Operations systems management. Capacity and Flexibility

1. Suppose that Sony believes that acceptance rate is 50%, and Philips
knows Sony’s belief with certainty. Philips still believes that the
acceptance rate is 5%, but Sony assumes that Philips believes that the
acceptance rate is 50%. When will each of the players invest (if at
all)? Explain your answer.

2. A Philips employee approached Sony’s vice president for research
and development and offered to provide Sony with Philips’s estimates
of the acceptance rate. Assuming that Sony is not likely to change
their own estimates of the acceptance rate, what is the value of this
information?

The Phillips case

Do you think Phillips should invest in 1983 or in 1984?

Suppose Phillips decides not to invest in 1983.

Suppose also that Phillips observes that popular buyers accept quickly.


What should Philips do in 1984?

Second period analysis: Philips

In this case Phillips should invest.

Notice that 2310 > 938 and 4579 > 1603

















Capacity and flexibility

Second period analysis: Philips

The situation is simpler. We now need to consider only the following
information

Sony

Second period analysis: Sony

What should Sony do?

Assuming that Sony does not know the state of the market, and assumes 5%
quick acceptance and 95% slow acceptance.

Sony compares: 0.05*1207 + 0.95*(-95) = -30 with 0.05*563 + 0.95*0 = 28

It is clear that Sony should not build in 1984

First period analysis: Phillips

Knowing that Sony will not invest in 1984, what should Phillips do?

Assuming Phillips invests in the first year then the expected profit is:


0.05*4585 + 0.95*533 = 736

Assuming Phillips invests in the second year then the expected profit
is:

0.05*4579 + 0.95*546 = 748

Clearly Phillips Should invest on the second year

A second scenario: Fast acceptance

Let us assume a different scenario in which Sony and Phillips believe
that the probability of fast acceptance is 50%.

What will change in our analysis?

On the second period if Phillips observes quickly acceptance they will
invest. (as before)

Sony compares: 0.5*1207 + 0.5*(-95) = 556 with 0.5*563 + 0.5*0 = 281.5

It is clear that Sony should build in 1984. (unlike before)

Knowing that Sony will invest in 1984, what should Phillips do?

Assuming Phillips invests in the first year then the expected profit is:


0.5*4585 + 0.5*533 = 2559

Assuming Phillips invests in the second year then the expected profit
is:

0.5*2310 + 0.5*422 = 1366

Clearly, Phillips Should invest on the first year

A third scenario: Information sharing

The value of flexibility was reduced.

Notice, in the above two examples we assumed that Phillips is able to
observe the real demand but Sony is unable to do the same.

What happens if Sony gets the same information as Phillips.

In analyzing this situation we assume that the probability of fast
acceptance is 0.05

Assuming that Phillips does not build in 1983, it is clear that both
Phillips and Sony will build in 1984.

What should Phillips do?

Assuming Phillips invests in the first year then the expected profit is:


0.05*4585 + 0.95*533 = 736

Assuming Phillips invests in the second year then the expected profit
is:

0.05*2310 + 0.95*422 = 516

It is clear that Phillips should invest in the first period.

A fourth scenario: Some information sharing

Do we know whether Sony will get the information?

Suppose we assume that the chance that Sony will get the information is
10% what should Phillips do?

As before, If Phillips does not invest in 1983 and acceptance is fast,
then Phillips invests in 1984.

Phillips does not know whether Sony got the information. So, Phillips
expected profit is:
0.1*(2310*0.05+0.95*422)+0.9*(0.05*4578+0.95*546)=724

If Phillips invests in 1983 the profit is 736

Clearly Phillips should invest in the first period.

End of the story

Phillips did not invest in the pressing facility in 1983. It assumed
that the rate of acceptance will be low.

Sony built a small facility with the capacity of 3.5m (1984)

At the end of 1985 Sony had capacity of 12m

In 1987 it had the capacity of 26.4m

In 1987 Phillips open its first facility with capacity of 12m-15m

Lessons to learn

What were the uncertainties that Phillips faced?

Acceptance rates

Sony knowledge about the acceptance rates

What was the approach Phillips took?

A single probability for the speed of acceptance (5%)

A single probability for Sony’s knowledge of the acceptance rate (0%)

Future profits, single numbers

How accurate these number can be?

Past data and experience tell us that these numbers can’t be very
accurate.

How robust is Phillips strategy of not investing?

Our calculations show that it is not robust at all.

What are the results?

A non-robust forecast. We have seen that small changes in the
assumptions will result in a major shift in the policy.

What should Phillips have done?

Create three to four scenarios:

The present scenario

A scenario that assumes that Sony gets the information with a small
probability, and acceptance rate is small

A scenario in which Sony gets the information and acceptance rate is
high

What will a robust policy in this case be?

Invest in the first period.

