Consensual spinoffs(work in progress)
Simon Parker (University of Western Ontario)
Mark Sanders (University of Utrecht)
Research questions
• Why do we see new venture spinoffs rather than in-house development?
• Why are some of these “intrapreneurial”…– i.e. firm ownership stake and sometimes firm
managerial involvement
• …while others are “entrepreneurial”?– i.e. employee starts a truly independent new
venture?
Slide 2 of 20 17 June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Previous literature
• New ideas emerging within incumbents are not exploited, owing to– Inflexible internal routines (Winter, 1984)– Poor fit of with core business (Cassiman & Ueda,
2006)– Different information (Klepper & Thompson, 2006)– Multi-task conflicts (Hellmann, 2007)
• Litigation and conflict– Klepper (2006), Agarwal et al (2009)
Slide 3 of 20 17 June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Motivation and Approach
• Evidence of consensual spinoffs in some industries
• Rothwell (1984): ICT and CAD; Phillips Cluster (NL)
• We explore this possibility theoretically– Using a simple principal-agent model – In which one of 3 possible outcomes can arise
endogenously:• Internal in-house development (Case A)• Intrapreneurship (Case B)• Entrepreneurship (Case C)
Slide 4 of 20 17 June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Contribution
• Three players: employee, manager & owner– Everyone risk-neutral
• Two mechanisms explain why B can replace A as the optimal mode of organization:– Double Moral Hazard– Limited Liability
• And C dominates B as well as A if double moral hazard is particularly acute in spinoff project
Slide 5 of 20 17 June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Contribution (contd.)
• We explain the spectrum of possible (consensual) organizational forms– Predict which types of ventures will end up in
which outcome• E.g. our model predicts that more radical innovations
more likely to be developed via spinoffs than commercialized in-house
– More generally, the choice for one form over another is shown to depend on the parameters in the model
Slide 6 of 20 17 June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
The model: set-up
Slide 7 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
(R&D) Worker
Manager
Owner
Case A Case B Case C
Double Moral Hazard
Slide 8 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Double Moral Hazard
€
p1,2 = (1+ e1,2
M)β1,2 (1+ e1,2
E)γ1,2 θ−β 1,2 −γ1,2
P1,2 = probability of success in project 1,2.e1,2
M,E= costly effort of (M)anager and (E)mployee in projects 1 and 2.β1,2 , γ1,2 are parameters such that β1,2 +γ1,2 < 1θ = normalization parameter such that probability < 1
Note that efforts are substitutes and free riding will occur when:1. Success of the project is a noisy signal of effort2. Effort is unobservable to all but the agents themselves…
…which we assume.
First Best
Slide 9 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Total Surplus
€
maxe1,2
M ,EV = Π j + p j (e j
M ,e jE ) Π j − Π j( ) − e j
M − e jE
( )j=1
2
∑
Π is the pay-off received in the good (upperbar) and bad (underbar) outcome.
€
e jM = β j
1−γ j
1−β j −γ j γ j
γ j
1−β j −γ j θ j
−β j −γ j
1−β j −γ j Π j − Π j( )1
1−β j −γ j −1
e jE = β j
β j
1−β j −γ j γ j
1−β j
1−β j −γ j θ j
−β j −γ j
1−β j −γ j Π j − Π j( )1
1−β j −γ j −1
Problem for the Agents
Slide 10 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Problem for the Manager and Employee (case A)
€
maxe1,2
M ,EV M ,E = wM ,E + p j (e j
M ,e jE )b j
M ,E − e jM ,E
( )j=1
2
∑
bjM,E is the bonus received in the good outcome and the other agent’s effort level is taken as given.
€
e jM (e j
E ,b jM ) = β j
1
1−β j θ j
−β j −γ j
1−β j 1+ e jE
( )
γ j
1−β j b jM
1
1−β j −1
e jE (e j
M ,b jE ) = γ j
1
1−γ j θ j
−β j −γ j
1−γ j 1+ e jM
( )
β j
1−γ j b jE
1
1−γ j −1
Spinoffs as a Solution for DMH
Slide 11 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Problem for the Owner (case A)
€
maxb j
M ,EV O = Π j + p j (b j
M ,b jE ) Π j − Π j − b j
M − b jE
( )( ) − wM − wE
j=1
2
∑
€
b jM ≈ β j Π j − Π j( )
b jE ≈ γ j Π j − Π j( )
€
e jM = β j
2−2γ j
1−β j −γ j γ j
2γ j
1−β j −γ j θ j
−β j −γ j
1−β j −γ j Π j − Π j( )1
1−β j −γ j −1
e jE = β j
2β j
1−β j −γ j γ j
2−2β j
1−β j −γ j θ j
−β j −γ j
1−β j −γ j Π j − Π j( )1
1−β j −γ j −1
(Double) Moral Hazard reduces induced effort levels below optimal.Given that marginal productivity of both is bounded (not infinite as effort drops to 0)……full specialization may be preferred to joint development.
