東華大學圖書館 |

Essays on Textual Analysis for Insurers' Disclosures.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Essays on Textual Analysis for Insurers' Disclosures./
作者:	Zhang, Jian.
面頁冊數:	1 online resource (153 pages)
附註:	Source: Dissertations Abstracts International, Volume: 84-03, Section: A.
Contained By:	Dissertations Abstracts International84-03A.
標題:	Communication. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29319639click for full text (PQDT)
ISBN:	9798841755234

Essays on Textual Analysis for Insurers' Disclosures.
Zhang, Jian.

Essays on Textual Analysis for Insurers' Disclosures. - 1 online resource (153 pages)

Source: Dissertations Abstracts International, Volume: 84-03, Section: A.

Thesis (Ph.D.)--Temple University, 2022.

Includes bibliographical references

This thesis develops three essays examining the feasibility of textual disclosures in predicting corporate performance and how such disclosures relate to ownership structure and disclosure requirements. By using a topic modeling approach (Roberts et al., 2016a), this thesis identifies the topics discussed in the disclosures and estimates the importance of different topics.Chapter 2 examines the value of the informational content of forward-looking statements (FLS) in the Management's Discussion and Analysis (MD&A) section of P&C insurers' annual regulatory filings for both publicly traded and privately owned insurers. We use a structural topic modeling approach to extract the topics discussed in the FLS and examine whether these topics can explain or predict loss reserve errors and predict a firm's current and future performance. We have four main findings. First, the current year's topical discussions are better predictors of future expenses, losses, surplus, and earnings than typical financial ratios such as net income and ROA. Second, both a topic's importance (topic prevalence) and the year-to-year change in the topic's prevalence provide information about future insurer performance. Third, the predictive power decreases as we try to predict the performance further into the future. Finally, we compare the explanatory power of topics between mutual and stock companies and find the predictive power is higher for non-publicly traded mutual insurers than the other types of insurers.Chapter 3 investigates how the ownership type of insurers and managerial discretion shapes the management's discussion and analysis (MD&A) section in the property and casualty insurers' statutory annual reports. We compare and contrast the content in the disclosure between insurance companies of different ownership types. We obtain insights from applying the managerial discretion hypothesis to explain how insurers discuss different topics in their annual filings. We find at least forty percent of the topics in MD&As are significantly related to the ownership type. Moreover, we use three proxies related to the level of managerial discretion to find that insurers change their discussion of topics as the proxy for managerial discretion varies. These results suggest that insurers discuss less about their performance when the level of managerial discretion is lower. In addition, public stock companies (with higher levels of managerial discretion) have greater levels of discussion about topics related to stock prices. In addition, while our structural topic model was able to discern differences in topics between mutuals and stocks and between topics related to proxies for managerial discretion, a commonly used measure of text similarity (cosine similarity) was not able to discern differences between types of firms.In chapter 4, we study how the impact of enactments of disclosure regulations affects the content of textual disclosures by U.S. property and liability insurers. Since 2002 two major disclosure changes have affected insures in the US: The Sarbanes-Oxley Act (SOX), and the Model Audit Rule (MAR). SOX and MAR apply to different types of (insurance) companies and become effective in different years. To identify the effect of regulation changes on insurer behavior, we use a two-stage difference-in-difference method (Gardner, 2021) which can account for the application of the new requirements at different times for companies headquartered in different states. We use two measures to summarize the content of textual disclosure: the importance of topics estimated by a topic modeling approach and the year-to-year cosine similarity score of disclosures. We find that disclosure requirements affect the prevalence of eighty percent of the topics, and the enactment of requirements significantly increases the discussion for topics related to investment, reserves, losses, and exposures.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798841755234Subjects--Topical Terms:

524709
Communication.
Subjects--Index Terms:

Textual analysisIndex Terms--Genre/Form:

542853
Electronic books.

Essays on Textual Analysis for Insurers' Disclosures.
LDR:05291nmm a2200409K 4500 001 2357346
005 20230621091245.5
006 m o d
007 cr mn ---uuuuu
008 241011s2022 xx obm 000 0 eng d
020 $a 9798841755234
035 $a (MiAaPQ)AAI29319639
035 $a AAI29319639
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Zhang, Jian. $3 1057605
245 1 0 $a Essays on Textual Analysis for Insurers' Disclosures.
264 0 $c 2022
300 $a 1 online resource (153 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 84-03, Section: A.
500 $a Advisor: Grace, Martin.
502 $a Thesis (Ph.D.)--Temple University, 2022.
504 $a Includes bibliographical references
520 $a This thesis develops three essays examining the feasibility of textual disclosures in predicting corporate performance and how such disclosures relate to ownership structure and disclosure requirements. By using a topic modeling approach (Roberts et al., 2016a), this thesis identifies the topics discussed in the disclosures and estimates the importance of different topics.Chapter 2 examines the value of the informational content of forward-looking statements (FLS) in the Management's Discussion and Analysis (MD&A) section of P&C insurers' annual regulatory filings for both publicly traded and privately owned insurers. We use a structural topic modeling approach to extract the topics discussed in the FLS and examine whether these topics can explain or predict loss reserve errors and predict a firm's current and future performance. We have four main findings. First, the current year's topical discussions are better predictors of future expenses, losses, surplus, and earnings than typical financial ratios such as net income and ROA. Second, both a topic's importance (topic prevalence) and the year-to-year change in the topic's prevalence provide information about future insurer performance. Third, the predictive power decreases as we try to predict the performance further into the future. Finally, we compare the explanatory power of topics between mutual and stock companies and find the predictive power is higher for non-publicly traded mutual insurers than the other types of insurers.Chapter 3 investigates how the ownership type of insurers and managerial discretion shapes the management's discussion and analysis (MD&A) section in the property and casualty insurers' statutory annual reports. We compare and contrast the content in the disclosure between insurance companies of different ownership types. We obtain insights from applying the managerial discretion hypothesis to explain how insurers discuss different topics in their annual filings. We find at least forty percent of the topics in MD&As are significantly related to the ownership type. Moreover, we use three proxies related to the level of managerial discretion to find that insurers change their discussion of topics as the proxy for managerial discretion varies. These results suggest that insurers discuss less about their performance when the level of managerial discretion is lower. In addition, public stock companies (with higher levels of managerial discretion) have greater levels of discussion about topics related to stock prices. In addition, while our structural topic model was able to discern differences in topics between mutuals and stocks and between topics related to proxies for managerial discretion, a commonly used measure of text similarity (cosine similarity) was not able to discern differences between types of firms.In chapter 4, we study how the impact of enactments of disclosure regulations affects the content of textual disclosures by U.S. property and liability insurers. Since 2002 two major disclosure changes have affected insures in the US: The Sarbanes-Oxley Act (SOX), and the Model Audit Rule (MAR). SOX and MAR apply to different types of (insurance) companies and become effective in different years. To identify the effect of regulation changes on insurer behavior, we use a two-stage difference-in-difference method (Gardner, 2021) which can account for the application of the new requirements at different times for companies headquartered in different states. We use two measures to summarize the content of textual disclosure: the importance of topics estimated by a topic modeling approach and the year-to-year cosine similarity score of disclosures. We find that disclosure requirements affect the prevalence of eighty percent of the topics, and the enactment of requirements significantly increases the discussion for topics related to investment, reserves, losses, and exposures.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Communication. $3 524709
653 $a Textual analysis
653 $a Insurer
653 $a Textual disclosure
653 $a Corporate performance
653 $a Ownership structure
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0511
690 $a 0501
690 $a 0459
690 $a 0703
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a Temple University. $b Business Administration/Risk Management and Insurance. $3 1682143
773 0 $t Dissertations Abstracts International $g 84-03A.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29319639 $z click for full text (PQDT)