 

Popular Buyers accept quickly

P. builds in 1984

P. does not build in 1984

Sony

Builds in 1984

2310 1207

 

938 3295

Does not build in 1984

4579 563

1603 1423

 

quickly Slowly

P. builds in 1984

Sony

Builds in 1984

2310 1207

422 -95

Does not build in 1984

4579 563

546 0

 

Popular Buyers accept quickly

P. builds in 1984

P. does not build in 1984

Sony

Builds in 1984

 2310 1207

 938 3295

Does not build in 1984

4579 563

1603 1423

Philips Compact Disc Introduction.

In 1979, executives in the Consumers Electronics Group at Philips NV of
the Netherlands decided to investigate the market opportunities for the
newly developed Compact Disc (CD), an optical disc product for
reproducing pre-recorded music for home entertainment. Although
Philip’s managers recognized that the technical superiority of CD’s
in sound reproduction might appeal to consumers they also realized that
its commercial success would depend on resolving a number of competitive
and marketing issues.

Experience in both the video and audio segments heightened the
management team’s concern for standardization. Several firms in Europe
and Japan had announced development of digitised audio products. If
incompatible digital formats were introduced simultaneously, consumer
acceptance might be permanently damaged. A second and related question
concerned forecasts of profitability on the hard ware (i.e., CD players)
and software (i.e., discs) sides of the business. The team wanted to
develop pricing policies that reflected the opportunity for
differentiation as well as the advantages of a broad installed base. The
timing of changes in prices would also be crucial. Influential,
committed audiophiles would be willing to pay more for the system than
casual users.

Finally the team was faced with difficult decisions regarding the timing
of capital investments. If Philips were to commit relatively early to
significant disc pressing capacity, for example, it might attract other
consumer electronics firms to its standards. On the other hand too much
pressing capacity might intimidate other firms and undermine
standardization.

Technology development.

Early research on digital reproduction methods using lasers was
conducted at the Massachusetts Institute of Technology during the 1950s.
Engineers in the Philips Research Laboratories began to explore
potential applications in the late 1960’s. By the early 1970s the
Consumer Electronics Group demonstrated its first video prototype based
on optical scanning.

Laser Vision products used scanning over analog images engraved on
discs. Despite their superior reproduction capability, the discs sold in
limited quantities because of their playback-only capability.

Consumers were unwilling to accept the product at a high price premium
over the newly introduced videotape recorder.

Philips therefore began developing audio applications for optical
scanning technology in an effort to partially recover eight years and
several hundred million dollars of development expenditures. By 1979 it
had developed a prototype audio CD player and a 4.5-inch disc that held
60 minutes of music. Although the optical engraving process for CDs was
similar to the process for CDs was similar to the process for Laser
Vision, the engraved information was digitised for the CD (as opposed to
analog for Laser Vision). Digital reproduction required conversion of
sampled waves to binary codes, which were then recorded by optical
scanners as engraved pits on a specially treated disc. For playback, a
high-speed laser read the pits off the surfaces of the discs and
electronically reconverted the binary code to the original medium. If
correctly performed, the process left little room for error in
reproduction.

The CD would differ from conventional audio products in its accuracy and
durability.

Battles to establish a standard in the video segment of the consumer
electronics industry reached beyond the Laser Vision/VCR issue. Within
the VCR sub segment vigorous competition had emerged between adherents
to Beta and to VHS technology. Sony had been the innovator with its
introduction of the Betamax videotape format in the mid 70s. Both RCA
and Philips challenged sales of high-end videotape recorders with
disc-based technologies. The Japan Victor’s corporation’s (JVC)
subsequent introduction of an incompatible format under the VHS label
compounded a competitive situation that had proven costly for all firms.
It appeared that the dominant standard would not emerge for several
years.

An efficiently scaled audio CD manufacturing line with the capacity to
produce two million discs per year would cost $25 million and take a
year and a half to build. The principal activities in the production
process were mastering, electroforming, molding, coating, sealing,
labelling, centering, and final testing, followed by packaging.
Efficient production presumed rough capacity balance for these
activities. With the exception of the mastering equipment, which
incorporated newly designed semiconductors and laser technology
virtually all the technology involved had been adapted from familiar
consumer electronics applications. Nevertheless, the tolerances and
cleanliness required of CD equipment made it prohibitively expensive to
start up a production line in less than a year and a half and ensured
that it had little salvage value except as scrap.

