PurposeAlthough risk culture is a key determinant for an effective risk management, identifying t... more PurposeAlthough risk culture is a key determinant for an effective risk management, identifying the risk culture of a firm can be challenging due to the abstract concept of culture. This paper proposes a novel approach that uses unsupervised machine learning techniques to identify significant features needed to assess and differentiate between different forms of risk culture.Design/methodology/approachTo convert the unstructured text in our sample of banks' 10K reports into structured data, a two-dimensional dictionary for text mining is built to capture risk culture characteristics and the bank's attitude towards the risk culture characteristics. A principal component analysis (PCA) reduction technique is applied to extract the significant features that define risk culture, before using a K-means unsupervised learning to cluster the reports into distinct risk culture groups.FindingsThe PCA identifies uncertainty, litigious and constraining sentiments among risk culture feat...
We study the effects of transparency disclosures on the risk culture, corporate culture, and perf... more We study the effects of transparency disclosures on the risk culture, corporate culture, and performance of U.S. banks. Using stress test regulation, textual analysis, and a regression discontinuity design, we exploit the regulation's quasi-experimental properties around the bank-size policy thresholds. We find that stress-test banks substantially improve their risk and corporate cultures. Strong risk culture banks further reduce their risk densities, risky loans, and costs of debt while increasing profitability. Surprisingly, these effects do not exist among the subset of banks with strong corporate culture only. When examining banks with both strong risk and corporate cultures, the effects switch signs showing evidence of moral hazard. Our findings highlight the differentiating role of risk culture and the restraining impact organizational culture has in mitigating risk.
Using machine learning (ML) and contrasting with simple textual sentiment score and principal com... more Using machine learning (ML) and contrasting with simple textual sentiment score and principal components analysis (PCA) methods, we examine the time series of content within over 700,000 sell-side analyst research reports from 1983 to 2017. We find that analyst reports have significantly changed across a variety of dimensions including length and content of four existing and two new dictionaries related to valuation and alternative metrics. We find that the naive net tone of reports only explains contemporaneous, not future, equity returns. On the other hand, we find that ML methods provide substantially different results from naive sentiment and PCA approaches on determining the impact of analyst reports on financial markets. We also examine the ability of reports to predict changes in firm short interest and volatility. Overall, we find that sell-side analyst reports have stronger impact on smaller cap stocks.
In this paper, we examine the effect of different risk cultures on a banks performance, risk-taki... more In this paper, we examine the effect of different risk cultures on a banks performance, risk-taking incentives and corporate governance. We extend beyond banks' risk culture evaluation and classification using interviews and surveys, and apply textual analysis and cluster analysis to banks' annual 10-K reports. Benefiting from the sentiment dictionary built by Loughran and McDonald (2011) and risk culture framework proposed by McConnell (2013), we develop a twodimensional dictionary and extract paragraph level features reflecting sentiment and risk culture topics from 10-K reports. We use legal expense data in a regression analysis based supervised learning for feature reduction and implement a k-means clustering to categorize the 10-K reports into three risk culture classes: high, moderate and low risk cultures. For our sample of U.S. bank holding companies, we find that banks in the high and moderate risk culture classes have better performance, lower total compensation fo...
We study the effects of transparency disclosures on the risk culture, corporate culture, and perf... more We study the effects of transparency disclosures on the risk culture, corporate culture, and performance of U.S. banks. Using stress test regulation, textual analysis, and a regression discontinuity design, we exploit the regulation's quasi-experimental properties around the bank-size policy thresholds. We find that stress-test banks substantially improve their risk and corporate cultures. Strong risk culture banks further reduce their risk densities, risky loans, and costs of debt while increasing profitability. Surprisingly, these effects do not exist among the subset of banks with strong corporate culture only. When examining banks with both strong risk and corporate cultures, the effects switch signs showing evidence of moral hazard. Our findings highlight the differentiating role of risk culture and the restraining impact organizational culture has in mitigating risk.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the ad... more This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Although risk culture is a key element in the risk management process, identifying the risk cultu... more Although risk culture is a key element in the risk management process, identifying the risk culture of a firm can be challenging due to the abstract concept of culture. We introduce text mining and unsupervised machine learning algorithms to define the risk culture for U.S. bank holding companies and examine the relation between risk culture and performance. Applying principal component analysis on textually extracted features from 10-K filings identifies uncertainty, litigious and constraining sentiments among risk culture features to be significant in defining risk culture of banks. Cluster analysis of these features proposes three distinct risk culture clusters which we label as good, fair and poor. Consistent with regulatory expectations, sound risk culture in banks is characterized by high profitability ratios, bank stability, lower default risk and good governance.
