Presentation is loading. Please wait.

Presentation is loading. Please wait.

S TAGGERED B OARDS AND F IRM V ALUE, R EVISITED Martijn Cremers (University of Notre Dame) Lubomir P. Litov (Univ. of Arizona & WFC, Univ. of Pennsylvania)

Similar presentations


Presentation on theme: "S TAGGERED B OARDS AND F IRM V ALUE, R EVISITED Martijn Cremers (University of Notre Dame) Lubomir P. Litov (Univ. of Arizona & WFC, Univ. of Pennsylvania)"— Presentation transcript:

1 S TAGGERED B OARDS AND F IRM V ALUE, R EVISITED Martijn Cremers (University of Notre Dame) Lubomir P. Litov (Univ. of Arizona & WFC, Univ. of Pennsylvania) Simone M. Sepe (Univ. of Arizona & IAST-TSE) University of Nebraska, September 29 th 2014

2 2 Staggered Boards (SBs): –boards with 3 classes of directors. –one class standing for re-election each year, each serving 3-year terms. Quintessential corporate governance failure or Strengthening board commitment to long-term value creation? T HE S TAGGERED B OARD C ONTROVERSY

3 Empirical literature to date documents that firms with SBs have lower financial value (e.g., Bebchuk & Cohen, 2005; Faleye, 2007; ISS, 2013; Cohen and Wang, 2013) –Purely cross-sectional result. –Literature interprets this as SBs reducing firm value. Our paper challenges this interpretation using data. –Finding that staggering up (down) is associated with increase (decrease) in financial value (proxied by Q). 3 3 N O C ONTROVERSY IN E MPIRICAL L ITERATURE

4 I.T HE T RADITIONAL (D ELAWARE ) V IEW Representative Democracy model (Strine, 2006)  Director Primacy  With accountability mechanisms  Shareholders: less expertise, incentives, information II. T HE S HAREHOLDER E MPOWERMENT V IEW Direct Democracy (Bebchuk, HLR 2005)  Shareholder empowerment T HEORETICAL B ACKGROUND

5 R EPRESENTATIVE VS. D IRECT D EMOCRACY G OVERNANCE R OME G OVERNANCE VS.A THENS G OVERNANCE

6 66 6 S EPARATION OF O WNERSHIP AND C ONTROL Gives rise to twin ‘Agency Problems’: Moral Hazard (of agent) Due to management–shareholder conflict of interest. Addressed by Shareholder Empowerment View. Limited Commitment of shareholders to defer judgment. Due to asymmetric information + adverse selection. Addressed by Director Primacy View. Trade-off Between market discipline and board decision making. Not ‘one-size-fits-all’ corporate governance. T HEORY : W HICH A GENCY P ROBLEM ?

7 7 A N O NGOING D EBATE  Lipton, Wachtell, Rosen & Katz, February 2013 Memo, “The Shareholder Rights Project Is Wrong” (The New York Times): “It is surprising that a major academic institution would countenance the formation of a clinical program to advance a narrow agenda that would exacerbate the short-term pressures under which American companies are forced to operate.”  Professor Bebchuk, CLR 2013: “None of the organizations that press for board insulation in the name of long- term value …, such as … Wachtell, Lipton, Rosen & Katz, have thus far attempted to conduct or commission research that would use the substantial data available on the financial performance of firms and shareholders to validate their myopic activists hypothesis.”

8 S HAREHOLDER R IGHTS P ROJECT

9 Staggered Boards –1978–1989 from Cremers & Ferrell (2013) database, & hand-collected information.hand-collected information –1990–2011 from Risk Metrics, previously Investor Responsibility Research Center (IRRC). »Hand-checked missing years in the 1993–2006 using proxy statements (SEC’s EDGAR). –Charter vs. Bylaws Staggered Boards. –Bebchuk and Cohen (2005); Cremers and Ferrell (2013). Firm Value –Q (Compustat). –Stock Return (CRSP). D ATA

