Voluntary Disclosures: Evidence from the Dodd-Frank Act Anya - PowerPoint PPT Presentation

The Interaction of Mandatory and Voluntary Disclosures: Evidence from the Dodd-Frank Act Anya Kleymenova and Li Zhang March 31, 2017 PwC Young Scholars Research Symposium III, University of Illinois at Urbana-Champaign

Research Setting  Why is this important? − Reduction of systemic risk − Increased mandatory disclosure requirements − Directly affect a portion of financial institutions  Mixed theoretical predictions on voluntary disclosure for affected and unaffected banks − Positive or negative externalities − Pooling vs. separating equilibria − Decreased TBTF guarantees and increased risk of default

Institutional Setting  DFA applicability − Systemically important financial institutions (SIFIs), >$50bln − Large banks, $10-$50bln − Mid-size banks, $1-$10bln and small banks <$1bln  Main changes affecting disclosure − Stress tests − Resolutions of TBTF banks − Swap and derivatives disclosures − Corporate governance and compensation − Credit ratings

Motivation Positive benefits from disclosure  Cost of capital consequences − Decreased cost of equity capital (e.g., Glosten and Milgrom, 1985; Kim and Verrecchia, 1994; Chae, 2005) − Improved bond pricing (Morgan and Stiroh, 2001; Easton, Monahan and Vasvari, 2009; Sengupta, 1998; Balasubramnian and Cyree, 2014) − Funding providers can distinguish bank types (Balasubramnian and Cyree, 2014; Ellahie, 2016) − Increased incentives to monitor (Mehran and Millineaux, 2012)  Positive externalities − Increased voluntary disclosures of affected and non-affected banks (Ball et al., 2012; Bischof and Daske, 2013)

Motivation (cont.) Negative costs of disclosure  Cost of capital consequences − Bond prices have high sensitivity to negative news (Easton, et al., 2009)  Effects of increased mandatory disclosure for banks − Cost of inviting further regulatory scrutiny (Armstrong et al., 2016) − Real effects of disclosure (Bond et al. 2012; Gigler et al. 2014; Kanodia and Lee, 1998; Sapra 2002) − Endogenous costs of disclosure (Goldstein and Sapra, 2013; Thakor, 2015; Kashyap and Diamond, 2016)

Research Questions  Are there any spillover effects of increased mandatory disclosure on other banks?  What is the impact on voluntary disclosure of banks subject to increased mandatory disclosure requirements?  What are the resulting consequence to the quality of disclosure?

Sample  Bank-specific data − BHCs FR Y-9C quarterly reports  Financial and market data − Compustat and CRSP  Textual analysis-based data − SEC Edgar 10-Ks and 10-Qs  Management forecasts − IBES  Large banks  Medium-size banks  Small banks  Other non-regulated financial institutions and insurance companies

Empirical Research Design (Similarity) (1) SIMILARITY = β 0 + β 1 POST + β 2 SIZE + β 3 BM + β 4 LOSS + β 5 RET + β 6 RET_VOL + β 7 COVERAGE + β 8 INS_OWNERSHIP + ε Where • SIMILARITY = cosine similarity of 10Ks (MD&As) following Hanley and Hoberg (2010) • Post = 1 for 2011-2014, 0 otherwise • Size = natural logarithm of market value of equity • BM = ratio of book value of equity to market value of equity • Loss = 1 for reported negative income, 0 otherwise • RET = cumulative stock returns over prior 12 months • RET_VOL = volatility of daily stock returns over prior 12 months • COVERAGE = number of analysts covering the firm • INS_OWNERSHIP = percentage of shares held by institutional investors

Similarity Test: Large vs. Other Banks (Table 3) Dependent Variable: Dependent Variable: SIMILARITY_10K SIMILARITY_MDA (1) (2) Intercept 0.851*** 0.449*** (43.62) (11.78) POST 0.018*** 0.056*** (2.91) (3.14) SIZE -0.007** -0.009 (-2.03) (-1.23) BM -0.008 -0.014 (-1.54) (-1.32) LOSS -0.002 -0.003 (-0.34) (-0.18) RET -0.009 -0.001 (-1.31) (-0.06) RET_VOL -0.056 0.512 (0.27) (1.05) COVERAGE 0.002** 0.003 (2.20) (1.05) 0.002 -0.011 INS_OWNERSHIP (0.11) (-0.35) Year FE Yes Yes No of OBS 1,816 1,816 R-Squared 0.93% 1.09%

Empirical Research Design (Voluntary Disclosure) (2) ISSUE = β 0 + β 1 TR + β 2 POST + β 3 TR×POST + β 4 SIZE + β 5 BM + β 6 LOSS + β 7 RET + β 8 RET_VOL + β 9 COVERAGE + β 10 INS_OWNERSHIP + ε Where • Issue = 1 if issues management forecast, 0 otherwise • Post = 1 for 2011-2014, 0 otherwise • TR = 1 for large banks, 0 otherwise • TR x Post = 1 for large banks in the post period, 0 otherwise

