Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS...
Transcript of Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS...
![Page 1: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/1.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
On the bosonic nature of business opportunities
Giulio Bottazzi
Scuola Superiore Sant’Anna
ISS Conference- July 2014 - Jena
Bosonic opportunities
![Page 2: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/2.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Who’s this guy?
Peter Higgs, Nobel Prize laureate for the theoretical prediction in 1969 ofthe Higgs’ boson whose existence “recently was confirmed through thediscovery of the predicted fundamental particle, by the ATLAS and CMSexperiments at CERN’s Large Hadron Collider"Why Higgs decided to call his particle “boson”?
Bosonic opportunities
![Page 3: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/3.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Who’s this guy?
Peter Higgs, Nobel Prize laureate for the theoretical prediction in 1969 ofthe Higgs’ boson whose existence “recently was confirmed through thediscovery of the predicted fundamental particle, by the ATLAS and CMSexperiments at CERN’s Large Hadron Collider"
Why Higgs decided to call his particle “boson”?
Bosonic opportunities
![Page 4: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/4.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The origins of “bosons”
Actually he did not decide. He was this guy who decided, many yearsbefore. Paul A.M. Dirac. According to quantum theory sub-atomicparticles are divided in two “groups”: Fermions and Bosons. Thename “boson” was in honour of Satyendra Nath Bose
Bosonic opportunities
![Page 5: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/5.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The origins of “bosons”
Actually he did not decide. He was this guy who decided, many yearsbefore. Paul A.M. Dirac. According to quantum theory sub-atomicparticles are divided in two “groups”: Fermions and Bosons. Thename “boson” was in honour of Satyendra Nath Bose
Bosonic opportunities
![Page 6: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/6.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The origins of “bosons”
Actually he did not decide. He was this guy who decided, many yearsbefore. Paul A.M. Dirac. According to quantum theory sub-atomicparticles are divided in two “groups”: Fermions and Bosons. Thename “boson” was in honour of Satyendra Nath Bose
Bosonic opportunities
![Page 7: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/7.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Bosonic and fermionic behavior
Bosons and fermions follow peculiar statistical rules when theydistribute in the “energy levels”: Bose-Einstein and Fermi-Diracstatistics.
The difference of Bose-Einstein statistics with respect to classical(macroscopic) objects is better explained with an example.
Bosonic opportunities
![Page 8: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/8.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The simplest case
Two balls have to be assigned to two bins.
A B
Bosonic opportunities
![Page 9: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/9.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Assign the 1st ball
We assume that the probability for the two bins is the same
A B
A B A B
1/2 1/2
Bosonic opportunities
![Page 10: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/10.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Assign the 2nd ball
Again the probability for the two bins is the same
A B
A B A B
A BA B A B A B
1/2 1/2
1/4
1/2 1/2 1/2
1/4 1/4 1/4
1/2
Bosonic opportunities
![Page 11: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/11.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Probability of different occupancies
Occupancy = way of distributing balls in bins. Classic balls: the evenoccupancy is the most probable. Bose-Einstein: All occupancies areequally probable.
P{ }=
Classical
1/4
P{ }=
1/2P{ }=
1/4
A B
A B
A B
Bosonic opportunities
![Page 12: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/12.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Probability of different occupancies
Occupancy = way of distributing balls in bins. Classic balls: the evenoccupancy is the most probable. Bose-Einstein: All occupancies areequally probable.
P{ }=
Classical Bose-Einstein
1/4 1/3
P{ }=
1/2 1/3P{ }=
1/4 1/3
A B
A B
A B
Bosonic opportunities
![Page 13: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/13.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Distribution of the number of bins
Classical: the highest probability associated with the meanassignnment. Bose-Einstein: higher probability assigned to the“extreme” events.
P{ }=
Classical Bose-Einstein
1/4 1/3
P{ }= 1/2 1/3
P{ }= 1/4 1/3
Bosonic opportunities
![Page 14: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/14.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many bins and balls
Let N be the number of bins and M the number of balls. When N andM becomes large: Bose-Einstein→ Geometric and Binomial→Normal
Bosonic opportunities
![Page 15: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/15.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many bins and balls
Let N be the number of bins and M the number of balls. When N andM becomes large: Bose-Einstein→ Geometric and Binomial→Normal
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0 50 100 150 200
N=100 M/N=100
Bose-EinsteinGeometric
BinomialNormal
Bosonic opportunities
![Page 16: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/16.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Pharma industry NCEs
New Chemical Entity (NCE): new molecules with novel therapeuticalproperties
The total number of NCEs introduced over the period 1975-1994 is154.