10 Year#% #% , , , , , , , , , , , , , , Total31, D ATA - O BSERVATIONS

11 A LMOST N O T IME - SERIES V ARIATION IN

12 V ARIATION OF SB IN

13 Category## (2011) No SB SB in No SB SB in No SB SB in

14 Confirm previous cross-sectional results:  Negative association between SBs and firm value. Reversed in time-series  Controlling for firm fixed effects, association is positive (levels analysis).  Average firm value after staggering up (down) is higher (lower). SB S AND F IRM V ALUE

15 C ROSS -S ECTIONAL A NALYSIS Dep. Variable is Q [t] (1)(2)(3)(4)(5) Period: Staggered Board [t-1] ** ** (firm cluster)(2.38)(0.51)(2.52)(1.06)(1.17) [no cluster][4.98][0.96][4.33][1.94][1.83] G-Index [t-1] (0.57) Ln (Assets) [t-1] *** *** *** (3.74)(3.33)(0.44)(4.11)(3.24) Delaware Incorporation [t-1] (0.76)(1.26)(0.49)(0.31)(0.28) ROA [t-1] *** 2.7 *** *** *** *** (32.74)(15.98)(22.75)(24.06)(19.11) CAPX/Assets [t-1] ** (1.14)(1.23)(1.22)(0.58)(2.17) R&D/ Sales [t-1] *** *** *** *** *** (12.01)(5.19)(10.31)(9.33)(7.17) Industry M&A Volume [t-1] *** *** (3.04)(0.10)(0.92)(2.74)(0.85) N31,5748,5009,61713,4575,253 Adjusted R-Squared

16 C ROSS -S ECTIONAL A NALYSIS : C ONTROLLED FOR Q Dep. Variable is Q [t] (1)(2)(3)(4)(5) Period: Q [t-1] *** *** *** *** *** (91.25)(39.13)(38.52)(66.69)(32.55) Staggered Board [t-1] (firm cluster)(1.35)(0.15)(0.69)(1.13)(0.99) [no cluster][1.40][0.14][0.68][1.06][1.02] G-Index [t-1] (0.06) Ln (Assets) [t-1] ** *** ** (1.43)(2.36)(1.28)(2.78)(2.27) Delaware Incorporation [t-1] (0.99)(0.65)(0.98)(0.13)(0.03) ROA [t-1] *** 0.28 *** *** *** *** (9.63)(3.38)(6.94)(5.53)(5.74) CAPX/Assets [t-1] *** *** ** *** *** (4.16)(4.54)(2.25)(3.48)(3.04) R&D/ Sales [t-1] *** *** *** *** *** (5.24)(4.76)(5.54)(3.63)(3.70) Industry M&A Volume [t-1] * * (1.87)(0.30)(1.81)(1.11) N 28,464 7,4798,59612,3895,027 Adjusted R-Squared

17 T IME -S ERIES A NALYSIS Dep. Variable is Q [t] (1)(2)(3)(4)(5) Period: Variables Staggered Board [t-1] ** ** * (firm cluster)(2.11)(1.26)(0.11)(2.19)(1.82) [no cluster][4.65][2.35][0.17][3.43][2.15] G-Index [t-1] (0.33) Ln (Assets) [t-1] *** *** *** *** *** (12.01)(4.55)(4.65)(11.04)(8.10) ROA [t-1] *** *** 2.79 *** *** *** (20.27)(10.5)(11.34)(8.39)(7.74) CAPX/Assets [t-1] ** ** (0.60)(0.80)(2.36)(0.26)(2.19) R&D/ Sales [t-1] *** *** (2.72)(1.26)(3.41)(0.95)(0.35) Industry M&A Volume [t-1] *** * (3.59)(0.45)(1.84)(1.56)(0.93) N31,5748,5009,61713,4575,253 Adjusted R-Squared

18 Confirm previous cross-sectional results:  Negative association between SBs and firm value. Reversed in time-series:  Controlling for firm fixed effects, association is positive (levels analysis).  Average firm value after staggering up (down) is higher (lower).  In changes, association is also positive (first difference analysis).  Firm value increases (decreases) after staggering up (down). SB S AND F IRM V ALUE

