Risk Management and Corporate Governance...

Transcription

1 Risk Management and Corporate Governance: The Importance of Independence and Financial Knowledge (with Olfa Maalaoui and Thouraya Triki) Georges Dionne HEC Montreal October 2014

2 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Corporate governance scandals (Enron, Worldcom, etc.) reported in the press revealed serious flaws in the corporate governance system and boards are blamed and accused of failure in their watchdog role SOX enacted in 2002 to restore credibility to the US corporate governance system by setting stricter rules on the functioning and the independence of the external auditor 1 an audit committee has to be entirely composed of independent members (Section 301 of SOX) and 2 at least one member should be considered as financially knowledgeable (Section 407 of SOX).

3 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Following SOX, the NYSE among other exchanges adopts a new set of rules designed to complement SOX requirements 1 a majority of independent directors sitting on the board (Section 303A.01 of NYSE listed Cies manual), 2 an audit committee with a minimum of three members (Section 303A.07 of NYSE listed Cies manual), 3 each member of the audit committee being financially literate, (Section 303A.07 of NYSE listed Cies manual), and 4 at least one member of the audit committee with an accounting knowledge, (Section 303A.07 of NYSE listed Cies manual),

4 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Despite the rules and the regulatory attention given to corporte governance, reports of the OECD attribute the 2007 financial crisis to the failure of the boards in overseeing risk management systems. The reason often being the board limited knowledge: In dealing with losses through to the end of 2007, the report noted that some firms made strategic decisions to retain large exposures to super senior tranches of collateralised debt obligations that far exceeded the firms understanding of the risks inherent in such instruments, and failed to take appropriate steps to control or mitigate those risks...in a number of cases boards were not aware of such strategic decisions and had not put control mechanisms in place to oversee their risk appetite, a board responsability. (e.g. UBS with 18.7 bl, SG. 4.9 bl, Bear Stearns, etc.) (Kirkpatrick (2009), OECD annual report, page 7)

5 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results An important message revealed by the recent financial crisis is the necessity to certify a minimum level of financial knowledge of directors. How can we test that?

6 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results What we do? We analyze the quality of directors sitting on the board and the audit committee and investigate the impact of the financial knowledge of directors and independence on the firm risk management activities in the gold mining industry. 1 We consider the SOX and NYSE requirements and test whether these requirements affect the firm s risk management policy and/or its performance 2 We focus on the financial knowledge requirement and break down the definition of financial knowledge in order to specify different types and levels of financial literacy 3 Using different specifications of financial knowledge we test again their effects on the hedging behavior and the performance of the firm

7 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results The financial knowledge of directors (too much room for interpretation) Section 303A.07 of the NYSE s listed company manual leaves the definition of "financially knowledgeable directors" to the discretion of the board of directors The SOX definition of financial knowledge primarily focuses on whether the director has prior accounting related experience. These definitions leave too much room for interpretation

8 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Our definition of financial knowledge We decompose financial knowledge of directors into three categories. A director is considered as financially knowledgeable if he is : Financially active, i.e. having current or past activities/positions linked to finance (i.e. present or former CFO, an insurer, a financial analyst, a financial consultant, a banker, etc.). Financially educated, i.e. his educational background includes financial literacy (MBA, BBA, B.Comm,...) Possessing an accounting background, i.e. if he has an accounting background (CA, CPA,...), in accordance with the SOX view

9 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Our definitions of financial literacy allow for overlaps. For instance, a financially active director may also be educated and/or have an accounting background. Thus, we further decompose the three categories. A director is : Financially active, educated, and accountant (FinAct_Edu_Acc), Financially active, educated, not accountant (FinAct_Edu_NotAcc), Financially active, not educated, not accountant (FinAct_NotEdu_NotAcc), Not financially active, educated, accountant (NotFinAct_Edu_Acc), Not financially active, educated, not accountant (FinAct_Edu_NotAcc).

10 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results

11 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results

12 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results At least few reasons explain why this reasearch is important: We introduce a new set of explanatory variables that have never been used in the literature to determine the risk management behavior of the firm Hedging in the gold mining industry implies dealing with derivatives and other financially sophisticated tools, yet this requires the presence of financially qualified directors. How the presence of qualified directors affect the risk management? It is quiet easy to hedge gold price risk exposure, thus what would make a difference in the observed risk management activities would likely reflect differences in firm characteristics (control variables) and directors financial expertise. We decompose the financial knowledge of directors and test the separate effects of education and experience on the risk management activity of the firm

13 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results Another unexplored question is whether a more financially qualified director is beneficial to shareholders through the risk management channel.