Once a facility was established, managers would be able to identify the
sources of contamination and curb them by adjusting procedures,
modifying equipment, and automating linkages that proved to be
bottlenecks. The historical data on the improvement of audio CD yields
over time was limited. Video discs which encoded rather different
information using similar manufacturing activities had not exhibited
consistent progress even though they had been produced far longer: in
1979 the yields on some production lines were still stuck at about 50%.
The smaller size of audio CDs would permit more consistency in heating
however relaxing some of the constraints that kept videodiscs yields
low. It appeared that by the time 5 million audio CDs had been produced
on a potentially efficient line, its average yield would reach 70% and
with experience levels of 10 Million discs, as much as 90%. Based on
these estimates, the ongoing cost (excluding all capital charges) of
producing a useable disc would be $3.00 at start up, $2.34 after
producing 5 million discs, $1.77 after the production of 10 million
discs and very close to the lower limit, based on competitively priced
inputs and 100% yields, of $0.69 after 50-60 million had been produced.

As time passed, start up time would be reduced to a year and the capital
equipment used in new pressing plants would also improve and become less
expensive. According to a moderate estimate, the one time cost of
installing efficiently scaled new capacity would drop from its initial
level of $12.50 per unit (of disc pressing capacity per year) to $8.35
after one year, $5,58 after two years, $3.73 after three years, $2.49
after four years and $1.67 subsequently. The time required to build a
plant was also expected to decrease gradually so after two years new
plants would require only a year for construction. While it would not be
economical to refit old production capacity to take advantage of these
improvements, independent equipment suppliers would allow established
firms to capitalize on them when existing plants were made incrementally
larger.

Experience in manufacturing CDs might also be useful later if the
technology was successfully extended to solve data storage problems for
businesses. Commercial applications wren possible because the pits
engraved on CDs could just as easily reflect binary representations of
business data as music. A phone book for the entire USA for example,
could be stored on the equivalent of just 7 CD audio discs. CD ROMs (CD
-read only memory disks) was the name attached to digital technology in
business applications. Commercial acceptance for CD-Rom had been limited
both because computer companies had blocked access to distribution
channels and because businesses were reluctant to ship off
organizational records to CD pressing plants.

Demand.

There were several reasons why compact discs were expected to be of
particular appeal to buyers of prrecorded classical and jazz music.
First the greatest increment in the equality of sound would occur in
these recordings because of their large dynamic range.

Second, buyers in these segments had repeatedly demonstrated a
willingness to upgrade components to improve the quality of their
hardware systems. Finally buyer acceptance for higher quality recordings
had been tested in part with digitally mastered LPs.

Digitally mastered LPs had first been offered by classical and jazz
labels in the early 1970s at a significantly higher price than
conventional LPs. The masters for such recordings were prepared using
digital techniques, although the final LP was presses in vinyl, which
could only be manufactured using standard mechanical methods. Consumer
acceptance was so great that by 1982 they accounted for a majority of
unit sales of classical music at a 30% price premium. Although they did
not offer the same improvement in quality over conventional LPs, as
would the Compact disc, they did provide a benchmark for assessing the
value attached to extra quality by buyers in these segments.

The classical and jazz segments accounted for approximately 7.6% of the
sales of prerecorded music. Within this segment buyers of prerecorded
tapes were unlikely to switch to CDs because they apparently attached a
lower value to quality relative to other characteriscs such as
portability convenience or recording capability. The relevant benchmark
for forecasting unit sales therefore seemed to be the total market for
classical and jazz LPs. Penetration by CDs into the classical and jazz
segment would take several years. If acceptance of CD hardware in this
segment followed a classical S curve penetration rates would be 0.1% in
the year of introduction, and 1%, 5%15% 29% 50% 75% and 100% in
subsequent years.

Buyers in the popular, rock, country, soul and other segments (hereafter
called the popular segment) bought 92.3% of pre-recorded music (by
volume) in the USA. The rate of acceptance of CD in the popular segment
was more difficult to forecast for several reasons. First the
improvement in sound quality would not be as large because dynamic range
on popular music was not as wide. Second, the CD technology required a
change in recording style: artists would have to curtail the use of
engineered splices and dubbing because these would be audible in digital
reproduction. This in turn would make their music sound inherently
different. Third although portable Cd players would be available if
buyer acceptance were high, it was unclear that consumers would view the
new format as comparable to cassette tapes in portability. Fourth buyers
in the popular segment might not be willing to make their collections of
LPs and cassettes obsolete. Finally the playback-only capability of the
CD might also limit its appeal.

Market researchers had studied the likely extent of switching using
several different methodologies and had come to rather different
conclusions. Philips and Polygram executives were convinced that
additional testing would not yield more satisfying results until after
the product was introduced to the mass market and consumers were more
familiar with the product. They did know the in the USA, legal
restrictions would force pressing companies to charge the same per disc
price for their classical/jazz and popular labels and therefore left no
opportunity to tailor prices to the valuations of buyers in different
segments.

In the best-case scenario, popular segment buyers would convert from LPs
to Cds at the same rate as classical and jazz buyers. Polygram assigned
a higher probability to this outcome than their counterparts at Philips
headquarters, but even the record company’s managers conceded that the
chances of this happening were 15% at most. Their second more likely
scenario put the rate of conversion in the popular segment at 1/3 the
rate in the classical and jazz segments.