We utilize network analysis to evaluate the relationship between firms' characteristics and board... more We utilize network analysis to evaluate the relationship between firms' characteristics and board of director networks. In a sample of 20 largest firms by market value from the energy and the utility industrials, 10 from each sector, we cluster the firms by their firm-level characteristics, as well as develop a multiplex network of the firms' board members consisting of two major layers, one for board members' direct connections and another for their indirect connections. Additionally, 4 sub-layers of each major layer of the multiplex network represent corporate, non-profit organization, education and government/military connections between the board members. Each layer of the multiplex network is weighted so that their combined effect can be depicted as a single layered directors' network. The weights of the multiplex directors' network are learned by relating a firm's director network characteristics with the firm cluster characteristics. We observe that director networks display significant connectivity at all multiplex network layers and firms belonging to the same cluster display similar director network characteristics, specifically enhanced by appropriate weighting of director network layers. The optimal value of network sublayer weights are found by optimizing the cohesion of firms in a cluster by their director multiplex network measures.
Abstract—Weather elements such as temperature and solar
radiation play a crucial role in renewabl... more Abstract—Weather elements such as temperature and solar radiation play a crucial role in renewable power production as well as in power consumption. In this research, we address risk management issues of a power producer for its weather, price and production risk exposures. We investigate the extent of natural hedge embedded in the dependence of power production using renewable energy sources and power consumption on weather risk elements. We further develop a framework to construct explicit cross hedges for the risk management objectives of a renewable power producer using weather derivatives: a relatively new class of financial instruments for weather risk management.
PurposeAlthough risk culture is a key determinant for an effective risk management, identifying t... more PurposeAlthough risk culture is a key determinant for an effective risk management, identifying the risk culture of a firm can be challenging due to the abstract concept of culture. This paper proposes a novel approach that uses unsupervised machine learning techniques to identify significant features needed to assess and differentiate between different forms of risk culture.Design/methodology/approachTo convert the unstructured text in our sample of banks' 10K reports into structured data, a two-dimensional dictionary for text mining is built to capture risk culture characteristics and the bank's attitude towards the risk culture characteristics. A principal component analysis (PCA) reduction technique is applied to extract the significant features that define risk culture, before using a K-means unsupervised learning to cluster the reports into distinct risk culture groups.FindingsThe PCA identifies uncertainty, litigious and constraining sentiments among risk culture feat...
We study the effects of transparency disclosures on the risk culture, corporate culture, and perf... more We study the effects of transparency disclosures on the risk culture, corporate culture, and performance of U.S. banks. Using stress test regulation, textual analysis, and a regression discontinuity design, we exploit the regulation's quasi-experimental properties around the bank-size policy thresholds. We find that stress-test banks substantially improve their risk and corporate cultures. Strong risk culture banks further reduce their risk densities, risky loans, and costs of debt while increasing profitability. Surprisingly, these effects do not exist among the subset of banks with strong corporate culture only. When examining banks with both strong risk and corporate cultures, the effects switch signs showing evidence of moral hazard. Our findings highlight the differentiating role of risk culture and the restraining impact organizational culture has in mitigating risk.
Using machine learning (ML) and contrasting with simple textual sentiment score and principal com... more Using machine learning (ML) and contrasting with simple textual sentiment score and principal components analysis (PCA) methods, we examine the time series of content within over 700,000 sell-side analyst research reports from 1983 to 2017. We find that analyst reports have significantly changed across a variety of dimensions including length and content of four existing and two new dictionaries related to valuation and alternative metrics. We find that the naive net tone of reports only explains contemporaneous, not future, equity returns. On the other hand, we find that ML methods provide substantially different results from naive sentiment and PCA approaches on determining the impact of analyst reports on financial markets. We also examine the ability of reports to predict changes in firm short interest and volatility. Overall, we find that sell-side analyst reports have stronger impact on smaller cap stocks.