Voluntary Disclosure: Large Banks vs. Others (Table 4) Dependent Dependent Dependent Variable: Variable: Variable: ISSUE ISSUE_GOODNEWS ISSUE_BADNEWS (1) (2) (3) Intercept -3.350*** -4.387*** -4.130*** (-8.20) (-10.25) (-8.82) TR -0.452 -0.295 -0.578 (-1.10) (-0.65) (-1.46) POST -1.019*** -1.043*** -0.592** (-5.50) (-3.65) (-2.50) TR × POST -1.369** -0.952* -1.234* (-2.40) (-1.72) (-1.87) SIZE -0.038 -0.0001 -0.053 (-0.71) (-0.01) (-0.91) BM -0.758*** -0.841*** -0.655*** (-3.29) (-3.86) (-2.65) LOSS -0.586*** -0.790*** -0.348** (-3.97) (-3.99) (-2.03) RET -0.159 -0.025 -0.297 (-1.03) (-0.15) (-1.56) RET_VOL 6.506 7.061 2.527 (1.23) (1.30) (0.45) COVERAGE 0.109*** 0.076*** 0.111*** (7.57) (5.76) (7.99) INS_OWNERSHIP 2.755*** 2.754*** 2.604*** (6.79) (6.87) (6.14) Industry/Quarter FE Yes Yes Yes No of OBS 30,552 30,552 30,552 Pseudo R-Squared 21.86% 20.09% 19.53%

Voluntary Disclosure: Medium Banks vs. Others (Table 5) Dependent Dependent Dependent Variable: Variable: Variable: ISSUE ISSUE_GOODNEWS ISSUE_BADNEWS (1) (2) (3) Intercept -3.537*** -4.541*** -4.288*** (-8.37) (-10.21) (-9.00) MID -0.231 -0.344 0.070 (-0.64) (-0.87) (0.19) POST -1.184*** -1.284*** -0.690*** (-6.07) (-4.15) (-2.81) MID × POST -0.899 -0.686 -1.261 (-1.15) (-0.86) (-1.55) SIZE -0.014 0.025 -0.033 (-0.26) (0.49) (-0.55) BM -0.720*** -0.847*** -0.588** (-3.12) (-3.79) (-2.45) LOSS -0.582*** -0.806*** -0.336* (-3.88) (-3.95) (-1.95) RET -0.145 0.020 -0.315* (-0.94) (0.13) (-1.66) RET_VOL 7.671 7.684 3.425 (1.43) (1.38) (0.61) COVERAGE 0.118*** 0.084*** 0.118*** (7.94) (6.18) (8.34) INS_OWNERSHIP 2.660*** 2.653*** 2.517*** (6.61) (6.68) (6.00) Quarter FE Yes Yes Yes No of OBS 33,082 33,082 33,082 Pseudo R-Squared 25.69% 21.73% 20.75%

Voluntary Disclosure: Forward-looking Forecast (Table 6) Dependent Variable: FLS_10Q Intercept 0.038*** (2.98) TR -0.021 (-1.61) POST 0.204*** (29.20) TR × POST -0.031*** (-3.52) SIZE -0.016*** (-7.57) BM -0.001 (-0.20) LOSS -0.005 (-1.02) RET 0.002 (0.31) RET_VOL 0.364** (2.03) COVERAGE 0.003*** (3.61) INS_OWNERSHIP 0.103*** (8.12) Industry/Quarter FE Yes No of OBS 30,552 R-Squared 26.23%

Empirical Research Design (Forecast Characteristics) (3) MF_WIDTH = β 0 + β 1 TR + β 2 POST + β 3 TR×POST + β 4 SIZE + β 5 BM + β 6 LOSS + β 7 RET + β 8 RET_VOL + β 9 COVERAGE + β 10 INS_OWNERSHIP + ε Where • MF_WIDTH = width of the annual management forecast divided by the stock price • Post = 1 for 2011-2014, 0 otherwise • TR = 1 for large banks, 0 otherwise • TR x Post = 1 for large banks in the post period, 0 otherwise

Forecast Characteristics Test: Large Banks vs. Others (Table 7) All annual Good news annual Bad news annual Dependent Variable: management management management MF_WIDTH forecasts forecasts forecasts (1) (2) (3) Intercept -0.005 -0.002 -0.007 (-1.19) (-0.47) (-1.48) TR 0.0004 -0.001 0.003 (0.12) (-0.27) (0.72) POST 0.001 0.001 0.001 (0.64) (0.50) (1.14) -0.007* -0.007* -0.008* TR × POST (-1.89) (-1.95) (-1.80) SIZE 0.001 0.001 0.001 (1.17) (0.94) (1.19) BM 0.005*** 0.003** 0.005*** (3.33) (2.45) (3.30) LOSS 0.003*** 0.002** 0.003** (2.78) (2.18) (2.46) RET -0.001 0.001 -0.003 (-0.36) (0.57) (-1.49) RET_VOL 0.334*** 0.279** 0.360*** (3.93) (2.47) (4.76) COVERAGE -0.0002 -0.0002 -0.0001 (-1.50) (-1.61) (-1.42) 0.0004 0.001 0.0003 INS_OWNERSHIP (0.18) (0.19) (0.13) Industry/Quarter FE Yes Yes Yes No of OBS 2,304 1,126 1,178 R-Squared 24.42% 21.45% 31.61%

Additional Analyses  Regression Discontinuity research design − Explore banks’ disclosure behavior around size thresholds  Bank-specific tests − Specific textual disclosures (categories, phrases and words) − Specific characteristics of disclosures (Zur, 2015) − Forward-looking − Other regulatory changes

Conclusions  DFA impact on mandatory disclosure − Evidence of spillover effects on other banks  Changes in voluntary disclosures − Decrease in voluntary disclosures for large banks − No change in voluntary disclosures of mid-size banks − Decrease in frequency of forward-looking disclosures  Changes in precision of voluntary disclosure − Decrease in the width of the range forecast

Contribution  Spillover effects to other banks from mandatory disclosure − Preliminary evidence of positive effects of mandatory disclosure regulations on other banks  Informs the debate on the consequences of increased mandatory disclosure requirements for financial institutions  First study to investigate the effect of the DFA disclosure requirements on banks voluntary disclosures

Appendix