The value of investment in CD pressing plants was also contingent on the
timing of the next generation in audio formats. Digital cassette tapes
would offer the same durability and quality in reproduction as a CD,
together with recording capability and greater portability. Sony already
had a prototype in development (although the costs of production were
very high). The digital cassette introduction was forecasted to be 15 or
20 years away partly because of resistance among performance artists.

Capacity Uncertainty and Flexibility
Agenda
Capacity: The main issues
Case presentation
The role of flexibility in making capacity decisions
The role of commitments in making capacity decisions
Analysis of the case
Capital Investments
Intel:
Approx 3 to 7 billion a year
CISCO:
Few hundred million a year
Solectron
36 million in 4th Quarter of 2005
General Motors
2 Billion in 2005
Capacity: Definitions
Capacity is the volume of output per period of time that a business or facility can produce.
Maximum or design capacity defines the highest rate of output that a process or activity can theoretically achieve.
Effective or planned capacity identifies the output rate that managers expect to get for a given activity or process.
Capacity: Definitions
demonstrated capacity deals with actual rather than planned production, and measures the actual level of output for a process or activity over time.
Capacity utilization is the percentage of the facility’s capacity that is used by actual production.

Capacity: Short Term

Short-term capacity is managed through the day-to-day decisions organizations make in response to short-term fluctuations in demand, and usually entail human resource adjustments or investment in quality improvement or process optimization.
Capacity: Intermediate-term
Intermediate-term capacity decisions are made on a month-to-month basis in response to, for example, seasonal fluctuations in demand, and, as for short-term capacity decision, usually entail adjustments to human resources – hiring, firing, overtime, and subcontracting

Capacity: Long Term
Long-term capacity decisions, rely on projections about highly uncertain future demand and the strategic moves of competitors, and involve sizeable investments in resources – such as process and information technology and physical facilities – that take some time to put in place.

Capacity: The main Issues
What is the forecasted demand for the company’s products or services in the short-, intermediate- and long-term? How will demand grow or shrink over time?
How much capacity should the company have to cover expected demand? What service level does the company want to offer? How much reserve capacity does the company wish to have to buffer against unexpected fluctuations in demand?
Capacity: The main Issues
In what increments and when, or at what intervals, should the company add capacity?
What type of capacity should the company add? Human resources? Process and information technology? Facilities? Can the company extract more output from its existing resources, thereby increasing its capacity? How will it trade off short-, intermediate-, and long-term investments in capacity?
Where in the value chain, internal and/or external to the company, should capacity be added?

Capacity: Considerations
Explore multiple capacity expansion (or contraction) scenarios
Quantify the sensitivity of the answer to varying market assumptions
Use discounted cash flow techniques to quantify the costs of various capacity expansion timing policies
Evaluate competitive responses to capacity changes
Capacity: Considerations
Lead – Try Not to Run Short.
Lag – Maximize Capacity Utilization. In the lag strategy, average demand is never fulfilled and the company holds less than necessary capacity.
Stay Even -- Build to Forecast. In the stay-even policy, average demand is met, but the company has excess capacity half of the time, and too little capacity the other half of the time.
Lead and Lag policies

Evaluating Investments
Example 1:
$10 million investment:
Annual returns for next 5 years:
3 Million/ year
Discount rate: 15%

Should the firm make the investment?
Irreversible Investment & Demand Uncertainty
XYZ Inc is planning to build a plant to make Gizmos
Cost of investment: $1.6 million
Current selling price of a Gizmo: $60.
Annual Sales volumes: 10000 units
Profit: $200,000
Next year
50% chance price will increase to $70 and profits go up to $300,000 per year – remain at that level forever
50% chance price will drop to $50 and profits go down to $100,000, remain at that level forever
Should the firm Invest in the plant?
Assume salvage value is zero
NPV if it decides to invest?
Expenditure: 1.6 million
You will get in perpetuity
$200,000 this year
0.5*100,000 + 0.5*300,000 = $200,000 then on
Discounted earnings:


Net :
-1,600,000 + 2,200,000 = $600,000

Should they go ahead?
What if it decides to wait for one year?
When is it reasonable to assume you can wait?




If they wait

Assuming that next period the price is $300



If the price is $100 don’t invest
Cost Uncertainties
Research and development projects
Pharmaceuticals
Aerospace

Production Cost
Cost reduction in manufacturing
Raw material costs
Labor costs..

Construction Projects
Real estate development projects
Power plants

Capacity and the role of Flexibility

Clearly, it pays to wait.

