Merger Negotiations with Stock Market...

Motivation Markup projections: Theory Markup projections: Evidence Bidder projections: Theory Bidder projections: Evidence

Merger Negotiations with Stock MarketFeedback

Merger Negotiations with Stock Market Feedback

Sandra Betton, Concordia UniversityB. Espen Eckbo, Dartmouth College

Rex Thompson, Southern Methodist UniversityKarin Thorburn, Norwegian School of Economics

Third Paris Spring Corporate Finance Conference, May 20, 2011

Betton, Eckbo, Thompson and Thorburn (2011)



Figure: Figure 1: Information arrival process in event time.

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15

Targ

et P

erce

ntag

e C

umul

ativ

e A

bnor

mal

Ret

urns

Days Relative to Day of Initial Takeover Bid announcement

Pre-runup period: < day -42Synergies in takeoverperceived as negligibleby the market

Runup period: day -42 to -2.Market receive a signal, s, of synergy gains and revisesprobability ¶ of takeover, creating a target stock price runup of VR

Bid event: day -1 and +1.Target receives initial bid.Market revalues target to reflect the expectedfinal bid premium VP, creating a markup ofVP - VR

Sample averagetarget runup=9%

Sample averagemarkup = 34%




Information environment

Information environment

• Market receives signal s about synergy gains S .

• S known to bidder and target. Market knows only thedistribution over S given the signal.

• Negotiations establishes a sharing rule θ for S and γ forbidding cost C .

• Rational bidding threshold: K = γCθ .

• Target benefit function: B(S ,C ) (= 0 when S < K ).

• Prior takeover probability π(0) and prior target stock pricenormalized to zero.




Model

Rational market pricing conditional on the rumor s:

• Target runup prior to the first bid announcement:

VR = π(s)Es [B(S ,C )|s, bid ] =

∫ ∞

KB(S ,C )g(S |s)dS (1)

• Expected final offer and markup at first bid announcement:

VP = Es [B(S ,C )|s, bid ] =1

π(s)VR (2)

VP − VR =1− π(s)

π(s)VR (3)




Figure: Figure 2A: Target revaluations under deal anticipation.

get V

alua

tion

V(P)

V(P) ‐ V(R)

V(R)

Expected change in target valuations during the runup (VR), throughout the entire bid process (VP), and the expected markup ( VP ‐ VR)

Benefit function has target and bidder equally sharing synergy gainss. Bidder bears a larger share of bid costs. Uncertainty in the signal, s, is uniform. Expected markup hits zero when deal is perfectly anticipated.

Revision

in Targ

Synergy signal s in the runup period




Figure: Figure 2B: Markup projections under deal anticipation.

V P‐V

R

Projection of Vp‐VR on VR

Theoretical fit of V(P)‐V(R ) on V(R )

Benefit function has target and bidder equally sharing synergy gainss. Bidder bears a larger share of bid costs.Uncertaintly in the signal, s, is uniform. Projection hits zero when deal is perfectly anticipated.

V

VR




Model

Adding a known target stand-alone value change T

• Target runup:

VRT = π(s)Es [B(S ,C )+T |s, bid ]+[1−π(s)]T = VR+T (4)

• Expected final offer and markup at first bid announcement:

VPT = Es [B(S ,C ) + T |s, bid ] = VP + T (5)

VPT − VRT =1− π(s)

π(s)[VRT − T ] (6)




Figure: Figure 3: Markup projections with stand-alone change T inrunup. Solid line (Avg.): vertical markup summation acrossdifferent Ts

Mar

kup

VPT

-VR

T

Target runnup VRT

With T = 1

Avg. V(P)-V(R )

With T = 0

Π(s) = 1




Figure: Figure 4B: Markup projections with runup feedback

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Prob

abili

ty o

f bid

Rev

isio

ns in

targ

et V

alua

tions

s

Synergy signal s

B: Valuation changes with transfer of runup

V(P)*

V(P)* - V(R)*

V(R)*

Prob of bidΠ(s) = 1.




proposition 3 (2)

Deal anticipation with runup fed back into the offer price

Proposition 3: The hypothesis that runups caused by dealanticipation are transferred from bidders to targets is rejected by azero or negative average relation between markups and runups.




Table 3: Nonlinear projections

Markup measure Runup measure Linear projection Linear res. Nonlinear res.VP − VR VR VP − VR = a+ bVR ser. corr. ser. corr.