In G.Bottazzi, G.Dosi, M.Lippi, F.Pammolli and M.Riccaboni,International Journal of Industrial Organization 2001 we analized the150 top worldwide pharma firms.
How are the NCEs distributed among these top firms?
Bosonic opportunities
![Page 17: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/17.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Pharma industry NCEs
New Chemical Entity (NCE): new molecules with novel therapeuticalproperties
The total number of NCEs introduced over the period 1975-1994 is154.
In G.Bottazzi, G.Dosi, M.Lippi, F.Pammolli and M.Riccaboni,International Journal of Industrial Organization 2001 we analized the150 top worldwide pharma firms.
How are the NCEs distributed among these top firms?
Bosonic opportunities
![Page 18: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/18.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Pharma industry NCEs
New Chemical Entity (NCE): new molecules with novel therapeuticalproperties
The total number of NCEs introduced over the period 1975-1994 is154.
In G.Bottazzi, G.Dosi, M.Lippi, F.Pammolli and M.Riccaboni,International Journal of Industrial Organization 2001 we analized the150 top worldwide pharma firms.
How are the NCEs distributed among these top firms?
Bosonic opportunities
![Page 19: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/19.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Pharma industry NCEs
New Chemical Entity (NCE): new molecules with novel therapeuticalproperties
The total number of NCEs introduced over the period 1975-1994 is154.
In G.Bottazzi, G.Dosi, M.Lippi, F.Pammolli and M.Riccaboni,International Journal of Industrial Organization 2001 we analized the150 top worldwide pharma firms.
How are the NCEs distributed among these top firms?
Bosonic opportunities
![Page 20: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/20.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The distribution of NCEs
In this case it is N = 150 and M = 154.
Bosonic opportunities
![Page 21: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/21.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The distribution of NCEs
The “mainly bosonic” nature of NCEs was easily established.
But we lack a list of “business opportunities” in the different sectors.
Thus to asses their bosonic nature we have to revert to an indirectproof.
It turns out that the shape of the firm growth rate distribution providessuch proof. . . but we need a little more theory!
Bosonic opportunities
![Page 22: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/22.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The distribution of NCEs
The “mainly bosonic” nature of NCEs was easily established.
But we lack a list of “business opportunities” in the different sectors.
Thus to asses their bosonic nature we have to revert to an indirectproof.
It turns out that the shape of the firm growth rate distribution providessuch proof. . . but we need a little more theory!
Bosonic opportunities
![Page 23: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/23.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The distribution of NCEs
The “mainly bosonic” nature of NCEs was easily established.
But we lack a list of “business opportunities” in the different sectors.
Thus to asses their bosonic nature we have to revert to an indirectproof.
It turns out that the shape of the firm growth rate distribution providessuch proof. . . but we need a little more theory!
Bosonic opportunities
![Page 24: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/24.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
The distribution of NCEs
The “mainly bosonic” nature of NCEs was easily established.
But we lack a list of “business opportunities” in the different sectors.
Thus to asses their bosonic nature we have to revert to an indirectproof.
It turns out that the shape of the firm growth rate distribution providessuch proof. . . but we need a little more theory!
Bosonic opportunities
![Page 25: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/25.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rate
Let S(t) be the size of the firm at time t and s(t) = log S(t). Observedgrowth is the cumulative effect of diverse independent shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t)
Let µε and σ2ε be the mean and variance of the shocks, then
E[g] = Nµε V[g] = Nσ2ε .
If N becomes large and µε, σ2ε ∼ 1/N we have many micro-shocks.
What we expect to observe as the distribution of g?
Bosonic opportunities
![Page 26: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/26.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rate
Let S(t) be the size of the firm at time t and s(t) = log S(t). Observedgrowth is the cumulative effect of diverse independent shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t)
Let µε and σ2ε be the mean and variance of the shocks, then
E[g] = Nµε V[g] = Nσ2ε .
If N becomes large and µε, σ2ε ∼ 1/N we have many micro-shocks.
What we expect to observe as the distribution of g?