In this paper, we examine the effect of different risk cultures on a banks performance, risk-taki... more In this paper, we examine the effect of different risk cultures on a banks performance, risk-taking incentives and corporate governance. We extend beyond banks' risk culture evaluation and classification using interviews and surveys, and apply textual analysis and cluster analysis to banks' annual 10-K reports. Benefiting from the sentiment dictionary built by Loughran and McDonald (2011) and risk culture framework proposed by McConnell (2013), we develop a twodimensional dictionary and extract paragraph level features reflecting sentiment and risk culture topics from 10-K reports. We use legal expense data in a regression analysis based supervised learning for feature reduction and implement a k-means clustering to categorize the 10-K reports into three risk culture classes: high, moderate and low risk cultures. For our sample of U.S. bank holding companies, we find that banks in the high and moderate risk culture classes have better performance, lower total compensation fo...
We study the effects of transparency disclosures on the risk culture, corporate culture, and perf... more We study the effects of transparency disclosures on the risk culture, corporate culture, and performance of U.S. banks. Using stress test regulation, textual analysis, and a regression discontinuity design, we exploit the regulation's quasi-experimental properties around the bank-size policy thresholds. We find that stress-test banks substantially improve their risk and corporate cultures. Strong risk culture banks further reduce their risk densities, risky loans, and costs of debt while increasing profitability. Surprisingly, these effects do not exist among the subset of banks with strong corporate culture only. When examining banks with both strong risk and corporate cultures, the effects switch signs showing evidence of moral hazard. Our findings highlight the differentiating role of risk culture and the restraining impact organizational culture has in mitigating risk.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the ad... more This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Although risk culture is a key element in the risk management process, identifying the risk cultu... more Although risk culture is a key element in the risk management process, identifying the risk culture of a firm can be challenging due to the abstract concept of culture. We introduce text mining and unsupervised machine learning algorithms to define the risk culture for U.S. bank holding companies and examine the relation between risk culture and performance. Applying principal component analysis on textually extracted features from 10-K filings identifies uncertainty, litigious and constraining sentiments among risk culture features to be significant in defining risk culture of banks. Cluster analysis of these features proposes three distinct risk culture clusters which we label as good, fair and poor. Consistent with regulatory expectations, sound risk culture in banks is characterized by high profitability ratios, bank stability, lower default risk and good governance.
We utilize network analysis to evaluate the relationship between firms' characteristics and board... more We utilize network analysis to evaluate the relationship between firms' characteristics and board of director networks. In a sample of 20 largest firms by market value from the energy and the utility industrials, 10 from each sector, we cluster the firms by their firm-level characteristics, as well as develop a multiplex network of the firms' board members consisting of two major layers, one for board members' direct connections and another for their indirect connections. Additionally, 4 sub-layers of each major layer of the multiplex network represent corporate, non-profit organization, education and government/military connections between the board members. Each layer of the multiplex network is weighted so that their combined effect can be depicted as a single layered directors' network. The weights of the multiplex directors' network are learned by relating a firm's director network characteristics with the firm cluster characteristics. We observe that director networks display significant connectivity at all multiplex network layers and firms belonging to the same cluster display similar director network characteristics, specifically enhanced by appropriate weighting of director network layers. The optimal value of network sublayer weights are found by optimizing the cohesion of firms in a cluster by their director multiplex network measures.
Abstract—Weather elements such as temperature and solar
radiation play a crucial role in renewabl... more Abstract—Weather elements such as temperature and solar radiation play a crucial role in renewable power production as well as in power consumption. In this research, we address risk management issues of a power producer for its weather, price and production risk exposures. We investigate the extent of natural hedge embedded in the dependence of power production using renewable energy sources and power consumption on weather risk elements. We further develop a framework to construct explicit cross hedges for the risk management objectives of a renewable power producer using weather derivatives: a relatively new class of financial instruments for weather risk management.
Uploads
Papers by Abena Owusu
radiation play a crucial role in renewable power production as
well as in power consumption. In this research, we address risk
management issues of a power producer for its weather, price and
production risk exposures. We investigate the extent of natural
hedge embedded in the dependence of power production using
renewable energy sources and power consumption on weather
risk elements. We further develop a framework to construct
explicit cross hedges for the risk management objectives of a
renewable power producer using weather derivatives: a relatively
new class of financial instruments for weather risk management.
radiation play a crucial role in renewable power production as
well as in power consumption. In this research, we address risk
management issues of a power producer for its weather, price and
production risk exposures. We investigate the extent of natural
hedge embedded in the dependence of power production using
renewable energy sources and power consumption on weather
risk elements. We further develop a framework to construct
explicit cross hedges for the risk management objectives of a
renewable power producer using weather derivatives: a relatively
new class of financial instruments for weather risk management.