Designing an Experiment - University of Queensland · Motivation Corporate Governance has been...
Transcript of Designing an Experiment - University of Queensland · Motivation Corporate Governance has been...
Designing an Experiment
Look at 3 Experiments
1. Schultz, Tan and Walsh (2009) “Endogeneity and the corporate governance - performance relation”
2. Schultz, Smith, Tan and Walsh (2011) “Corporate Governance Regulation: the Cost of Divergence from Equilibrium ”
3. Heider and Ljundgqvist (2012) “As Certain as Debt and Taxes: Estimating the Tax Sensitivity of Leverage from Exogenous State Tax Changes”
First Experiment
Schultz, Tan and Walsh “Endogeneity and the corporate governance - performance relation”
Excellent paper – winner of best paper award AJM 2010
Looks at Corporate Gov – Performance relation using state of the art endogeneity tests
Second Experiment
Schultz, Smith, Tan and Walsh (2011) “Corporate Governance Regulation: the Cost of Divergence from Equilibrium
Even better paper – uses an exogenous event (the changes to the NYSE listing requirements) to study the costs of regulation.
Motivation
Corporate Governance has been shown to be a largely endogenous characteristic of the firm (Demsetz, 1983; Demsetz & Lehn, 1985; Demsetz & Villalonga, 2001; Hermalin & Weisbach, 2003; Wintoki, Linck, & Netter, 2007, Coles, Lemmon, & Meschke, 2011)
Implications for policy makers
Estimated $1.4 trillion spent on compliance with SOX (Zhang, 2005)
Compliance costs for the average firm is over $3 million (Financial Executives Institute, 2007)
Motivation
If firms are at an internal equilibrium, then a forced change in governance structure will not improve firm performance. In fact, it may be detrimental to firm performance
Optimal level of corporate governance is endogenously determined at the firm level (Demsetz, 1983)
Sub-optimal levels may reduce performance
Costly to re-structure governance mechanisms e.g. formation of committees, hiring of additional directors
Graphically
Performance
A
B
CG Structure
Each firm optimises CG Structure/performance relation
Graphically
Performance
A
B
CG Structure
Each firm has reached an equilibrium structure of CG
Graphically
Performance
A
B
CG Structure
A Senate committee requires a study to examine the CG and performance relation
Graphically
Performance
A
B
CG Structure
Based on the study, all firms forced to adopt “optimal level” of CG B is at its optimum but A Has now been shifted below its optimum
Background
The recent changes to the CG requirements by NYSE allows a natural experiment in the CG-firm outcome relation
If firms are in equilibrium in terms of CG, an exogenous change in governance will lead to no improvement in firm performance
May lead to a decline in performance as CG structure is sub-optimal
Strong Prediction of less than or equal to zero change in performance
Methodology
In addition to many requirements, recent changes in NYSE rules have also mandated the requirement of:
1. Independent audit committee 2. Independent nomination committee * 3. Independent compensation committee * 4. Business ethics code 5. CG policy guidelines 6. Majority of independent directors on the board
These are the requirements that are quantifiable
and readily observable by the researcher
* Indicates the rules that had sufficient variation from pre to post rule change
Methodology
Measure the deviations in these characteristics (pre and post – mandated requirements): CG Deviation = (post-period level
minus pre-period level)
Methodology
Do CG Deviations improve firm performance?
Firm performance = f(CG, controls, CG deviations)
Based upon the equilibrium model of CG:
0_ ≤DeviationsCGβ
Results
Tobin’s Q is negatively affected by CG deviations.
1. A deviation in nomination committee decreases Q at the 5% significance level.
2. A deviation in governance policy decreases Q at
the 5% significance level.
Results - Robustness
All of the other Performance Proxes (Total Return, Profit Rate, Return on Assets) are consistent with the null Hypothesis of
Same conclusion when abnormal performance is used (post minus pre performance)
0_ ≤DeviationsCGβ
Discussion
Paper provides evidence that if firms are at equilibrium with their Corporate Governance structure then forced changes like the NYSE listing requirement changes will lead to less than or equal to zero change in performance. This is what is found in the empirical results.
Paper would be even better if it examined other exogenous Corporate Governance events, such as $1 million cap on bonuses, SEC requirement for semi-annual earnings reports, disclosure of insider trading. Look at more US changes in Corporate Governance Look at changes in Corporate Governance Internationally
Third Experiment
Heider and Ljundgqvist (2012) “As Certain as Debt and Taxes: Estimating the Tax Sensitivity of Leverage from Exogenous State Tax Changes”
Looks at Debt and taxes relation Studies of Debt and taxes are fraught with
endogeneity issues. This paper uses many exogenous shocks
(changes in State Taxes) to get around the endogeneity issue.
Third Experiment
This paper looks at exogenous changes to State Taxes. There are a lot:
38 corporate tax increases in 25 States affecting 1,824 firms over 1990-2011 (pretty evenly split between the 1990’s and 2000’s)
67 corporate tax decreases in 29 States affecting 7,021 firms (again a pretty even split between 1990’s and 2000’s)
Third Experiment
Paper does a pretty good job of showing that the tax changes are exogenous eg looking at what happens to firms in the same industry in neighbouring States where there are no tax changes.
Finds a positive relation between increases in Taxes and Leverage --- Firms increase leverage on average by 114 basis points in response to a rise in their home State taxes
Effect is asymmetric – firms do not decrease their leverage in response to tax cuts.
Discussion
This is an outstanding paper – the way to deal with endogeneity is to find a exogenous event, even better to find over a 100 of them!
The Debt and taxes relation has been studied for many years but this is the most sensible paper yet in the area.
Gives you a road map for your own research design.