What is the value of flexibility?
773-600=173

Research and Development Example
Initial investment in product development is $100,000

Investment in the second phase depends on the outcome of first phase:
If the first phase is successful you can launch and earn discounted revenue of $220,000
If the first phase runs into technical problems you expect to spend $300,000 more, and then earn discounted revenue of $220,000
Estimated probability of success 50%

Should you invest?
Implications
Sequential decision making can lead to different decisions
Example if there is uncertainty in demand building two smaller plants in sequence may be better than committing to a large plant.
Projects value may increase if you have the flexibility to make a decision after some uncertainty gets resolved.

Nature of competitive environments
A small firm in a market with many players
Price Takers:

Large firms in concentrated markets
Price setters

Large but asymmetric power in concentrated markets
Capacity / Capital Investments
Two firms in a concentrated market:

Current Situation:
Costs $10/ units
Profits: $10 Million/ year

New Technology is available:
Investment cost to adopt: $15 Million
Costs : $7.50/ unit
Profits depends on other actions

Suppose you know that the NPV of the project is -$5 Million if the other firm also adopts,
Should you invest?
NPV of project will depend on other’s actions
Evaluating Investment Decision
Need to define the base case or the alternative to consider has to be carefully defined

In the previous case you have to compare profits if you don’t adopt against profits of adopting, given that your rival will adopt & not based on current situation

Capacity ,Commitment & Leadership
The role of commitments
Two companies that choose capacity decision simultaneously.The pay offs are the following:
Capacity ,Commitment & Leadership
What is a Nash Equilibrium?
(x*,y*) is a Nash equilibrium if:




Nash Equilibrium
Can we predict what will the two companies do?
There is a single Nash Equilibrium
Firm 2’s preferred solution?
What would firm 2 want firm 1 to choose?
Capacity and the role of Commitment
Can firm 2 improve its position?
Suppose firm 2 takes a leadership position and announces its capacity ahead of time.
Firm 2 can announce credibly (by building capacity) that it will be aggressive.
In this case, the best firm 1 can do is to announce that it is expanding at lower level.
Credible Commitment
Hernan Cortez, conqueror of the Aztec empire in Mexico

Upon landing in Mexico destroyed all but one of his ships
Lost the flexibility to retreat and sail back!
Capacity and the role of Commitment
Notice that by taking this action firm 2 lost flexibility to learn the market conditions and to adopt.
Actually, by losing flexibility (making a commitment) the firm may become stronger.
For a commitment to be successful it must be:
Visible
Understandable
Credible (irreversible)
Examples
In 2000 Airbus announces high capacity A380 super-jumbo jet and commits to building it by devoting resources and booking orders from Singapore Airlines, FedEx, Quantas

Boeing announces that it is canceling its 747X project
Examples
Band-wagons of capacity expansions in:
Semi-conductor industry
Chemical industry
Capacity & Investment Decisions
Role of uncertainty in information
Flexibility of waiting for uncertainty to be resolved is valuable

Need to select base case correctly
Not status quo but:
What could happen (others may adopt etc).

In concentrated industries commitment can influence others strategies
Phillips vs. Sony
The Phillips case
Do you think Phillips should invest in 1983 or in 1984?
Suppose Phillips decides not to invest in 1983.
Suppose also that Phillips observes that popular buyers accept quickly. What should Philips do in 1984?
Second period analysis: Philips
In this case Phillips should invest. Notice that 2310 > 938 and 4579 > 1603
Second period analysis: Philips
The situation is simpler. We now need to consider only the following information:
Second period analysis: Sony
What should Sony do?
Assuming that Sony does not know the state of the market, and assumes 5% quick acceptance and 95% slow acceptance.
Sony compares: 0.05*1207 + 0.95*(-95) = -30 with 0.05*563 + 0.95*0 = 28
It is clear that Sony should not build in 1984.
First period analysis: Phillips
Knowing that Sony will not invest in 1984, what should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.05*4585 + 0.95*533 = 736
Assuming Phillips invests in the second year then the expected profit is:
0.05*4579 + 0.95*546 = 748
Clearly Phillips Should invest on the second year
A second scenario: Fast acceptance
Let us assume a different scenario in which Sony and Phillips believe that the probability of fast acceptance is 50%.
What will change in our analysis?
On the second period if Phillips observes quickly acceptance they will invest. (as before)
Sony compares: 0.5*1207 + 0.5*(-95) = 556 with 0.5*563 + 0.5*0 = 281.5
It is clear that Sony should build in 1984. (unlike before)

A second scenario: Fast acceptance
Knowing that Sony will invest in 1984, what should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.5*4585 + 0.5*533 = 2559
Assuming Phillips invests in the second year then the expected profit is:
0.5*2310 + 0.5*422 = 1366
Clearly, Phillips Should invest on the first year

A third scenario: Information sharing
The value of flexibility was reduced.
Notice, in the above two examples we assumed that Phillips is able to observe the real demand but Sony is unable to do the same.
What happens if Sony gets the same information as Phillips.
A third scenario: Information sharing
In analyzing this situation we assume that the probability of fast acceptance is 0.05
Assuming that Phillips does not build in 1983, it is clear that both Phillips and Sony will build in 1984.

