Abstract
Using unique data and a new powerful Monte Carlo‐based statistical tool, we examine the effects of concentrated ownership and owner–management (CO‐OM) on the creditor–shareholder agency conflicts in small firms. A significant CO‐OM effect from the small business owner's view, but insignificant from the commercial lenders' perspective, is found. Special features of informational asymmetry problems in small firms with CO‐OM are also highlighted. Theoretical and empirical contributions are made to the small business management and corporate governance literature. Findings obtained from this research have important implications for small business practitioners as well as researchers, and this study can serve as a reference for policymakers and institutional lenders to assist small firms in successfully raising money through debt financing. In addition, a new powerful methodology is introduced to deal with various potential statistical biases and can be further applied to this line of research.
* The authors would like to thank Professor Jess H. Chua at the University of Calgary, two anonymous referees, the associate editor (Professor Gary J. Castrogiovanni), the editor (Professor Chandra S. Mishra), and referees from the 2005 Administrative Sciences Association of Canada (ASAC) Annual Meeting and from the 2005 Financial Management Association (FMA) Annual Meeting for their valuable comments and insights. All remaining errors are the sole responsibility of the authors.
* The authors would like to thank Professor Jess H. Chua at the University of Calgary, two anonymous referees, the associate editor (Professor Gary J. Castrogiovanni), the editor (Professor Chandra S. Mishra), and referees from the 2005 Administrative Sciences Association of Canada (ASAC) Annual Meeting and from the 2005 Financial Management Association (FMA) Annual Meeting for their valuable comments and insights. All remaining errors are the sole responsibility of the authors.
Notes
* The authors would like to thank Professor Jess H. Chua at the University of Calgary, two anonymous referees, the associate editor (Professor Gary J. Castrogiovanni), the editor (Professor Chandra S. Mishra), and referees from the 2005 Administrative Sciences Association of Canada (ASAC) Annual Meeting and from the 2005 Financial Management Association (FMA) Annual Meeting for their valuable comments and insights. All remaining errors are the sole responsibility of the authors.
1 SMEs completing the mail survey did not participate in the telephone surveys.
2 The firm could not be a franchise, holding company or not‐for‐profit agency. Furthermore, the firm had to have no more than 500 employees and have sales exceeding $30,000.
3 We concentrate on existing bank LOCs following the mainstream literature (for example, Berger and Udell Citation1995) that argues LOCs are ideal for studying lender–owner relationships given banks specialize in lending to highly information‐problematic classes of borrowers.
4 A score of 1 indicates the least satisfaction, while a score of 5 indicates the best.
5 To guarantee the robustness of the analysis, we also assigned a value of 1 to the dummy variable measuring the borrower's satisfaction if a firm owner had answered scores of 4 or 5 to all three questions, and 0 otherwise. Statistical results did not change qualitatively. We also repeated the same procedures individually with the three dimensions of satisfaction and no qualitative change was found.
6 The survey did not differentiate whether or not this third party was or was not related to the firm, only that this person was not considered the owner.
7 Sample firms were classified into 20 categories in the survey, and therefore, 19 industry dummies were constructed. The industry average debt ratio from Dun & Bradstreet Industry Norms & Ratios was also collected so that we had a continuous variable to control the industry effect. This serves both as a robustness check and controls the risk aversion of SME owners as discussed in the literature.
8 For a detailed review for the history of Boosting, see Ridgeway (Citation1999) and Friedman, Hastie, and Tibshirani (Citation2000).
9 Whether a variable should be left in the training model and testing model is judged by their relevance scores. If the score of a variable is greater than 1, it is considered a variable with valuable information and is included in the model; otherwise it is excluded.
Additional information
Notes on contributors
Zhenyu Wu
Zhenyu Wu is an associate professor in the Department of Finance and Management Science, College of Commerce at the University of Saskatchewan.
Peggy L. Hedges
Peggy Hedges is a senior instructor in the Haskayne School of Business at the University of Calgary.
Shali Zhang
Shali Zhang was a master student in the Department of Mathematics and Statistics at the University of Calgary.