But why spinoff?
Slide 12 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
…Cases B and C eliminate the problem of double moral hazard in project 1 and both projects, respectively……but the owner can also force (full) specialization by setting bonuses (to 0)……so double moral hazard is not sufficient to explain spinoffs.
An additional benefit of spinoff might be the limited liability for the owners.
Assume:Owners take an equity share in spinoffs.Equity cannot have a negative value.This Limited Liability then reduces the downside risk, but……of course creditors will demand compensation for the default risk……reducing the return in the good state as well.
Limited Liability
Slide 13 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
€
V B = Π1 + p1(e1M ,0) Π1 − Π1( ) − e1
M − e2M − e2
E
+max 0,Π 2[ ] + p2(e2M ,e2
E ) Π 2 +1− p2
p2
min Π 2,0[ ] − max 0,Π 2[ ] ⎛
⎝ ⎜
⎞
⎠ ⎟
is the markets expectation of the spinoff’s success probability.risk neutral creditors require a premium CC to compensate for default risk:for a negative lower support pay-off in case of failure is 0 and in case of success is:
€
p2
€
p2CC + (1− p2)Π 2 = 0
€
Π2 − CC = Π 2 +1− p2
p2
Π 2
Problem for the Agents
Slide 14 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Problem for the Manager (case B):
€
maxe1,2
MV M = wM + p1(e1
M ,0)b1M
+ p2(e2M ,e2
E )b2M − e1
M − e2M
€
e1M (b1
M ) = β1
1
1−β 1θ−β1 −γ1
1−β 1 b1M
1
1−β 1 −1
e2M (e2
E ,b2M ) = β 2
1
1−β 2 θ−β 2 −γ 2
1−β 2 1+ e2E
( )γ 2
1−β 2 b2M
1
1−β 2 −1
Problem for the Agents
Slide 15 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Surplus for the Employee (case B):
€
maxe2
EV E = p2(e2
M ,e2E )(1−α B )Π 2
LL − e2E
and αB is the equity share taken by the parent firm in the spinoff.
€
e2E (e2
M ,α B ) = γ 2
1
1−γ 2 θ−β 2 −γ 2
1−γ 2 1+ e2M
( )β 2
1−γ 2 (1−α B )Π 2LL
( )1
1−γ 2 −1
Where
€
Π2LL ≡ Π 2 +
1− p2
p2
Π 2
Spinoffs to exploit LL
Slide 16 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Surplus for the Owner (case B):
€
maxb1,2
M ,α BV O = Π1 + p1(b1
M ) Π1 − Π1 − b1M
( )
−wM + p2(b2M ,α B ) α B Π 2
LL − b2M
( )
€
b1M ≈ β1 Π1 − Π1( )
b2M ≈ β 2Π 2
LL
α B ≈1− γ 2
Bonuses still set proportional to the spread between best and worst outcome.LL reduces spread for project 2 (when lower support <0).No bonus and/or wage but equity share paid to employee.
Limited Liability
Slide 17 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Induced Efforts (case B)
€
e1M = β1
2
1−β1 θ1
−β1 −γ1
1−β1 Π1 − Π1( )1
1−β1 −1
e2M = β 2
2β 2
2β 2
1−β 2 −γ 2 γ 2
2γ 2
1−β 2 −γ 2θ−β 2 −γ 2
1−β 2 −γ 2 Π 2LL
( )1
1−β 2 −γ 2 −1
e2E = γ 2
2β 2
2β 2
1−β 2 −γ 2 γ 2
2γ 2
1−β 2 −γ 2θ−β 2 −γ 2
1−β 2 −γ 2 Π 2LL
( )1
1−β 2 −γ 2 −1
Note that induced efforts are lower than optimal in project 2 due to DMH and LL.