Total markup Total runup a = 0.36 0.030 0.015OPP−2

− 1P−2

P−42− 1 b=-0.24 (−11.9) (2.36) (1.15)

Total markup Total runup a = 0.36 0.045 0.027OPP−2

− 1P−2

P−42− 1 b = −0.22 (−10.1) (3.21) (2.19)

Expected markup Total runup a = 0.31 0.027 0.016

π[ OPP−2

− 1]P−2

P−42− 1 b = −0.17 (−9.5) (2.11) (1.25)




Figure: Figure 2B: Markup projections under deal anticipation.

V P‐V

R

Projection of Vp‐VR on VR

Theoretical fit of V(P)‐V(R ) on V(R )

Benefit function has target and bidder equally sharing synergy gainss. Bidder bears a larger share of bid costs.Uncertaintly in the signal, s, is uniform. Projection hits zero when deal is perfectly anticipated.

V

VR




Figure: Figure 5A: Empirical markup projections (using offer prices)

0.7

0.8

A: Projections of total markup on total runup using offer prices:

Best Linear Fit

Best Fit of Flexible Form

0 4

0.5

0.6

0.7

mar

kup

Best Linear Fit


Raw Data

0.2

0.3

0.4

Tot

al m

arku

p

0

0.1

0.2

‐0.7 ‐0.5 ‐0.3 ‐0.1 0.1 0.3 0.5 0.7 0.9Total runup from day -42 to day -2




Figure: Figure 6A: Projections of bidder gains on target runupwithout feedback

0

0.5

1

1.5

2

2.5

0 0.5 1 1.5 2 2.5 3 3.5

Bid

der e

xpec

ted

bene

fit, ν

P

Target Runup, VR

A: Bidder does not transfer runup VR to target




Figure: Figure 6B: Projections of bidder gains on target runup withfeedback and rational bidding

Bid

der e

xpec

ted

bene

fit, ν

P

Target Runup, VR

B: Bidder transfers VR to the target but bids only on beneficial deals (alters the bid threshold K)




Figure: Figure 6C: Projections of bidder gains on target runup withfeedback but not rational bidding

Bid

der e

xpec

ted

bene

fit, ν

P

Target Runup, VR

C: Bidder transfers VR to the target but does not alter the bid threshold K (suboptimal behavior).




Table 6: Bidder regressions

Dep var: Bidder CAR[-42,1] (1) (2) (3) (4) (5) (6)

Intercept -0.116 -0.116 -0.110 -0.114 -0.097 -0.099(0.091) (0.102) (0.979) (0.102) (0.486) (0.288)

Total Target Runup 0.049 0.054

VR =P−2

P−42− 1 (0.006) (0.003)

Net Target Runup 0.078 0.082

VRT =P−2

P−42− M−2

M−42(0.000) (0.000)

Augmented Target Runup 0.049

VR = (P−2

P−42− 1) + R0 (0.006)

Market Model Target Runup 0.148VRT = CAR(−42, 2) (0.000)Control variables no yes no yes no noAdjusted R2 0.019 0.025 0.019 0.049 0.043 0.049

N 3,691 3,689 3,660 3,691 3,624 3,623




Figure: Figure 6A: Projections of bidder gains on target runupwithout feedback

0

0.5

1

1.5

2

2.5

0 0.5 1 1.5 2 2.5 3 3.5

Bid

der e

xpec

ted

bene

fit, ν

P

Target Runup, VR

A: Bidder does not transfer runup VR to target




Figure: Figure 7A: Bidder gain on target runup

0.15

0.2

A: Projections of Bidder Market Model CAR(-42, 1) target runup

Best Linear Fit


0

0.05

0.1

0.15

Mod

el C

AR

(-42

, 1)

Best Linear Fit


Raw Data

‐0.15

‐0.1

‐0.05

0

‐0.7 ‐0.5 ‐0.3 ‐0.1 0.1 0.3 0.5 0.7 0.9

Bid

der

Mar

ket M

odel

C

‐0.2

‐0.15

Target Runup from day -42 to day -2




Conclusions

Conclusions: We show that...

• With deal anticipation, projection of markups on runups isnonlinear

• Empirical projections are nonlinear and consistent with dealanticipation in the runup

• Empirical projections are inconsistent with a transfer of thetarget runup to the target

• Projections of bidder gains on target runup yield positiveslope, as predicted under deal anticipation

• Bidders raise the offer price with the market runup prior tothe initial bid

• Toehold acquisitions in the runup period fuel runups butlowers offer premiums



Merger Negotiations with Stock Market...

Documents

Transcript of Merger Negotiations with Stock Market...