A third scenario: Information sharing








A third scenario: Information sharing
What should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.05*4585 + 0.95*533 = 736
Assuming Phillips invests in the second year then the expected profit is:
0.05*2310 + 0.95*422 = 516
It is clear that Phillips should invest in the first period.

A fourth scenario: Some information sharing
Do we know whether Sony will get the information?
Suppose we assume that the chance that Sony will get the information is 10% what should Phillips do?
As before, If Phillips does not invest in 1983 and acceptance is fast, then Phillips invests in 1984.
Phillips does not know whether Sony got the information. So, Phillips expected profit is: 0.1*(2310*0.05+0.95*422)+0.9*(0.05*4578+0.95*546)=724
A fourth scenario: Some information sharing
If Phillips invests in 1983 the profit is 736
Clearly Phillips should invest in the first period.
End of the story
Phillips did not invest in the pressing facility in 1983. It assumed that the rate of acceptance will be low.
Sony built a small facility with the capacity of 3.5m (1984)
At the end of 1985 Sony had capacity of 12m
In 1987 it had the capacity of 26.4m
In 1987 Phillips open its first facility with capacity of 12m-15m
Lessons to learn
What were the uncertainties that Phillips faced?
Acceptance rates
Sony knowledge about the acceptance rates
What was the approach Phillips took?
A single probability for the speed of acceptance (5%)
A single probability for Sony’s knowledge of the acceptance rate (0%)
Future profits, single numbers
Lessons to learn
How accurate these number can be?
Past data and experience tell us that these numbers can’t be very accurate.
How robust is Phillips strategy of not investing?
Our calculations show that it is not robust at all.
Lessons to learn
What are the results?
A non-robust forecast. We have seen that small changes in the assumptions will result in a major shift in the policy.
What should Phillips have done?
Create three to four scenarios:
The present scenario
A scenario that assumes that Sony gets the information with a small probability, and acceptance rate is small
A scenario in which Sony gets the information and acceptance rate is high
Lessons to learn
What will a robust policy in this case be?
Invest in the first period.
Diseño predeterminado



Capacity Uncertainty and Flexibility
Agenda
Capacity: The main issues
Case presentation
The role of flexibility in making capacity decisions
The role of commitments in making capacity decisions
Analysis of the case
Capital Investments
Intel:
Approx 3 to 7 billion a year
CISCO:
Few hundred million a year
Solectron
36 million in 4th Quarter of 2005
General Motors
2 Billion in 2005
Capacity: Definitions
Capacity is the volume of output per period of time that a business or facility can produce.
Maximum or design capacity defines the highest rate of output that a process or activity can theoretically achieve.
Effective or planned capacity identifies the output rate that managers expect to get for a given activity or process.
Capacity: Definitions
demonstrated capacity deals with actual rather than planned production, and measures the actual level of output for a process or activity over time.
Capacity utilization is the percentage of the facility’s capacity that is used by actual production.

Capacity: Short Term

Short-term capacity is managed through the day-to-day decisions organizations make in response to short-term fluctuations in demand, and usually entail human resource adjustments or investment in quality improvement or process optimization.
Capacity: Intermediate-term
Intermediate-term capacity decisions are made on a month-to-month basis in response to, for example, seasonal fluctuations in demand, and, as for short-term capacity decision, usually entail adjustments to human resources – hiring, firing, overtime, and subcontracting

Capacity: Long Term
Long-term capacity decisions, rely on projections about highly uncertain future demand and the strategic moves of competitors, and involve sizeable investments in resources – such as process and information technology and physical facilities – that take some time to put in place.

Capacity: The main Issues
What is the forecasted demand for the company’s products or services in the short-, intermediate- and long-term? How will demand grow or shrink over time?
How much capacity should the company have to cover expected demand? What service level does the company want to offer? How much reserve capacity does the company wish to have to buffer against unexpected fluctuations in demand?
Capacity: The main Issues
In what increments and when, or at what intervals, should the company add capacity?
What type of capacity should the company add? Human resources? Process and information technology? Facilities? Can the company extract more output from its existing resources, thereby increasing its capacity? How will it trade off short-, intermediate-, and long-term investments in capacity?
Where in the value chain, internal and/or external to the company, should capacity be added?