Bosonic opportunities
![Page 27: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/27.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rate
Let S(t) be the size of the firm at time t and s(t) = log S(t). Observedgrowth is the cumulative effect of diverse independent shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t)
Let µε and σ2ε be the mean and variance of the shocks, then
E[g] = Nµε V[g] = Nσ2ε .
If N becomes large and µε, σ2ε ∼ 1/N we have many micro-shocks.
What we expect to observe as the distribution of g?
Bosonic opportunities
![Page 28: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/28.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rates
Because of the Central Limit Theorem . . .
Bosonic opportunities
![Page 29: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/29.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rates
Because of the Central Limit Theorem . . .
µg
Normal bell-shaped density
σ2g
Bosonic opportunities
![Page 30: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/30.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Firm growth rates
Because of the Central Limit Theorem . . .
µg
Normal bell-shaped density
σ2g
Bosonic opportunities
![Page 31: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/31.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Random opportunities assignment
The previous model had no competition in it: each firm takes a largebut FIXED number of opportunities.
Assume instead that micro-schocks arise from opportunities that aredistributed among firms, like microscopic particles among energylevels.
Thus N becomes a random variable N. Observed growth as thecumulative effect of diverse shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t) . (1)
Bosonic opportunities
![Page 32: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/32.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Random opportunities assignment
The previous model had no competition in it: each firm takes a largebut FIXED number of opportunities.
Assume instead that micro-schocks arise from opportunities that aredistributed among firms, like microscopic particles among energylevels.
Thus N becomes a random variable N. Observed growth as thecumulative effect of diverse shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t) . (1)
Bosonic opportunities
![Page 33: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/33.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Random opportunities assignment
The previous model had no competition in it: each firm takes a largebut FIXED number of opportunities.
Assume instead that micro-schocks arise from opportunities that aredistributed among firms, like microscopic particles among energylevels.
Thus N becomes a random variable N. Observed growth as thecumulative effect of diverse shocks
g(t) = s(t + 1)− s(t) = ε1(t) + ε2(t) + . . . =
N∑j=1
εj(t) . (1)
Bosonic opportunities
![Page 34: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/34.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many firms and opportunities
With N = E[N] it is again
E[g] = Nµε V[g] = Nσ2ε .
What happens if N → +∞, and µε, σ2ε ∼ 1/N, that is if we have
many micro-shocks? What we expect to observe as the distribution ofg?
It depends on the assignment procedure.
If the opportunities are distributed as classical particles, we are backto the Normal distribution.
If opportunities are distributed as “bosons” then in G.Bottazzi andA.Secchi, Rand Journal of Economics, 37, 2006 we showed that . . .
Bosonic opportunities
![Page 35: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/35.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many firms and opportunities
With N = E[N] it is again
E[g] = Nµε V[g] = Nσ2ε .
What happens if N → +∞, and µε, σ2ε ∼ 1/N, that is if we have
many micro-shocks? What we expect to observe as the distribution ofg?
It depends on the assignment procedure.
If the opportunities are distributed as classical particles, we are backto the Normal distribution.
If opportunities are distributed as “bosons” then in G.Bottazzi andA.Secchi, Rand Journal of Economics, 37, 2006 we showed that . . .
Bosonic opportunities
![Page 36: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/36.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many firms and opportunities
With N = E[N] it is again
E[g] = Nµε V[g] = Nσ2ε .
What happens if N → +∞, and µε, σ2ε ∼ 1/N, that is if we have
many micro-shocks? What we expect to observe as the distribution ofg?
It depends on the assignment procedure.
If the opportunities are distributed as classical particles, we are backto the Normal distribution.
If opportunities are distributed as “bosons” then in G.Bottazzi andA.Secchi, Rand Journal of Economics, 37, 2006 we showed that . . .
Bosonic opportunities
![Page 37: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/37.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Many firms and opportunities
With N = E[N] it is again
E[g] = Nµε V[g] = Nσ2ε .
What happens if N → +∞, and µε, σ2ε ∼ 1/N, that is if we have
many micro-shocks? What we expect to observe as the distribution ofg?
It depends on the assignment procedure.
If the opportunities are distributed as classical particles, we are backto the Normal distribution.
If opportunities are distributed as “bosons” then in G.Bottazzi andA.Secchi, Rand Journal of Economics, 37, 2006 we showed that . . .