Limited Liability
Slide 18 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Similarly we can find Induced Efforts (case C)
€
e1M = β1
2
1−β1 θ−β1 −γ1
1−β1 Π1 − Π1( )1
1−β1 −1
e2E = γ 2
2
1−γ1 θ−β 2 −γ 2
1−γ 2 Π 2LL
( )1
1−γ 2 −1
Note that now induced efforts are lower than optimal in project 2 due to LL only.
The Owners’ Choice
Slide 19 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
The owners will set bonuses and equity shares such that……they maximize their surplus……and incentive compatibility constraints are satisfied……and participation constraints are satisfied.
The Owners’ Choice
Slide 20 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
€
V OA = Π1 − wM + (1− β1 − γ1) θ−β 1 −γ1 Π1 − Π1( )β12β1γ1
2γ1
( )
1
1−β1 −γ1
Π 2 − wE + (1− β 2 − γ 2) θ−β 2 −γ 2 Π 2 − Π 2( )β 22β 2γ 2
2γ 2
( )
1
1−β 2 −γ 2
€
V OB = Π1 − wM + (1− β1) θ−β 1 −γ1 Π1 − Π1( )β12β1
( )
1
1−β 1 +
(1− β 2 − γ 2) θ−β 2 −γ 2 Π 2LL
( )β 22β 2γ 2
2γ 2
( )
1
1−β 2 −γ 2
< >
The Owners’ Choice
Slide 21 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
VOA<VOB
Proposition 1: It is more likely that the owner chooses B over A if:
The outside option for employees is larger (obvious)
The spread for project 1 is larger (DMH)
The lower support for project 2 is more negative (LL)
The upper support for project 2 is less positive (DMH)
The market success probability for project 2 is higher. (LL)
The Owners’ Choice
Slide 22 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
€
V OB = Π1 − wM + (1− β1) θ−β 1 −γ1 Π1 − Π1( )β12β1
( )
1
1−β 1 +
(1− β 2 − γ 2) θ−β 2 −γ 2 Π 2LL
( )β 22β 2γ 2
2γ 2
( )
1
1−β 2 −γ 2
€
V OC = Π1 − wM + (1− β1) θ−β 1 −γ1 Π1 − Π1( )β12β1
( )
1
1−β 1 +
(1− γ 2) θ−β 2 −γ 2 Π 2LL
( )γ 22γ 2
( )
1
1−γ 2
< >
The Owners’ Choice
Slide 23 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
VOC >VOB
Proposition 3: It is more likely that the owner chooses C over B if:
The lower support for project 2 is more negativeThe upper support for project 2 is less positiveThe market estimate of success probability for project 2 is lower.
We also show that VOA>VOB implies VOA>VOC for VOB < > VOC.The value of β2 is also an important driver of the choice between C and B……and can be tied to the double moral hazard problem.
Note that for β2 = 0 and for β2 = 1-γ2 the owner would always prefer case C…… but B can be preferred for values between these extremes.
Conclusions
Slide 24 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Proposal
Case A: In House Not A: Consensual Spinoff
Case B: Intrapreneurial Case C: Entrepreneurial
Weak DMH; Positive or small negative downside risk
Strong DMH project 1; Large negative downside risk
Intermediate DMH; Small negative downside risk
Strong or Weak DMH project 2; Large negative downside risk
Implications and Limitations
Slide 25 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Our model predicts that:Firms may strategically decide to allow spinoffs particularly when:
Radical InnovationsYoung Industries
Our model implies that:Governments may want to facilitate and stimulate consensual spinoffs in:
Entrepreneurial EconomiesAt the Technology Frontier
In light of our results:Firms may want to revise their strategy on IPR and R&D.Governments may want to revise Limited Liability and Labor Regulation.
Implications and Limitations
Slide 26 of 20 17June 2010
Third Searle Entrepreneurship Symposium, Chicago IL
Our model at this point……cannot explain conflictual Spinoffs.…has modeled limited liability in a particular (non GE) way.…has complex comparative statics in β and γ.
…is a useful starting point for a more unified theory.…uses well-known modeling tools and concepts and can thereby be firmly embedded in the extant literature.…can (and should) be tested more formally in the data.
Your comments, questions and discussion are most appreciated……and thank you for your kind attention.