Capacity: Considerations
Explore multiple capacity expansion (or contraction) scenarios
Quantify the sensitivity of the answer to varying market assumptions
Use discounted cash flow techniques to quantify the costs of various capacity expansion timing policies
Evaluate competitive responses to capacity changes
Capacity: Considerations
Lead – Try Not to Run Short.
Lag – Maximize Capacity Utilization. In the lag strategy, average demand is never fulfilled and the company holds less than necessary capacity.
Stay Even -- Build to Forecast. In the stay-even policy, average demand is met, but the company has excess capacity half of the time, and too little capacity the other half of the time.
Lead and Lag policies

Evaluating Investments
Example 1:
$10 million investment:
Annual returns for next 5 years:
3 Million/ year
Discount rate: 15%

Should the firm make the investment?
Irreversible Investment & Demand Uncertainty
XYZ Inc is planning to build a plant to make Gizmos
Cost of investment: $1.6 million
Current selling price of a Gizmo: $60.
Annual Sales volumes: 10000 units
Profit: $200,000
Next year
50% chance price will increase to $70 and profits go up to $300,000 per year – remain at that level forever
50% chance price will drop to $50 and profits go down to $100,000, remain at that level forever
Should the firm Invest in the plant?
Assume salvage value is zero
NPV if it decides to invest?
Expenditure: 1.6 million
You will get in perpetuity
$200,000 this year
0.5*100,000 + 0.5*300,000 = $200,000 then on
Discounted earnings:


Net :
-1,600,000 + 2,200,000 = $600,000

Should they go ahead?
What if it decides to wait for one year?
When is it reasonable to assume you can wait?




If they wait

Assuming that next period the price is $300



If the price is $100 don’t invest
Cost Uncertainties
Research and development projects
Pharmaceuticals
Aerospace

Production Cost
Cost reduction in manufacturing
Raw material costs
Labor costs..

Construction Projects
Real estate development projects
Power plants

Capacity and the role of Flexibility

Clearly, it pays to wait.

What is the value of flexibility?
773-600=173

Research and Development Example
Initial investment in product development is $100,000

Investment in the second phase depends on the outcome of first phase:
If the first phase is successful you can launch and earn discounted revenue of $220,000
If the first phase runs into technical problems you expect to spend $300,000 more, and then earn discounted revenue of $220,000
Estimated probability of success 50%

Should you invest?
Implications
Sequential decision making can lead to different decisions
Example if there is uncertainty in demand building two smaller plants in sequence may be better than committing to a large plant.
Projects value may increase if you have the flexibility to make a decision after some uncertainty gets resolved.

Nature of competitive environments
A small firm in a market with many players
Price Takers:

Large firms in concentrated markets
Price setters

Large but asymmetric power in concentrated markets
Capacity / Capital Investments
Two firms in a concentrated market:

Current Situation:
Costs $10/ units
Profits: $10 Million/ year

New Technology is available:
Investment cost to adopt: $15 Million
Costs : $7.50/ unit
Profits depends on other actions

Suppose you know that the NPV of the project is -$5 Million if the other firm also adopts,
Should you invest?
NPV of project will depend on other’s actions
Evaluating Investment Decision
Need to define the base case or the alternative to consider has to be carefully defined

In the previous case you have to compare profits if you don’t adopt against profits of adopting, given that your rival will adopt & not based on current situation

Capacity ,Commitment & Leadership
The role of commitments
Two companies that choose capacity decision simultaneously.The pay offs are the following:
Capacity ,Commitment & Leadership
What is a Nash Equilibrium?
(x*,y*) is a Nash equilibrium if:




Nash Equilibrium
Can we predict what will the two companies do?
There is a single Nash Equilibrium
Firm 2’s preferred solution?
What would firm 2 want firm 1 to choose?
Capacity and the role of Commitment
Can firm 2 improve its position?
Suppose firm 2 takes a leadership position and announces its capacity ahead of time.
Firm 2 can announce credibly (by building capacity) that it will be aggressive.
In this case, the best firm 1 can do is to announce that it is expanding at lower level.
Credible Commitment
Hernan Cortez, conqueror of the Aztec empire in Mexico

Upon landing in Mexico destroyed all but one of his ships
Lost the flexibility to retreat and sail back!
Capacity and the role of Commitment
Notice that by taking this action firm 2 lost flexibility to learn the market conditions and to adopt.
Actually, by losing flexibility (making a commitment) the firm may become stronger.
For a commitment to be successful it must be:
Visible
Understandable
Credible (irreversible)
Examples
In 2000 Airbus announces high capacity A380 super-jumbo jet and commits to building it by devoting resources and booking orders from Singapore Airlines, FedEx, Quantas

Boeing announces that it is canceling its 747X project
Examples
Band-wagons of capacity expansions in:
Semi-conductor industry
Chemical industry
Capacity & Investment Decisions
Role of uncertainty in information
Flexibility of waiting for uncertainty to be resolved is valuable

Need to select base case correctly
Not status quo but:
What could happen (others may adopt etc).

