TECHNOLOGY INNOVATION IN MANUFACTURING: Real Options

1
TECHNOLOGY INNOVATION IN MANUFACTURING
Real Options GAMES Alcino
Azevedo and Dean Paxson
Managerial Presentation
Real Options Conference, Rio de Janeiro July,
2008
2
Relevant Considerations for Technology Innovations

• Competition and First Mover Advantage
  Efficiency (FMAE)
• Arrival of Better Technologies (l)
• Complimentarity (g)

3
The Textile Industry

• Tech 2 Tech 1
• Technological
  complementarity

Textile Fibers (cotton, polyester, wool, etc)
Make Up
Finish Treatments Dyeing, Printing, etc
Weaving (fabrics)
Spinning (yarns)
Domestic Textiles
4
The Investment Opportunity

• New Textile Product (US Market)
• Two Firms Lameirinho and Coelima _at_same size,
  mostly high quality exports
• First-mover market/Efficiency advantage.
• Investment
• New Weaving Machines
• High Quality Yarns (raw material to produce
  textile fabrics)
• Investment Decision
• Invest 13m Euros for Initial Annualized Net
  Revenues of 2.65m Euros, if First
• New weaving machines only for one textile firm,
  immediately
• Follower obtains 40 market share
• FMA of monopoly until follower enters, then 60
  market share
• FME because learn production efficiency while
  operating.

5
Model 1

• Underlying Variables
• Market Uncertainty
• X(t), Net Revenues
• Technical Uncertainty
• E(t), Efficiency .
• (Both following independent and possibly
  correlated gBm processes)
• Change in the variables
• X(t)E(t)(r- u X - u E) F
• F, annualized net revenues weighted with the
  efficiency variable

6
Model 1 (cont.)

• Investment Game
• Two Firms Lameirinho (Leader), Coelima
  (Follower)
• OneTechnology
• Tech 1 (weaving machines - currently available),
• Firms value functions (payoffs) and investment
  trigger values?

7
GAME STRUCTURE
Firm i Adopt
Wait  
Firm j
Firm j
Adopt Wait Adopt
Wait   Payoff Firm i FL(f)
FL(f) FL(f)
Fi(f) Payoff Firm j
FF(f) FF(f)
FF(f) Fi(f)
i(L,F) L Leader F Follower Figure
3.3 - Extensive-Form representation of a
Continuous Time Real Option Game (CTROG) with 2
players, 2 strategies available and 2 symmetric
payoffs.
8
Model 2

• Underlying Variables
• Market Uncertainty
• X(t), Net Revenues
• Technical Uncertainty
• E(t), Efficiency .
• (Both following independent and possibly
  correlated gBm processes)
• Change in the variables
• X(t)E(t)(r- u X - u E) F
• F, annualized net revenues weighted with the
  efficiency variable
• Technological Uncertainty
• l, Probability of New Technology arrival (Poisson
  distribution).

9
Model 2 (cont.)

• Investment Game
• Two Firms Lameirinho (Leader), Coelima
  (Follower)
• Two Technologies
• tech 1 (weaving machines - currently available),
• tech 2 (weaving machines - may arrive in the
  future with probability ?)
• Assumptions
• The Leader is commited to the adoption of tech 1
  and, by assumption, tech 1 is adopted before tech
  2 arrives
• The Follower is commited to the adoption of tech
  2, and adopts it as soon as it arrives and its
  investment rigger value is reached.
• Firms value functions (payoffs) and investment
  trigger values?

10
Model 3

• Underlying Variables
• Market Uncertainty
• X(t), Net Revenues (gBm)
• I1(t), I2(t), Cost of tech 1 and tech 2 (gBm).
• Technological Complementarity
• Tech 1 (weaving macines)
• Tech 2 (spinning machines)
• Tech 1,2 (both weaving plus spinning machines)
• ?1 ?2lt ?
• Where ?1, ?2 and ? are , respectively
• Cost savings when tech 1 is adopted (given as a
  proportion of the revenues)
• Cost savings when tech 2 is adopted
• Cost savings when tech 1 and tech 2 are adopted.

11
Model 3 (cont.)

• Investment Game
• Two Firms Lameirinho (Leader), Coelima
  (Follower)
• Two new Technologies tech 1 (new weaving
  machines), tech 2 (new spinning machines)
• At time t, firms have three options
• a) adopt tech 1 alone
• b) adopt tech 2 alone, or
• c) adopt both technologies (tech 1 and tech 2) at
  the same time.
• Firms value functions (payoffs) and investment
  trigger values?

12
Efficiency Dataset

• Figure 1 - The Efficiency of the Previous Version
  of the Sulzer-ruti Weaving Machines
• during the First 15 Months of Activity.

13
Impact of Efficiency Uncertainty

• Figure 2 - The Impact of an Unexpected Loss in
  the Efficiency of the New Weaving Machines on the
  Daily Production of the Weaving Mill.

Unexpected lost in production per day (840 m2)
Unexpected lost in Efficiency (3)
14
Inputs

• Table 1

15
Inputs (cont.)
Table 2
16
Results Model 1
17
Sensitivity Analysis
18
Results Model 2
19
Results Model3
20
Conclusions

• This paper uses three real option models to
  advise firms, operating in duopoly markets, on
  timing investments in new technologies, where
• There is a first-mover advantage
• Revenues from the adoption are uncertain
• The technical performance of the new
  technology(ies) after adoption is uncertain
• Investment costs are uncertain (decreasing over
  time)
• The new technologies are complementary (in case
  of adoption of more than one technology).
• In all cases, the Leader should invest now, but
  the Follower should wait (not long in Model 2, or
  in Model 3 where there are large
  complementarities between weaving spinning cost
  savings)

21
Conclusion (cont.)

• Our results are significantly affected by the
  following factors
• The first-mover advantage
• The drift and volatility of revenues/efficiency
  after adoption
• Technological progress
• Proportion of the revenues saved through the
  adoption of the new technology(ies) and
  technological complementarity
• Closed-form solutions leader/followers value
  functions and followers investment trigger
  values.
• Numerical solution leaders investment trigger
  values.
• Empirical engineering studies evolving machine
  efficiency during operations.

22
Similar ROC-Rio Investment Opportunities with
Competition Efficiency, New Technology
Arrivals, and Complementarities

• Petrochemical
• Nuclear
• Power
• Networks
• Airplanes Air Routes
• Bio Plants
• Current MBS Studies On
• Chilean Animal Food Plants
• Uganda Ethanol Plants
• ME Fertilizer Production
• Portugal Textile Equipment Replacement