Managerial Economics Topic 1

1. Managerial EconomicsTopic 1 Shift from optimization to equilibrium and consider properties of that equilibrium (I.e. efficiency)Shift from optimization to equilibrium and consider properties of that equilibrium (I.e. efficiency)

2. Introduction: Why economics? Economics provides basic tools to analyse ... production and exchange (both in markets and other organisations) prices competition investments changes in markets over time strategic interaction between firms Provides a �toolkit� that can be applied to a huge range of firm decisions.

3. Course Objectives Specific objectives are to understand: Pricing and value creation: what determines how much you pay or are paid? Changing price: what actions can you take that alter pricing outcomes? General objective: Be able to apply the tools of economics to help your managerial decision making

4. Skill background for subject Logical and intuitive thinking Interpretation of graphs Mathematics Mathematics practice = the problem sets Real-world applications: in-class discussion, and examples submitted by email. Economists and the real world.

5. General thoughts on �word problems� in economics. Word problems and Pattern recognition First apply the logic to the extreme, then use real-world intuition to identify where logical results are missing some real-world elements: in class discussion, in exercises, in texts Learning to tackle brand-new questions: the role of individual work in attempting problem sets The role of maths in pattern recognition and general skills; math fear Game Theory

6. Subject Organisation Office hours: Jenny George, Room 250 j.george@mbs.edu Office hours by appointment � send me an email Textbook Core Economics for Managers by Joshua Gans Co-opetition by Adam Brandenburger and Barry Nalebuff Extra: Good reference textbook, especially for the second half of the class: Michael Baye�s Managerial Economics and Business Strategy (McGraw-Hill, 5th edition)

7. Subject organisation, cont�d NOTE THAT � mistake in notes distributed . . .NOTE THAT � mistake in notes distributed . . .

8. Topic 1Non-strategic decision making: decision trees and economic costs

9. Overview of Topic 1: Decision Trees � the most basic analytical tool of decision making Non-strategic decisions (this week) Strategic decisions (later) Decision rules flowing from decision trees. The concept of economic profit ? Next week = Decisions to buy and sell � we start with the most basic decision facing a firm in a market

10. Types of Decisions

11. Decisions Non-strategic � if the decision only depends on your actions and not how your actions interact with other peoples� actions Strategic � your best action depends on other peoples� actions and vice-versa Uncertainty � if there is uncertainty or risk associated with your actions Examples: whether an R&D project is successful (or what products are developed); whether tastes for products change,�

12. Basic Tool - Decision Trees The basic tool for decision making is a decision tree Idea: a traveller comes to a fork in the road. She must make a decision whether to go right or left.

13. Example (non-strategic under certainty) Moonee Ponds $120,000 open a restaurant North Carlton $150,000 Niki don�t 0

14. Decision trees are only useful if there is� complexity then the process of laying out the problem sequentially is helpful uncertainty then the tree has to be solved mathematically other firms making decisions then it�s strategic choice: again has to be solved mathematically

15. Decision-Making Product development

16. Chang�s dilemma in 2006 Sarah Chang is the owner of a small electronics company. There is a proposal for the provision of an electronic timing system for the 2008 Olympic Games. For several years, Chang�s company has been developing a new microprocessor, a critical component in a timing system that would be superior to any product currently on the market.

17. Chang�s dilemma in 2006 Progress has been slow and Chang is unsure about whether the new product will be developed on time. If the R&D succeeds, then there is an excellent chance her company will win the $1m Olympic contract; awarded solely on the basis of quality. If it does not succeed, they might still win the contract with their original, but inferior, system for which there are closer substitutes. The costs involved in continuing R&D are $200,000. Developing a proposal itself will cost Chang�s company $50,000. Finally, the costs of producing the product � should they win the contract � will be $150,000. Should Chang continue R&D or not?

18. Framing the Decision I Chang�s main decision is between two alternatives � to continue R&D or to abandon the project

19. Four steps for using decision trees Step 1: Structure the tree Step 2: Make assessments (outcomes and probabilities) Step 3: Make a decision (using a decision rule) Step 4: Sensitivity analysis

20. Building a tree There are two types of �fork� (nodes) in a decision tree. Decision node where the decision maker is able to choose whichever path he or she wants State of nature node where �Nature� (or chance) determines which path is taken. We use squares for decision nodes and circles for state of nature nodes. When we �solve� the tree we will treat each type of node quite differently

21. Step 1: Structure

22. Step 1: Structure

23. Step 1: Structure

25. Step 2: Make assessments What numbers are we given in the question that we can incorporate into the tree?

27. Step 2: Make assessments What additional information does Chang need to make this decision?

28. Uncertainty in a decision tree We�ll assume that Chang knows the probability of success of any uncertain event: that�s obviously an exaggeration! Let�s say that the probability of winning the contract with the old product is only 5% = 0.05

30. More uncertainty What about the chance of the R&D succeeding? Let�s assume Sarah knows that the probability of this is 60%.