In concentrated industries commitment can influence others strategies
Phillips vs. Sony
The Phillips case
Do you think Phillips should invest in 1983 or in 1984?
Suppose Phillips decides not to invest in 1983.
Suppose also that Phillips observes that popular buyers accept quickly. What should Philips do in 1984?
Second period analysis: Philips
In this case Phillips should invest. Notice that 2310 > 938 and 4579 > 1603
Second period analysis: Philips
The situation is simpler. We now need to consider only the following information:
Second period analysis: Sony
What should Sony do?
Assuming that Sony does not know the state of the market, and assumes 5% quick acceptance and 95% slow acceptance.
Sony compares: 0.05*1207 + 0.95*(-95) = -30 with 0.05*563 + 0.95*0 = 28
It is clear that Sony should not build in 1984.
First period analysis: Phillips
Knowing that Sony will not invest in 1984, what should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.05*4585 + 0.95*533 = 736
Assuming Phillips invests in the second year then the expected profit is:
0.05*4579 + 0.95*546 = 748
Clearly Phillips Should invest on the second year
A second scenario: Fast acceptance
Let us assume a different scenario in which Sony and Phillips believe that the probability of fast acceptance is 50%.
What will change in our analysis?
On the second period if Phillips observes quickly acceptance they will invest. (as before)
Sony compares: 0.5*1207 + 0.5*(-95) = 556 with 0.5*563 + 0.5*0 = 281.5
It is clear that Sony should build in 1984. (unlike before)

A second scenario: Fast acceptance
Knowing that Sony will invest in 1984, what should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.5*4585 + 0.5*533 = 2559
Assuming Phillips invests in the second year then the expected profit is:
0.5*2310 + 0.5*422 = 1366
Clearly, Phillips Should invest on the first year

A third scenario: Information sharing
The value of flexibility was reduced.
Notice, in the above two examples we assumed that Phillips is able to observe the real demand but Sony is unable to do the same.
What happens if Sony gets the same information as Phillips.
A third scenario: Information sharing
In analyzing this situation we assume that the probability of fast acceptance is 0.05
Assuming that Phillips does not build in 1983, it is clear that both Phillips and Sony will build in 1984.

A third scenario: Information sharing








A third scenario: Information sharing
What should Phillips do?
Assuming Phillips invests in the first year then the expected profit is:
0.05*4585 + 0.95*533 = 736
Assuming Phillips invests in the second year then the expected profit is:
0.05*2310 + 0.95*422 = 516
It is clear that Phillips should invest in the first period.

A fourth scenario: Some information sharing
Do we know whether Sony will get the information?
Suppose we assume that the chance that Sony will get the information is 10% what should Phillips do?
As before, If Phillips does not invest in 1983 and acceptance is fast, then Phillips invests in 1984.
Phillips does not know whether Sony got the information. So, Phillips expected profit is: 0.1*(2310*0.05+0.95*422)+0.9*(0.05*4578+0.95*546)=724
A fourth scenario: Some information sharing
If Phillips invests in 1983 the profit is 736
Clearly Phillips should invest in the first period.
End of the story
Phillips did not invest in the pressing facility in 1983. It assumed that the rate of acceptance will be low.
Sony built a small facility with the capacity of 3.5m (1984)
At the end of 1985 Sony had capacity of 12m
In 1987 it had the capacity of 26.4m
In 1987 Phillips open its first facility with capacity of 12m-15m
Lessons to learn
What were the uncertainties that Phillips faced?
Acceptance rates
Sony knowledge about the acceptance rates
What was the approach Phillips took?
A single probability for the speed of acceptance (5%)
A single probability for Sony’s knowledge of the acceptance rate (0%)
Future profits, single numbers
Lessons to learn
How accurate these number can be?
Past data and experience tell us that these numbers can’t be very accurate.
How robust is Phillips strategy of not investing?
Our calculations show that it is not robust at all.
Lessons to learn
What are the results?
A non-robust forecast. We have seen that small changes in the assumptions will result in a major shift in the policy.
What should Phillips have done?
Create three to four scenarios:
The present scenario
A scenario that assumes that Sony gets the information with a small probability, and acceptance rate is small
A scenario in which Sony gets the information and acceptance rate is high
Lessons to learn
What will a robust policy in this case be?
Invest in the first period.

1. Suppose that Sony believes that acceptance rate is 50%, and Philips
knows Sony’s belief with certainty. Philips still believes that the
acceptance rate is 5%, but Sony assumes that Philips believes that the
acceptance rate is 50%. When will each of the players invest (if at
all)? Explain your answer.

2. A Philips employee approached Sony’s vice president for research
and development and offered to provide Sony with Philips’s estimates
of the acceptance rate. Assuming that Sony is not likely to change
their own estimates of the acceptance rate, what is the value of this
information?