19 F IRST D IFFERENCE A NALYSIS Dep. Variables:∆ Q [t, t+1] ∆ Q [t, t+2] ∆ Q [t, t+3] ∆ Q [t, t+4] ∆ Q [t, t+5] Variables(1)(2)(3)(4)(5) ∆ Staggered Board [t-1, t] ** ** ** *** ** (2.16)(2.11)(2.37)(3.22)(2.22) ∆ Ln(Assets) [t-1, t] *** *** *** *** *** (15.35)(21.56)(23.55)(22.50)(20.69) ∆ ROA [t-1, t] *** *** 1.42 *** *** *** (22.70)(16.48)(12.69)(10.54)(9.71) ∆ CAPX/Assets [t-1, t] ** *** *** *** *** (1.99)(6.96)(6.94)(8.40)(6.64) ∆ R&D/ Sales [t-1, t] ** * (1.97)(1.29)(1.63)(1.56)(1.93) ∆ Industry M&A Volume [t-1, t] *** *** *** *** (6.07)(3.29)(2.67)(1.52)(2.94) Sample Period (years) # of firms in regression2,8862,7662,5972,4562,311 N29,16628,00425,87523,86021,954 Adjusted R-Squared

20 F IRST D IFFERENCE A NALYSIS – M ATCHED S AMPLE Dep. Variable:∆ Q [t-1, t] ∆ Q [t-1, t+1] ∆ Q [t-1, t+2] ∆ Q [t-1, t+3] ∆ Q [t-1, t+4] Variables(1)(2)(3)(4)(5) ∆ Staggered Board [t-1, t] ** ** ** *** ** (2.54)(2.16)(2.38)(3.08)(2.27) ∆ Assets [t-1, t] *** *** *** *** *** (2.92)(3.81)(2.98)(4.05)(4.38) ∆ ROA [t-1, t] *** *** ** * (6.60)(4.00)(2.45)(1.87)(1.28) ∆ CAPX/Assets [t-1, t] ** ** (1.39)(1.53)(1.65)(2.12)(2.42) ∆ R&D/ Sales [t-1, t] (0.71)(1.58)(1.14)(1.01)(0.57) ∆ Industry M&A Volume [t-1,t] (1.50)(0.17)(0.05)(0.48)(0.36) Sample Period (Years) N1,2051, Adjusted R-Squared

21 Confirm previous cross-sectional results:  Negative association between SBs and firm value. Reversed in time-series:  Controlling for firm fixed effects, association is positive (levels analysis).  Average firm value after staggering up (down) is higher (lower).  In changes, association is also positive (first difference analysis).  Firm value increases (decreases) after staggering up (down).  Also, portfolio analysis has a positive alpha SB S AND F IRM V ALUE

22 S TAGGERING & D E -S TAGGERING P ORTFOLIO EW R ETURNS Portfolio “6m12”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) ** * ** (2.04)(0.19)(0.95) (1.72)(0.05)(1.05) (2.57)(0.43)(1.13) Alpha (Annual) 6.37%0.75%5.11% 5.43%-0.19%5.50% 9.22%1.71%5.90% Portfolio “12m12”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) ** ** * *** (1.54)(1.08)(2.24)(1.13)(1.59)(2.47)(1.85)(0.93)(2.65) Alpha (Annual) 6.54%-3.46%15.87%4.76%-4.98%16.71%7.20% %16.30% Portfolio “12m24”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) ** * *** * (2.30)(0.17)(1.44) (1.65)(0.31)(1.45) (2.7)(0.18)(1.68) Alpha (Annual) 4.92%0.47%5.15% 3.56%-0.80%4.99% 6.49%0.47%5.67%

23 S TAGGERING AND D E -S TAGGERING P ORTFOLIO VW R ETURNS Portfolio “6m12”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) (0.01)(0.43)(0.53) (0.13)(0.56)(0.58) (0.02)(0.4)(0.51) Alpha (Annual) -0.05%-1.57%3.08% -0.56%-2.03%3.39% -0.07%-1.47%3.05% Portfolio “12m12”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) ** ** ** (0.58)(1.28)(2.35)(0.34)(1.54)(2.34)(0.68)(1.58)(2.57) Alpha (Annual) 2.81%-4.11%17.64%1.51%-4.67%16.25%2.82%-4.88%17.32% Portfolio “12m24”Four Factors ModelThree Factors Model Market Factor Model LongShort Long - ShortLongShort Long - ShortLongShort Long - Short Alpha (Monthly) (0.04)(0.69)(0.50) (0.28)(0.53)(0.17) (0.12)(0.67)(0.47) Alpha (Annual) -0.10%-1.99%1.86% -0.72%-1.54%0.65% 0.29%-1.96%1.80%