32. And more� But wait� we still need to estimate the chance of winning the contract with the improved microprocessor. Let�s assume that it�s 90% And of course we can assume the chance with the old one is still 5%

34. Completed tree A completed decision tree: Has structure that takes into account every possibility Clearly distinguishes between times when a decision or choice is able to made and times when the outcome is due to chance/Nature Has outcomes at end of each branch that include all relevant costs and revenues Has probabilities on each branch that comes from a state of nature node (probabilities for all branches from a single node must add to 1)

35. Step 3: Making a decision While we built the tree by adding branches � the way to solve a tree is to start at the end and �roll back.� Looking forward and working backwards is a key skill in economic decision-making

36. Example (non-strategic under certainty) Moonee Ponds $120,000 open a restaurant $150,000 Niki North Carlton don�t 0 first solve the decision that is furthest to the right; eliminate the choice(s) you don�t want. Now move left.

37. State of Nature nodes: Expected value In deciding whether to invest in R&D, Chang wants to know, is it a risk worth taking? That�s the same as asking, is it a bet worth making? That�s a question we can easily answer, so long as Chang is risk-neutral = not worried about taking on large amounts of risk Her company is publicly traded, and her investors are highly diversified ? They don�t want Chang to protect them from risk, they want to protect themselves through diversification Then she should take any bet that pays off, on average Example: Flip a coin, Heads you get $2.10, Tails you lose $1. What would she get on average from this bet? 1000 flips: roughly 500 heads, 500 tails ? an average of ?? per flip. The quick way to calculate this is an expected value: = (Probability of heads)?(Payoff if heads) + (Prob of tails)?(Payoff if tails)

38. Step 3: At circles � calculate EMV

39. Step 3: At squares � make a choice

40. Step 3: Continuing�

41. Step 3: Continuing�

42. Solving the Tree

43. Solving the Tree

44. Solving the Tree

45. Uses of Decision Trees Decision Trees are used in situations that may be too complex to think through in your mind In Decision Analysis: used in situations where there is uncertainty, multiple decisions In �Managerial Economics�: used when the payoffs are not so obvious when the alternative choices are not so obvious to shape our decision-making, even when it�s too complex to write a tree Being systematic helps you to see though complexity and to remember all your alternative choices

46. Avoid common mistakes in decision making Ignore irrelevant information Ex: sunk costs Find the correct time frame Ex: fixed costs Find the correct units of measure Ex: marginal and lumpy decisions Consider all the alternatives Ex: exclusive and non-exclusive choices

47. (a) Ignore irrelevant information A common mistake in decision making is to include too much information � the decision becomes confused by extraneous detail �Sunk costs� are one particular form of irrelevant information. Sunk costs are an amount that you have already paid and nothing you do will change the amount that you paid. As there is nothing you can do to avoid the cost, the sunk cost should be ignored.

48. An example of sunk costs Mita runs petrol stations and express stores at several highway exits. Until recently, she didn�t sell any drinks. She brought in a new line of drinks, Fizzies, which have proved unpopular. She has 10,000 Fizzies left. She thinks she can sell half of the remaining drinks for $1.00, but only 15% of the drinks at the standard price of $2.50. If she paid $0.30 per drink, how much should she charge? What about if she paid $1.05 per drink? Mita cannot return unsold stock of Fizzies, but must throw the stock out.

49. If she cannot return unsold Fizzies, then the purchase price is irrelevant � it is a sunk cost

50. Definition: Economic Profit The economic profit of decision A relative to decision B is the accounting profits you earn from decision A, minus the accounting profits you earn from decision B. (It�s the payoff from branch A, minus the payoff on branch B.) sell at $1.00 $5000 - cost Mita sell at $2.50 $3750 - cost Whatever the sunk costs, the economic profit of charging $1 rather than $2.50 is $1250.

51. One implication: choose a baseline to start from

52. Questions How might your answer to �Mita�s problem� change if she was a wholesaler for a few big customers (i.e. if it were a strategic game?) In the 1980s, the manager of R&D in a division of IBM priced the innovations as a large markup over the cost of R&D. As a result, some of the innovations were never sold. When asked by an employee why he didn�t discount those innovations, he said �IBM doesn�t do R&D for free.� Go back to Sarah Chang�s problem: where could we have saved time and effort, if we had recognized the presence of sunk costs?

53. (b) The right time frame for a decision Variable Costs: Costs that vary depending on how much you produce. Fixed Costs: Costs that you incur, no matter how much you produce (even if you produce nothing) Ex: $800 a month in rent, if you cannot break your lease or sublet. Notice that no expense is a fixed cost forever: eventually your lease contract will finish. It�s important to specify the time frame over which a cost is fixed.