14 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results We consider the whole industry of gold mining in USA and Canada We benefit from having access to unique detailed data on all risk management activities of these firms. Tufano (JF, 1996) used the same data from and proposes new risk management variables that we use here as controls We construct the same measure of risk management activities covering the total period for which the data was available ( ) Adam (JFE, 2009) used the same period but focuses on the type rather than the level of risk management activities and did not consider governance issues.

15 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results We benefit from an ideal laboratory to test the benefits of the regulatory changes on risk management activities that we should observe when the new regulation becomes effective. Our sample data covers the period (1989 to 1999) preceding major corporate governance problems and the preparatory work that led to changes in corporate governance regulation

16 Motivation and overview This Research The financial knowledge of directors Why this Research is important? Summary of results What do we find? Our results mainly highlight Experience and especially education in finance of inependent directors is positively associated with the firms risk management activities The observed hedging ratio is negatively associated with the presence of directors with accounting background High education of directors and CEO increase the firm risk managment and its performance even if directors are related to the firm s activities A financially qualified board and audit committee increase the firm performance

17 The risk management activities The financial knowledge of directors The hedge ratio as dependent variable Following Tufano (1996), we measure the risk management activity of the firm by the delta percentage: For each firm-quarter, we first calculate the delta of each instrument that is used to hedge the production over the next three years. Then, we multiply each delta by the size (ounces of gold) that the firm hedged, i.e. the ounces of gold that are covered by the corresponding instrument (delta ounces) The delta of the portfoliois the sum over the different delta ounces. Finally, we divide the delta of the portfolio to be hedged by the expected production of gold over the same three years of the hedge (delta percentage).

18 The risk management activities The financial knowledge of directors The median of the observed hedge ratio

19 The risk management activities The financial knowledge of directors The financial knowledge of directors :We hand collect the information about the size and the composition of the board and the audit committee, the name of each director sitting on the board and on the audit committee, the education level of each director, the current and former functions of each director, the age of the CEO and his portfolio holding of common shares and exercisable options.

20 The risk management activities The financial knowledge of directors The financial knowledge of directors : Various sources are used: Publication Different who s who websites, The S&P register of corporations, directors and executives, The Northern Miner, The Financial Post directory of directors, The blue book of Canadian businesses, Encyclopedia of British Columbia, Electronic sources Firms websites available in the internet and google search engine, Proquest ABI/Inform, Factiva provided by Reuters, Other sources Questionnaires, s, Faxes, Phone calls!

21 Why the Heckman model? Our dependent variable, the hedge ratio, has nonnegative values, yet about 15% of the firms in our sample have a hedge ratio equal to zero. One can argue that a standard Tobit model or a Tobit model with a firm fixed effect are appropriate models for our analysis However, Tobit model requires that the empirical realizations of the dependent variable are censored to zero In our case, when the observed hedge ratio is zero, it does not reflect any censoring but it reflects the directors decision of not to hedge

22 The advantage of using the Heckman two-stage model Instead of using the Tobit model, we choose to use the Heckman (1979) two-stage model to separately model : 1 The determinants of the firm s decision to hedge or not to hedge using a binary variable 2 The determinants of the intensity of the hedge for firms who hedge using an OLS with firms fixed effect The first stage in the Heckman model yields to the calculation of the Inverse Mills ratio, which is used in the second stage as a regressor to control for sample selection biases.

23 The Heckman first-stage estimation results

24 Control variables in the Heckman second-stage estimation

25 Measuring the compliance indexes For SOX, we construct the following dummy variables and the resulting SOX index 1 A1=1 if the audit committee is entirely composed of independent directors and 0 otherwise. 2 A9=1 if at least one director in the audit committee is considered as financially knowledgeable and 0 otherwise. Compliance_SOX 1 = A1 + A9, Compliance_SOX 2 = A1 A9.

26 Measuring the compliance indexes For NYSE, we consider the following dummy variables and index 1 B1=1 if the majority of directors in the board is independent and 0 otherwise. 2 A2=1 if the audit committee contains at least three members and 0 otherwise. 3 A5=1 if each member of the audit committee is financially literate. 4 A12=1 if at least one director in the audit committee has an accounting background and 0 otherwise. Compliance_NYSE 1 = B1 + A2 + A5 + A12, Compliance_NYSE 2 = B1 A2 A5 A12.