Bosonic opportunities
![Page 38: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/38.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Laplace density
. . . when E[N]→ +∞ the distribution converges to a Laplace (doubleexponential)
µg
Laplace density
σ2g
Bosonic opportunities
![Page 39: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/39.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Laplace density
. . . when E[N]→ +∞ the distribution converges to a Laplace (doubleexponential)
µg
Laplace density
σ2g
Bosonic opportunities
![Page 40: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/40.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Simulation Results for N = 100
1e-06 1e-05
0.0001
0.001 0.01 0.1
1
-5 -4 -3 -2 -1 0 1 2 3 4 5
M=0 M=100 M=10000
0.1
1
-1 -0.5 0 0.5 1
from G.Bottazzi and A.Secchi Explaining the Distribution of Firms Growth Rates, Rand Journal of Economics, 37, 2006
Bosonic opportunities
![Page 41: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/41.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Alternative models
µg
Normal bell-shaped density
σ2g
Classical opportunities: normaldistribution
µg
Laplace density
σ2g
Bosonic opportunities: Laplacedistribution
Bosonic opportunities
![Page 42: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/42.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Data - U.S.
COMPUSTAT U.S. publicly traded firms in the ManufacturingIndustry (SIC code ranges between 2000-3999) in the time window1982-2001.
analysis performed in
G.Bottazzi and A.Secchi Review of Industrial Organization, vol. 23, pp. 217-232, 2003
Bosonic opportunities
![Page 43: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/43.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - U.S.
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Aggregate
Two digits sectors
Bosonic opportunities
![Page 44: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/44.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - U.S.
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Aggregate
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Food
Two digits sectors
Bosonic opportunities
![Page 45: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/45.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - U.S.
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Aggregate
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Food
0.01
0.1
1
-6 -4 -2 0 2 4 6
Apparel
Two digits sectorsBosonic opportunities
![Page 46: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/46.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - U.S.
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Aggregate
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Food
0.01
0.1
1
-6 -4 -2 0 2 4 6
Apparel
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Instruments
Two digits sectorsBosonic opportunities
![Page 47: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/47.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - U.S.
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Aggregate
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Food
0.01
0.1
1
-6 -4 -2 0 2 4 6
Apparel
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6
Instruments
Two digits sectorsBosonic opportunities
![Page 48: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/48.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Data - Italy
MICRO.1 Developed by the Italian Statistical Office(ISTAT). Tenthousands firms with 20 or more employees in 97 sectors (3-digitATECO) in the time window 1989-1996. We use 55 sectors withmore than 44 firms.
analysis presented inG.Bottazzi, E.Cefis and G.Dosi Industrial and Corporate Change vol. 11, 2002;G.Bottazzi, E.Cefis, G.Dosi and A.Secchi Small Business Economics 29, pp. 137-159, 2007
Bosonic opportunities
![Page 49: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/49.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - ITA
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Aggregate
Three digits sectors
Bosonic opportunities
![Page 50: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/50.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - ITA
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Aggregate
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Pharmaceuticals
Three digits sectors
Bosonic opportunities
![Page 51: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/51.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - ITA
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Aggregate
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Pharmaceuticals
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Cutlery, tools and general hardware
Three digits sectorsBosonic opportunities
![Page 52: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/52.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - ITA
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Aggregate
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Pharmaceuticals
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Cutlery, tools and general hardware
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Footwear
Three digits sectorsBosonic opportunities
![Page 53: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/53.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - ITA
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Aggregate
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Pharmaceuticals
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Cutlery, tools and general hardware
0.01
0.1
1
-1.5 -1 -0.5 0 0.5 1 1.5
Footwear
Three digits sectorsBosonic opportunities
![Page 54: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/54.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Data - worldwide pharma industry
PHID Developed by the CERM research institute. Sales figures fortop pharmaceutical firms in United States, United Kingdom, France,Germany, Spain, Italy and Canada for the years 1987-97.
analysis presented inG.Bottazzi and A.Secchi Review of Industrial Organization, 26, 2005
Bosonic opportunities
![Page 55: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/55.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Empirical Growth Rates Densities - Pharma
0.001
0.01
0.1
1
-6 -4 -2 0 2 4 6 8∧g
Bosonic opportunities
![Page 56: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/56.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
A more general result
Theorem
Let g(λ, µ, σ) =∑N
j=1 εj(t) with ε i.i.d distrbuted according to acommon distribution with mean µ and variance σ2, and N distributedaccording to a distribution h of mean λ. Assume that there is an n′
such that for any n > n′ it is
limλ→+∞
λ supn>n′{h(n)− G(n)} = 0
where G(n) is the geometric distribution with mean λ, then
limλ→+∞
g(λ, µ/λ, σ/λ) ∼ Laplace
Bosonic opportunities
![Page 57: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/57.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
In other words . . .