54. Example January 1, 2006, the CEO of Prescott & Kydland declares the firm�s venture in China to be a failure, and begins to plan withdrawal from the market. The subsidiary in China is earning $10 revenue per unit, and sells 10,000 units per month; but the inputs cost $5 per unit. Rent is $25,000 per month and staff costs are $28,000 per month. There are no other costs. Staff can be instantaneously dismissed, and the lease expires in 4 months, after which it can be renewed on a monthly basis. There are no subletting options. What should P&K do? Close right away, or close on May 1?

55. Decision tree

56. More complicated example January 1, 2006, P&K are deciding about when to leave. The subsidiary in China is earning $10 revenue per unit, and sells 10,000 units per month; but the inputs cost $5 per unit. Rent is $10,000 per month and staff costs are $52,000 per month. Staff can be dismissed with two months� notice, and the lease expires in 4 months, after which it can be renewed on a monthly basis. There are no subletting options.

57. Introduction to marginal thinking We know that P&K will want to close at some point: by May, they will want to close for sure. Their decision tree is: Rollback: April: do I want to stay open in April, or close? March: do I want to stay open in March, or close? Feb: � Jan: � They can ask the marginal question about each month (Jan, then Feb,�) to get the right answer: am I better off staying open this month, or closing?

58. How to think about fixed costs P&K ask themselves the question in the following way: �Should we stay open this month, or should we shut down?� make your decision based on Variable Cost only: stay open if you are covering your Variable Costs, otherwise close Close on March 1st (give employees notice right now, but keep operating until March 1st)

59. (c) Find the correct units of measure:Marginal and lumpy decisions �Marginal� decisions= decisions involving the smallest possible units. (e.g. produce one more pizza, stay open one more hour) �Lumpy� decisions = decisions that have to be made about a group of small units, together (e.g. shut down business)

60. Examples Question = is there a margin? = can you do a little bit more/less of what you�re doing? If so, then frame the question in terms of the margin: �Is it worth adding one more car to the train?� useful when every marginal decision is the same useful when every marginal decision brings a little less profit ALWAYS check whether there is also a lumpy decision (or two!) to make: should we install extra capacity, or de-commission some production equipment? should we shut down entirely?

61. Cost concepts Cost function is total cost of inputs the firm needs to produce output q. Denoted C(q). Fixed cost (FC): the cost that does not depend on the output level, C(0). Variable cost (VC): that cost which would be zero if the output level were zero, C(q) � C(0). Average cost (AC) (aka �unit cost�): total cost divided by output level, C(q)/q. Marginal cost (MC): the unit cost of a small increase in output. = derivative of cost with respect to output, dC/dq Approximated by C(q)-C(q-1) MB: Point out that all these are a function of q Go over numerical example FC=32, VC=2q^2, then TC=?, AC=?, MC=?MB: Point out that all these are a function of q Go over numerical example FC=32, VC=2q^2, then TC=?, AC=?, MC=?

62. Examples Bagels: modest fixed cost (space), relatively constant marginal cost (labor and materials). Electricity generation: large fixed cost (plant), initially declining marginal cost (large plants are more efficient, and many plants have �startup� costs). Music CDs: large fixed cost (recording), small marginal cost (production and distribution). More examples like CDs: computer software, cable TV, airlines (?) More examples like CDs: computer software, cable TV, airlines (?)

63. T-shirt factory example To produce T-shirts: Lease one machine at $200 / week, on a long-term lease; no other uses for the machine. Machine requires one worker. The machine, operated by the worker, produces 10 T-shirts per hour. Worker is paid $10/hour on weekdays (up to 40 hours), $20/hour on Saturdays (up to 8 hours), $30 on Sundays (up to 8 hours). Assume no material costs

64. T-shirts: costs Suppose output level is 400 T-shirts per week. Then, Fixed cost: FC = $200. Variable cost: VC = 40 � $10 = $400. Average cost: AC = (200+400)/400 = $1.5 Marginal cost: MC = $2 after this point. (Note that producing an extra T-shirt would imply working on Saturday, which costs more.) Similar calculations can be made for other output levels, leading to the cost function �

67. T-shirts : how many to make Scenario A: Benetton�, sole buyer of T-shirts, offers price p = $1.8 per T-shirt. Benetton� is willing to buy as many T-shirts as factory wants to sell (at given price). Should factory increase output beyond 400 T-shirts per week, thus operating on Saturdays ? p = 1.8, AC = 1.5, MC = 2 Although factory is making money at q=400 (because p > AC), profits would be lower if it produced more (because p < MC); it would lose money at the margin. [Verify: compute profit at q=400,401]

68. No matter how much factory produces, price is below per-unit cost; i.e., no matter how much factory produces, it will lose money:� p < AC implies q � p < q � AC implies Revenue < Cost Optimal decision is to leave the market: but when? At the end of the lease: leave the market Before end of the lease: produce 400 T-shirts Important lesson: Scenario B: Benetton� offers p = $1.3 per T-shirt.