24 24 How to reconcile time-series with cross-sectional evidence?  Possible explanation is “reverse causality”  Having relatively low firm value could induces some firms to adopt a SB (rather than a SB causing low firm value).  Could explain cross-sectional result that firms with SBs tend to have low firm values.  However, reverse causality cannot explain the time series results, as this analysis shows that firm value tends to go up after the adoption of a SB W HICH D IRECTION OF C AUSALITY ?

25 P REDICTING M ODELS : S TAGGERING Dep. Variable is: Pr (Stagger in period t)Random EffectsCox Proportional Probit ModelHazard Model Variables(1)(2) Q [t-1] *** *** (3.84)(7.63) Ln (Assets) [t-1] * (0.47)(1.69) Delaware Incorporation [t-1] ** (0.54)(2.14) ROA [t-1] * *** (1.79)(5.05) Capital Expense/Assets [t-1] ** ** (2.29)(2.45) R&D/ Sales [t-1] ** (2.44)(1.54) Industry M&A Volume [t-1] *** (3.21)(0.23) Percentage Effect (i.e., Economic Significance)-35.1%-57.8% N15,35914,535 # of firms in regression1,7841,651

26 P REDICTING M ODELS : D E -S TAGGERING Dep. Variable is: Pr (De-Stagger in period t)Random EffectsCox Proportional Probit ModelHazard Model Variables(1)(2) Q [t-1] (0.68)(1.48) Ln (Assets) [t-1] *** *** (12.70)(6.22) Delaware Incorporation [t-1] (1.73)(0.14) ROA [t-1] (1.35)(0.46) Capital Expense/Assets [t-1] (1.45)(0.48) R&D/ Sales [t-1] (0.15)(0.09) Industry M&A Volume [t-1] (1.55)(0.42) Percentage Effect (i.e., Economic Significance)-6.2%-14.4% N17,36813,462 # of firms in regression1,8131,494

27 27 How to interpret positive association of SBs & firm value?  Managerial entrenchment view of SBs  Based on assumption that more shareholder power is always better (to reduce moral hazard)  However, in contexts of high asymmetric information or very noisy market prices, shareholders may not know best  Asymmetric information / adverse selection, rather than moral hazard, may be primary agency problem for some firms  SBs protect directors from shareholder pressure  Avoid myopia and promote long-term value creation A P OSITIVE A CCOUNT

28 Hypothesis: If limited commitment (asymmetric info + adverse selection) is the primary problem (rather than market discipline of the board), then: SBs may be more valuable in firms with (i) more intangible assets, firms that are (ii) more innovative, and (iii) more complex firms.  These features make valuation of corporate assets / performance more difficult, especially in the short-term Our results confirm this hypothesis:  Value-changes associated with changes in SBs:  Stronger / driven by firms with these features SB S & F IRM F EATURES

29 V ALUE OF SB S & A DVERSE S ELECTION Dep. Variable is Q [t] (1)(3)(5)(7) Staggered Board [t-1] ** ** (2.44)(2.45)(0.68)(1.28) R&D/ Sales [t-1] (0.56) Intangible Assets/ Total Assets [t-1] (1.64) Ranked Patent Citation Count [t-1] (0.38) Firm Sales [t-1] (0.88) R&D/ Sales [t-1] *Staggered Board [t-1] ** (2.54) Intangible/Total Assets [t-1] *Staggered Board [t-1] *** (2.86) Ranked Patent Citations [t-1] *Staggered Board [t-1] * (1.92) Firm Sales t-1] * Staggered Board [t-1] ** (2.34) Economic Significance (Staggered B.)4.49%4.61%1.71%2.34% Economic Significance (Int. Variable)10.8%3.96%4.97%3.44% N31,57431,33715,33831,558