27 Therefore, considering a combination of the six rules Compliance_SOX 1_NYSE 1 = A1 + A9 + B1 + A2 + A5 + A12, Compliance_SOX 2_NYSE 2 = A1 A9 B1 A2 A5 A12.

28 Compliance with SOX

29 Compliance with NYSE

30 Compliance with SOX and NYSE

31 Effect of the regulations on the firm performance: Simultaneous estimation of observed hedging and performance Panel A: The hedge ratio is the dependent variable

32 Simultaneous estimation of observed hedging and performance Panel B: The firm performance is the dependent variable

33 Effect of the regulations on the predicted level of hedging and firm performance Panel A: First stage: the dependent variable is the hedge ratio

34 Effect of the regulations on the predicted level of hedging and firm performance Panel C: Second stage: the dependent variable is the firm performanc

35 The financial experience of directors The audit committee

36 The financial experience of directors The board

37 The financial education of directors The audit committee

38 The board

39 The accounting background of directors The audit committee

40 The accounting background of directors The board

41 Measures of the quality of the board and the audit committee Dummy variables considered from the characteristics of the audit: 1 A1=1 if the audit committee is entirely composed of independent directors and 0 otherwise. 2 A2=1 if the audit committee contains at least three members and 0 otherwise. 3 A3=1 if the average tenure in the audit committee for directors is superior to 10 years and 0 otherwise. 4 A4=1 if at least 25% of the audit committee hold a Master degree or a PhD in finance and 0 otherwise. 5 A5=1 if each member of the audit committee is financially literate. 6 A6=1 if the majority of the audit committee is comprised of financially active directors and 0 otherwise. 7 A7=1 if the majority of the audit committee is comprised of financially educated directors and 0 otherwise. 8 A8=1 if the majority of the audit committee is comprised of directors with accounting background and 0 otherwise.

42 Dummy variables considered from the characteristics of the board 1 B1=1 if the majority of directors in the board is independent and 0 otherwise. 2 B2=1 if the CEO is the COB and 0 otherwise. 3 B3=1 if the average tenure of directors in the board is superior to 10 years and 0 otherwise. 4 B4=1 if at least 25% of directors in the board hold a Master or a PhD in finance and 0 otherwise. 5 B5=1 if the majority of directors in the board are considered as financially knowledgeable and 0 otherwise. 6 B6=1 if at least 25% of directors in the board are financially active and 0 otherwise. 7 B7=1 if at least 25% of directors in the board are financially educated and 0 otherwise. 8 B8=1 if at least 25% of directors in the board have an accounting background and 0 otherwise.

43 Using these dummy variables, we construct the following indexes to measure the quality of the audit committee and the bord Gov_index_audit = A1 + A2 A3 + A4 + A5 + A6 + A7 + A8 Gov_index_board = B1 B2 B3 + B4 + B5 + B6 + B7 + B8

44 Effect of directors "financial quality" on hedging

45 Effect of directors "financial quality" on firm performance Panel A: The hedge ratio is the dependent variable

46 Effect of directors "financial quality" on firm performance Panel B: The firm performance is the dependent variable

47 High education versus hedging and firm performance Education of members of the audit

48 Education of members of the board (Panel A) and the CEO (Panel B)

49 The effect of high education on firm performance The hedge ratio is the dependent variable

50 The high level of education is the dependent variable

51 Our results can be summarized as follows: Only SOX rules are found to have an effect on the hedging behavior using derivatives as well as the performance of the firm Composition of the audit: The presence of directors with experience, education and accounting background in the audit committee is positively related to hedging if directors are independent and negatively related to hedging when directors are related to the firm s activities. The presence of educated members is generally associated with less activity. Yet, when the audit is fully independent the presence of accountants increase the hedging activity.

52 Composition of the board: Experience of directors increase the hedging activity no matters whether the majority of directors are independent. Yet, the presence of accountants in the board is likely to reduce the hedging activity Educated directors in the board is associated with more hedging Accountants in the board seem to increase the propensity of hedging when most of them are dependent to the firm s activities High education of the board and the CEO increase the firm performance. High education seems to be a director "must have" especially for firms using sophisticated instruments to hedge their exposure

53 It is interesting to observe that starting from the recent crisis financial institutions in North America are obligated to have a risk committee on the board: At least three members All must be independent An adequate number of committee members should have suffi cient knowledge in the risk management of financial institutions When appropriate the committee should include individuals with technical knowledge in the risk discipline.