It in not necessary to have a Bose-Einstain statistics, but any way todistribute opportunities which leads, in the limit of a large number ofopportunities, to a geometric marginal distribution will work!
The meansing of the Geometric distribution is
Prob{N = n + 1} = cosntant× Prob{N = n}
that is the probability to get one more opportunity is proportional tothe number of opportunities already got.
This is the Gibrat’s Law of Proportionate Effect or “competitionamong objects whose market success...[is] cumulative orself-reinforcing” (B.W. Arthur)
Bosonic opportunities
![Page 58: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/58.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
In other words . . .
It in not necessary to have a Bose-Einstain statistics, but any way todistribute opportunities which leads, in the limit of a large number ofopportunities, to a geometric marginal distribution will work!
The meansing of the Geometric distribution is
Prob{N = n + 1} = cosntant× Prob{N = n}
that is the probability to get one more opportunity is proportional tothe number of opportunities already got.
This is the Gibrat’s Law of Proportionate Effect or “competitionamong objects whose market success...[is] cumulative orself-reinforcing” (B.W. Arthur)
Bosonic opportunities
![Page 59: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/59.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
In other words . . .
It in not necessary to have a Bose-Einstain statistics, but any way todistribute opportunities which leads, in the limit of a large number ofopportunities, to a geometric marginal distribution will work!
The meansing of the Geometric distribution is
Prob{N = n + 1} = cosntant× Prob{N = n}
that is the probability to get one more opportunity is proportional tothe number of opportunities already got.
This is the Gibrat’s Law of Proportionate Effect or “competitionamong objects whose market success...[is] cumulative orself-reinforcing” (B.W. Arthur)
Bosonic opportunities
![Page 60: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/60.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Conclusions
The business opportunities follow the Bose-Einstein statistics.
The business opportunities are destributed across firms in clusters,with a relatively high probability to have a large number of themassigned to the same firm.
The Gibrat’s Law applies to business opportunities.
Bosonic opportunities
![Page 61: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/61.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Conclusions
The business opportunities follow the Bose-Einstein statistics.
The business opportunities are destributed across firms in clusters,with a relatively high probability to have a large number of themassigned to the same firm.
The Gibrat’s Law applies to business opportunities.
Bosonic opportunities
![Page 62: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/62.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Conclusions
The business opportunities follow the Bose-Einstein statistics.
The business opportunities are destributed across firms in clusters,with a relatively high probability to have a large number of themassigned to the same firm.
The Gibrat’s Law applies to business opportunities.
Bosonic opportunities
![Page 63: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/63.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Self-reinforcing in the number of bins
As suggested in Y. Ijiri and H. Simon Proc. Nat. Acad. Sci. USA,1975 assign bins proportionally to bin’s “size”. Initially attach to eachbin the same size S = 1.
S=1 S=1
S=2 S=1 S=1 S=2
1/2 1/2
Bosonic opportunities
![Page 64: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/64.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Self-reinforcing in the number of bins
Increase the size with the number of balls.
S=1 S=1
S=2 S=1 S=1 S=2
S=2 S=2S=2 S=2 S=3 S=1 S=1 S=3
1/2 1/2
1/3
1/3 1/3 2/3
1/6 1/6 1/3
2/3
Bosonic opportunities
![Page 65: Scuola Superiore Sant’Annagiulio/doc/talk_ISS_2014.pdf · Scuola Superiore Sant’Anna ISS Conference- July 2014 - Jena Bosonic opportunities. Introduction ... name “boson”](https://reader035.fdocuments.in/reader035/viewer/2022070817/5f122951c07288412b65b758/html5/thumbnails/65.jpg)
IntroductionThe distribution of opportunity
Empirical evidence
Diversification patterns
from G.Bottazzi and A.Secchi Industrial and Corporate Change, 2006
Bosonic opportunities