30 Hypothesis: If limited commitment (asymmetric info + adverse selection) is the primary problem (rather than market discipline of the board), then: SBs may be more valuable in firms with (i) more productive labor, firms with (ii) more contractual imcompleteness (relationality), and firms with (iii) large customers. Our results confirm this hypothesis:  Value-changes associated with changes in SBs:  Stronger / driven by firms with these features F URTHER T ESTS OF L IMITED C OMMITMENT P ROBLEM

31 L ONG -T ERM C OMMITMENT H YPOTHESIS (1)(4)(5) Staggered Board [t-1] ** (2.20)(0.91)(1.45) Staggered Board [t-1] * Labor Productivity [t-1] *** (3.74) Labor Productivity [t-1] *** (8.31) Staggered Board [t-1] *Relativity [t-1] (1.62) Relativity [t-1] ** (2.01) N 30,79724,8809,628 Adjusted R-Squared

32 L ONG -T ERM C OMMITMENT H YPOTHESIS (C ONTINUED ) (1)(2)(3)(4) Staggered Board [t-1] * (1.45)(1.50)(1.54)(1.66) Staggered Board [t-1] *Large Customer 10% [t-1] ** (2.38) Large Customer 10% [t-1] *** (3.26) Staggered Board [t-1] *Large Customer 20% [t-1] *** (2.69) Large Customer 20% [t-1] *** (3.62) Staggered Board [t-1] *Large Customer 30% [t-1] *** (3.34) Large Customer 30% [t-1] *** (4.81) Staggered Board [t-1] *Large Customer 40% [t-1] *** (3.53) Large Customer 40% [t-1] *** (4.33) N 31,574 Adjusted R-Squared 0.715

33 Alternative Hypotheses: (i)Maybe firms with SBs simultaneously decrease other pro- management governance features and this results in higher value? (ii)Does having a SB alter the structure and level of executive compensation (increased board capture hypothesis)? (iii)Does SB changes level of CEO turnover? Results: (i)Changes in firm value before versus after the adoption of a SB independent of the level of shareholder rights at the firm. (ii)Having a SB produces more efficient executive compensation, for example providing better risk-taking incentives. (iii)CEO turnover is not different for firms with/without SB. SB S & O THER G OVERNANCE F EATURES

34 G OVERNANCE P ROVISIONS Dep. Variable is Q [t] (1)(2)(3)(4) Variables: Staggered Board [t-1] * *** ** (1.94)(2.97)(2.57)(2.58) CEO-Board Chairman Duality [t-1] ** (0.58)(1.96) Governance Index [t-1] ** ** (2.40)(2.44) CEO-Board Chairman Duality [t-1] * Staggered Board [t-1] *** - (2.81)- Governance Index [t-1] *Staggered Board [t-1] (0.64)- Economic Significance (Staggered B.)6.26%-4.86%- Economic Significance (Int. Variable)5.10%-1.28%- Sample Period (Years) N 18,552 23,525 Adjusted R

35 E XECUTIVE C OMPENSATION Dep. Variable is Q [t] (1)(2)(3)(4)(5)(6) Variables: Staggered Board [t-1] ** *** ** *** ** *** (2.58)(2.63)(2.54)(3.05)(2.53)(2.85) CEO Delta [t-1] *** *** (4.66)(7.60) CEO Vega [t-1] (1.28)(0.58) CEO Total Compensation [t-1] *** *** (4.29)(8.87) CEO Delta [t-1] * Staggered Board [t-1] (0.57) CEO Vega [t-1] * Staggered Board [t-1] *** (2.82) CEO Total Comp [t-1] * Stagg. Board [t-1] ** (2.44) Economic Significance (Staggered B.)6.62%-7.26%-5.97%- Economic Significance (Int. Variable)1.79%-5.53%-5.43%- Sample Period (Years) N17,573 15,983 17,965 Adjusted R

36 CEO T URNOVER Dependent Variable: Prob (Forced CEO Turnover [t] ) Prob (Forced CEO Turnover [t] ) Prob (CEO Turnover [t] ) Prob (CEO Turnover [t] ) (1)(2)(3)(4) Staggered Board [t] (0.77)(0.45)(0.12)(0.06) Excess Returns [t] *** *** *** *** (7.23)(3.62)(5.91)(3.74) Staggered Board [t] * Excess Returns [t] (0.22)-(0.25) Poison Pill [t] * 0.01 * (1.88)(1.9)(1.74)(1.75) Delaware Incorporation [t] (0.97) (0.27) N9,519 Pseudo R Number of events Sample Period

