- •BUSINESSES IN THE BOOK
- •Preface
- •Brief Contents
- •CONTENTS
- •Why Study Strategy?
- •Why Economics?
- •The Need for Principles
- •So What’s the Problem?
- •Firms or Markets?
- •A Framework for Strategy
- •Boundaries of the Firm
- •Market and Competitive Analysis
- •Positioning and Dynamics
- •Internal Organization
- •The Book
- •Endnotes
- •Costs
- •Cost Functions
- •Total Cost Functions
- •Fixed and Variable Costs
- •Average and Marginal Cost Functions
- •The Importance of the Time Period: Long-Run versus Short-Run Cost Functions
- •Sunk versus Avoidable Costs
- •Economic Costs and Profitability
- •Economic versus Accounting Costs
- •Economic Profit versus Accounting Profit
- •Demand and Revenues
- •Demand Curve
- •The Price Elasticity of Demand
- •Brand-Level versus Industry-Level Elasticities
- •Total Revenue and Marginal Revenue Functions
- •Theory of the Firm: Pricing and Output Decisions
- •Perfect Competition
- •Game Theory
- •Games in Matrix Form and the Concept of Nash Equilibrium
- •Game Trees and Subgame Perfection
- •Chapter Summary
- •Questions
- •Endnotes
- •Doing Business in 1840
- •Transportation
- •Communications
- •Finance
- •Production Technology
- •Government
- •Doing Business in 1910
- •Business Conditions in 1910: A “Modern” Infrastructure
- •Production Technology
- •Transportation
- •Communications
- •Finance
- •Government
- •Doing Business Today
- •Modern Infrastructure
- •Transportation
- •Communications
- •Finance
- •Production Technology
- •Government
- •Infrastructure in Emerging Markets
- •Three Different Worlds: Consistent Principles, Changing Conditions, and Adaptive Strategies
- •Chapter Summary
- •Questions
- •Endnotes
- •Definitions
- •Definition of Economies of Scale
- •Definition of Economies of Scope
- •Economies of Scale Due to Spreading of Product-Specific Fixed Costs
- •Economies of Scale Due to Trade-offs among Alternative Technologies
- •“The Division of Labor Is Limited by the Extent of the Market”
- •Special Sources of Economies of Scale and Scope
- •Density
- •Purchasing
- •Advertising
- •Costs of Sending Messages per Potential Consumer
- •Advertising Reach and Umbrella Branding
- •Research and Development
- •Physical Properties of Production
- •Inventories
- •Complementarities and Strategic Fit
- •Sources of Diseconomies of Scale
- •Labor Costs and Firm Size
- •Spreading Specialized Resources Too Thin
- •Bureaucracy
- •Economies of Scale: A Summary
- •The Learning Curve
- •The Concept of the Learning Curve
- •Expanding Output to Obtain a Cost Advantage
- •Learning and Organization
- •The Learning Curve versus Economies of Scale
- •Diversification
- •Why Do Firms Diversify?
- •Efficiency-Based Reasons for Diversification
- •Scope Economies
- •Internal Capital Markets
- •Problematic Justifications for Diversification
- •Diversifying Shareholders’ Portfolios
- •Identifying Undervalued Firms
- •Reasons Not to Diversify
- •Managerial Reasons for Diversification
- •Benefits to Managers from Acquisitions
- •Problems of Corporate Governance
- •The Market for Corporate Control and Recent Changes in Corporate Governance
- •Performance of Diversified Firms
- •Chapter Summary
- •Questions
- •Endnotes
- •Make versus Buy
- •Upstream, Downstream
- •Defining Boundaries
- •Some Make-or-Buy Fallacies
- •Avoiding Peak Prices
- •Tying Up Channels: Vertical Foreclosure
- •Reasons to “Buy”
- •Exploiting Scale and Learning Economies
- •Bureaucracy Effects: Avoiding Agency and Influence Costs
- •Agency Costs
- •Influence Costs
- •Organizational Design
- •Reasons to “Make”
- •The Economic Foundations of Contracts
- •Complete versus Incomplete Contracting
- •Bounded Rationality
- •Difficulties Specifying or Measuring Performance
- •Asymmetric Information
- •The Role of Contract Law
- •Coordination of Production Flows through the Vertical Chain
- •Leakage of Private Information
- •Transactions Costs
- •Relationship-Specific Assets
- •Forms of Asset Specificity
- •The Fundamental Transformation
- •Rents and Quasi-Rents
- •The Holdup Problem
- •Holdup and Ex Post Cooperation
- •The Holdup Problem and Transactions Costs
- •Contract Negotiation and Renegotiation
- •Investments to Improve Ex Post Bargaining Positions
- •Distrust
- •Reduced Investment
- •Recap: From Relationship-Specific Assets to Transactions Costs
- •Chapter Summary
- •Questions
- •Endnotes
- •What Does It Mean to Be “Integrated?”
- •The Property Rights Theory of the Firm
- •Alternative Forms of Organizing Transactions
- •Governance
- •Delegation
- •Recapping PRT
- •Path Dependence
- •Making the Integration Decision
- •Technical Efficiency versus Agency Efficiency
- •The Technical Efficiency/Agency Efficiency Trade-off
- •Real-World Evidence
- •Double Marginalization: A Final Integration Consideration
- •Alternatives to Vertical Integration
- •Tapered Integration: Make and Buy
- •Franchising
- •Strategic Alliances and Joint Ventures
- •Implicit Contracts and Long-Term Relationships
- •Business Groups
- •Keiretsu
- •Chaebol
- •Business Groups in Emerging Markets
- •Chapter Summary
- •Questions
- •Endnotes
- •Competitor Identification and Market Definition
- •The Basics of Competitor Identification
- •Example 5.1 The SSNIP in Action: Defining Hospital Markets
- •Putting Competitor Identification into Practice
- •Empirical Approaches to Competitor Identification
- •Geographic Competitor Identification
- •Measuring Market Structure
- •Market Structure and Competition
- •Perfect Competition
- •Many Sellers
- •Homogeneous Products
- •Excess Capacity
- •Monopoly
- •Monopolistic Competition
- •Demand for Differentiated Goods
- •Entry into Monopolistically Competitive Markets
- •Oligopoly
- •Cournot Quantity Competition
- •The Revenue Destruction Effect
- •Cournot’s Model in Practice
- •Bertrand Price Competition
- •Why Are Cournot and Bertrand Different?