69. (d) Exclusive versus non-exclusive choices Suppose you have 2 investment opportunities Key question: Can I take both investment opportunities? Do I want to take both investment opportunities, or just one? Or neither? Often the hidden cost of a choice is giving up on another choice Example: What is the cost of doing an MBA? Besides the price, there is an opportunity cost = what you would have earned, using the resource (your time) for another opportunity.

70. Consider all the alternatives If you are using multiple resources then you can ask three types of questions: Could the firm stop using those resources, and put each one to its best alternative use? Could the firm pursue another business opportunity with precisely those resources? Could the firm combine its resources with more outside resources to do something else?

71. Consider all the alternatives One of the most common problems with decision making is not considering all the alternatives: Can you make a small change (marginal) or only a big change (lumpy)? Have you included all relevant time frames (e.g. when fixed costs become variable)? Have you included all of the relevant alternatives and options? Being systematic can often assist you in creative thinking: For example, you notice that there are several decision nodes ? are there options you forgot at any of the nodes?

72. Takeaways The costs that matter in deciding between two possible decisions are those that differ between the two decisions. Other costs don�t matter. Opportunity costs involve no cash outlay, but matter to decisions. Sunk costs do involve cash payments, but do not matter to decisions (they can�t change). Use marginal cost when deciding how much to produce, average cost when deciding whether to produce.

73. Practice: Shoreham (electricity) Initial cost estimate: $101m. Fierce political fights during the 1980s. Ultimately cost $5b to build and another $200m to decommission. Comment on the following: �Given the $5b cost, this plant will never make money and should be shut down to save Long Island consumers from having the highest cost electricity in the country.� Cost is sunk. Of course, people will differ over how to value the risk. In the end, they had the highest electric rates in the country anyway. Cost is sunk. Of course, people will differ over how to value the risk. In the end, they had the highest electric rates in the country anyway.

74. The pen factory Factory produces pens of all colors Can produce 8000 pens in an 8-hour day Cost of �pen machine� is $1000/day ($125/hour) Materials/labor cost is 15�/pen Market for pens: 30� for first 5000 red pens 25� for blue pens Firm decides to produce 5000 red, 3000 blue, netting profit of 50 = 5000 x (.30-.15) + 3000 x (.25-.15) - 1000 Question: Are blue pens profitable?

75. Practice: health club Suppose you own a health club. Current revenues: $3800 (per month). Expenses: $2,500 labor, $2000 lease payment. Lease unbreakable. Liquidation not an option. Chance to sublet for $1200. Should you keep it open? What if liquidation is an option? The $2,000 lease payment is a sunk cost, and the $1,200 is the opportunity costs for keeping the health club open. � With this cost structure, one is losing $700 a month on an accounting basis.� On an economic basis however, the economic costs would be $2,500 + $2,000 [actual expenses] + $1,200 [opportunity cost] - $2,000 [sunk costs] = $3,700.� The economic profits are thus $100 a month.� In this case it makes sense to keep the health club open.� If you could declare bankruptcy, though, the costs would be $2,500 + $2,000 + $1,200 = $5,700 and it would make sense to shut down. The $2,000 lease payment is a sunk cost, and the $1,200 is the opportunity costs for keeping the health club open. � With this cost structure, one is losing $700 a month on an accounting basis.� On an economic basis however, the economic costs would be $2,500 + $2,000 [actual expenses] + $1,200 [opportunity cost] - $2,000 [sunk costs] = $3,700.� The economic profits are thus $100 a month.� In this case it makes sense to keep the health club open.� If you could declare bankruptcy, though, the costs would be $2,500 + $2,000 + $1,200 = $5,700 and it would make sense to shut down.

76. Practice: Redsyke quarry Expansion of M6 motorway (UK) requires use (as quarry) of plot currently used by farmer. Quarry offers farmer 3x market value of land. Farmer�s lawyer: compensation should be based on value of cows and sheep that must be removed from plot. Case went to court. How would you testify? There is an LBS case on this. Issue is: could farmer take cows and sheep to a neighboring plot of land? If so, then 3x market value of land is very generous compensation. Otherwise, farmer is right as profit on cows and sheep is true opportunity cost. Note: farmer lost. Additional interesting problems: Baye, chapter 5, Numbers 14, 15.There is an LBS case on this. Issue is: could farmer take cows and sheep to a neighboring plot of land? If so, then 3x market value of land is very generous compensation. Otherwise, farmer is right as profit on cows and sheep is true opportunity cost. Note: farmer lost. Additional interesting problems: Baye, chapter 5, Numbers 14, 15.