評估方法

An Assessment of Enterprise Financial Risk Based on Hybrid Model

董律里(Lu-Li Tung)

指導教授 ： 周雨田

國立交通大學/管理學院/經營管理研究所/碩士(2020年)

摘要

財務危機在文獻上的定義是指企業之現金流量無法償還債務，而這樣的情形是發生在企業破產或是清算之前的。財務危機的可能性對於公司本身與金融放款或是投資部門來說都是一項重要的指標，本研究基於混合模型的架構，來建立企業財務危機預測模型，提前偵測企業未來之財務風險，提供企業超前部署的機會。
本研究的分析對象為台灣電子業的上市上櫃公司，研究期間為2000年到2018年，使用了年報上的財務資訊以及總體經濟的資料來作為模型的特徵，並透過資料前處理以及特徵工程的方式，來建立資料集。在模型評比的標準上，由於財務危機發生的機會較小，因此本研究選用了召回率（Recall）、F1值以及AUC來因應類別不平衡的狀況。實證結果說明，不管是以隨機森林作為最終分類器或是以類神經網路為分類器的混合模型，其模型表現皆優於單一模型；而在變數的解釋性上，我們發現獲利能力以及經營表現的相關資訊，對於企業財務危機的預測最具有影響力。

關鍵字

財務危機 ； 企業風險 ； 機器學習 ； 混合模型

並列摘要

Financial distress is defined in the literature as a situation in which the cash flow of a business is unable to repay its debts, and this occurs before the business goes bankrupt or is liquidated. The possibility of financial distress is an important indicator for the company itself as well as for the financial lending or investment departments. This study is based on the framework of hybrid model to build a predictive model of financial distress, detect the future financial risk of the company in advance and provide the opportunity for the company to deploy ahead of schedule.
In this study, the target of analysis is listed companies in Taiwan's electronics industry, the study period is from 2000 to 2018, financial information from annual reports and data from the macroeconomy are used to characterize the model, and the data set is built through data pre-processing and feature engineering. On the model performance, the recall rate, F1 score, and AUC were chosen to account for the categorical imbalance because of the low probability of financial distress. The empirical results show that the model performance is better than that of a single model, whether it is a hybrid model with a random forest as the final classifier or a neural network-like classifier. In terms of the explanatory of variables, we find that information about profitability and operating performance is most influential in predicting the financial crisis of the firm.

並列關鍵字

Financial Distress ； Corporate Risk ； Machine Learning ； Hybrid Model

參考文獻

Altman, E. I. (1968). Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The Journal of Finance, 23(4), 589–609. https://doi.org/10.1111/j.1540-6261.1968.tb00843.x

Barboza, F., Kimura, H., & Altman, E. (2017). Machine learning models and bankruptcy prediction. Expert Systems with Applications, 83, 405–417. https://doi.org/10.1016/j.eswa.2017.04.006

Breiman, L. (1996). Bagging predictors. Machine Learning, 24(2), 123–140. https://doi.org/10.1007/bf00058655

Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324

Chen, M. Y. (2011). Predicting corporate financial distress based on integration of decision tree classification and logistic regression. Expert Systems with Applications, 38(9), 11261–11272. https://doi.org/10.1016/j.eswa.2011.02.173