37 Separating SB provisions  Charter v. Bylaws. Around 90% of SB are Charter-based. Economic significance is higher. SB bylaws based (statistically) insignificant. Controlling for M&A activities. Other proxies for asymmetric information (e.g., idiosyncratic volatility). O THER R OBUSTNESS C HECKS

38 38 SB: C HARTER V. B YLAWS C ROSS -S ECTIONAL A NALYSIS Dep. Variable: Q [t] Variables(1)(2)(3)(4)(5)(6) Staggered Board [t-1] ** ** * (0.36) (2.20)(2.47)(1.85) Staggered Board – Charter [t-1] ** (0.37) (2.17) Staggered Board – Bylaws [t-1] * (0.11)(0.14)(1.89)(0.01) Pre-1986 Indicator x Staggered Board [t-1] * (1.65)(0.11) Year fixed effects and other controls of Table 3 included Yes N 23,793 Adjusted R-Squared Firm Fixed EffectNo Yes NoYes Industry Fixed EffectYes No YesNo

39 39 We investigate 34 years of SBs data and find that, over time, staggering up (down) is associated with increases (decreases) in firm value.  First, we challenge existing cross-sectional evidence and interpretation on staggered boards  Second, we document that in the time series, staggered boards increase firm value: (i) SBs valuable to protect commitments to long-term value creation, especially when asymmetric information is important; (ii) Traditional board-centric model can efficiently serve the interests of shareholders? C ONCLUSIONS

40 Reducing Federal intervention on Corporate Law? How constraining managerial moral hazard?  Executive compensation?  Fiduciary duties? Bias in shareholder advisory services? D IRECTIONS FOR F UTURE L EGAL R ESEARCH

41 Data Checks

42 Hand Collection, –53 cases of incomplete (late) board classification in Risk Metrics: –128 cases of incomplete (late) board de-classification in Risk Metrics: D ATA C HECKS IN R ISK M ETRICS, YEARCOUNT YEARCOUNT

43 Examples incomplete classification in Risk Metrics: D ATA C ONSISTENCY – R ISK M ETRICS CompanyIRRC Date/YearActual Date Classify Bed Bath & Beyond199806/27/1997 City National Corp199804/16/1996 Progressive199804/25/1997 Charles Schwab Corp199805/06/1996 Whole Foods Market Inc /30/1998 Harmon Industries Inc /10/1998 Allegheny Energy Inc /15/1999 Qwest Communications200205/02/2001 WM Wrigley Jr Co200408/15/2001 Eagle Materials Inc /08/2004

44 Examples incomplete de-classification in Risk Metrics: Examples D ATA C ONSISTENCY – R ISK M ETRICS CompanyIRRC Date/YearActual Date De-Classify Ameritech Corp199804/17/1996 Foamex International Inc /23/1996 Mead Corp199804/25/1996 Occidental Petroleum Corp199804/25/1997 Time Warner Inc /15/1997 Travelers Group Inc /23/1997 Union Pacific Corp199804/19/1996 Apple Computer Inc /24/1999 Walt Disney Co200002/24/1998 Morrison Knudsen Corp200004/18/1998 Smurfit Stone Container200005/27/1999 Home Depot Inc /31/2000 Coca Cola Bottling Co200405/07/2003 Dow Jones & Co Inc /19/2003 Hasbro Inc /14/2013 Iomega Corp200405/20/2013 Pfizer Inc /24/2003 Sprint Corp200405/13/2003 Waste Management Inc /17/2002 Allegheny Energy Inc /13/2004 AT&T Inc /30/2004 Bristol Myers Squibb200605/06/2003 Eastman Kodak Co200604/27/2004 Fedex Corporation200609/27/2004 Host Marriott Corp200605/20/2004 Merck & Co200604/27/2004


Download ppt "S TAGGERED B OARDS AND F IRM V ALUE, R EVISITED Martijn Cremers (University of Notre Dame) Lubomir P. Litov (Univ. of Arizona & WFC, Univ. of Pennsylvania)"

Similar presentations


Ads by Google