This paper aims to comprehensively uncover bank risk factors from qualitative textual risk disclosures reported in financial statements, which contain a huge amount of information on bank risks. We propose a new semi‐supervised text mining approach named naive collision algorithm to analyse the textual risk disclosures, which can more accurately identify bank risk factors compared with the typical unsupervised text mining approach. We identified 21 bank risk factors in total, which is far more than identified in previous studies. We further analyse the importance of each bank risk factor and how the importance of each risk factor changes over time. 相似文献
Aims: This study analyzed discrepancies in the quantity of medical services supplied by physicians under different payment systems for patients with different health statuses and illnesses by means of a field experiment.
Methods: Based on the laboratory experiment of Heike Hennig-Schmidt, we designed a field experiment to examine fee-for-service (FFS), capitation (CAP), and diagnosis-related group (DRG) payment systems. Medical students were replaced with 220 physicians as experimental subjects, which more closely reflected the clinical choices made by physicians in the real world. Under the three payment mechanisms, the quantity of medical services provided by physicians when they treated patients with different health statuses and illnesses were collected. Finally, relevant statistics were computed and analyzed.
Results: It was found that payment systems (sig. = 0.000) and patient health status (sig. = 0.000) had a stronger effect on physicians’ choices regarding quantity of medical services than illness types (sig. = 0.793). Additionally, under the FFS and CAP payment systems, physicians overserved patients in good and intermediate health while underserving patients in bad health. Under the DRG payment system, physicians overserved patients in good health and underserved patients in intermediate and bad health. Correspondingly, under FFS and CAP, the proportional loss of benefits was the highest for patients in bad health and the lowest for patients in good and intermediate health; while under DRGs, patients in good and intermediate health had the largest and smallest loss of benefits, respectively.
Limitations: In order to increase external effects of experiment results, we used the field experiment to replace laboratory experiment. However, the external effects still existed because of the blurring and abstraction of the parameters.
Conclusions: Medical treatment cost and price affected the quantity of medical services provided by physicians. Therefore, we proposed that a mix of payment systems could address the common interests of physicians and patients, as well as influence incentives from payment systems. 相似文献
Following CEO turnovers, US firms adjust real business activities to manage earnings downwards (REM bath). This effect is most pronounced in firms with low levels of institutional ownership. REM baths early in CEOs’ tenure can be confounded with legitimate adjustments to business activities. However, we show that they are not accompanied by increases in R&D or capital expenses, nor are they explained by restructuring expenses. CEOs with short tenure record more negative REM measures in their first year of tenure, when compared with CEOs with long tenure. 相似文献
Operating cash flow (CFO) asymmetric timeliness occurs when CFO reflects bad news more quickly than good news. We examine the presence and determinants of CFO asymmetric timeliness in Australia, where substantial differences in reporting requirements of cash flow components, in characteristics of listed companies and in the degree of conservative financial reporting produce contrasting findings to those in the United States. We find supportive evidence for the novel ‘sticky cost behaviour’ explanation and also the product-pricing strategy, but not the life cycle hypothesis. These findings are useful for investors and analysts concerned with forecasting the future values of companies. 相似文献