- •Evidence on Market Structure and Performance
- •Price and Concentration
- •Chapter Summary
- •Questions
- •Endnotes
- •6: Entry and Exit
- •Some Facts about Entry and Exit
- •Entry and Exit Decisions: Basic Concepts
- •Barriers to Entry
- •Bain’s Typology of Entry Conditions
- •Analyzing Entry Conditions: The Asymmetry Requirement
- •Structural Entry Barriers
- •Control of Essential Resources
- •Economies of Scale and Scope
- •Marketing Advantages of Incumbency
- •Barriers to Exit
- •Entry-Deterring Strategies
- •Limit Pricing
- •Is Strategic Limit Pricing Rational?
- •Predatory Pricing
- •The Chain-Store Paradox
- •Rescuing Limit Pricing and Predation: The Importance of Uncertainty and Reputation
- •Wars of Attrition
- •Predation and Capacity Expansion
- •Strategic Bundling
- •“Judo Economics”
- •Evidence on Entry-Deterring Behavior
- •Contestable Markets
- •An Entry Deterrence Checklist
- •Entering a New Market
- •Preemptive Entry and Rent Seeking Behavior
- •Chapter Summary
- •Questions
- •Endnotes
- •Microdynamics
- •Strategic Commitment
- •Strategic Substitutes and Strategic Complements
- •The Strategic Effect of Commitments
- •Tough and Soft Commitments
- •A Taxonomy of Commitment Strategies
- •The Informational Benefits of Flexibility
- •Real Options
- •Competitive Discipline
- •Dynamic Pricing Rivalry and Tit-for-Tat Pricing
- •Why Is Tit-for-Tat So Compelling?
- •Coordinating on the Right Price
- •Impediments to Coordination
- •The Misread Problem
- •Lumpiness of Orders
- •Information about the Sales Transaction
- •Volatility of Demand Conditions
- •Facilitating Practices
- •Price Leadership
- •Advance Announcement of Price Changes
- •Most Favored Customer Clauses
- •Uniform Delivered Prices
- •Where Does Market Structure Come From?
- •Sutton’s Endogenous Sunk Costs
- •Innovation and Market Evolution
- •Learning and Industry Dynamics
- •Chapter Summary
- •Questions
- •Endnotes
- •8: Industry Analysis
- •Performing a Five-Forces Analysis
- •Internal Rivalry
- •Entry
- •Substitutes and Complements
- •Supplier Power and Buyer Power
- •Strategies for Coping with the Five Forces
- •Coopetition and the Value Net
- •Applying the Five Forces: Some Industry Analyses
- •Chicago Hospital Markets Then and Now
- •Market Definition
- •Internal Rivalry
- •Entry
- •Substitutes and Complements
- •Supplier Power
- •Buyer Power
- •Commercial Airframe Manufacturing
- •Market Definition
- •Internal Rivalry
- •Barriers to Entry
- •Substitutes and Complements
- •Supplier Power
- •Buyer Power
- •Professional Sports
- •Market Definition
- •Internal Rivalry
- •Entry
- •Substitutes and Complements
- •Supplier Power
- •Buyer Power
- •Conclusion
- •Professional Search Firms
- •Market Definition
- •Internal Rivalry
- •Entry
- •Substitutes and Complements
- •Supplier Power
- •Buyer Power
- •Conclusion
- •Chapter Summary
- •Questions
- •Endnotes
- •Competitive Advantage Defined
- •Maximum Willingness-to-Pay and Consumer Surplus
- •From Maximum Willingness-to-Pay to Consumer Surplus
- •Value-Created
- •Value Creation and “Win–Win” Business Opportunities
- •Value Creation and Competitive Advantage
- •Analyzing Value Creation
- •Value Creation and the Value Chain
- •Value Creation, Resources, and Capabilities
- •Generic Strategies
- •The Strategic Logic of Cost Leadership
- •The Strategic Logic of Benefit Leadership
- •Extracting Profits from Cost and Benefit Advantage
- •Comparing Cost and Benefit Advantages
- •“Stuck in the Middle”
- •Diagnosing Cost and Benefit Drivers
- •Cost Drivers
- •Cost Drivers Related to Firm Size, Scope, and Cumulative Experience
- •Cost Drivers Independent of Firm Size, Scope, or Cumulative Experience
- •Cost Drivers Related to Organization of the Transactions
- •Benefit Drivers
- •Methods for Estimating and Characterizing Costs and Perceived Benefits
- •Estimating Costs
- •Estimating Benefits
- •Strategic Positioning: Broad Coverage versus Focus Strategies
- •Segmenting an Industry
- •Broad Coverage Strategies
- •Focus Strategies
- •Chapter Summary
- •Questions
- •Endnotes
- •The “Shopping Problem”
- •Unraveling
- •Alternatives to Disclosure
- •Nonprofit Firms
- •Report Cards
- •Multitasking: Teaching to the Test
- •What to Measure
- •Risk Adjustment
- •Presenting Report Card Results
- •Gaming Report Cards
- •The Certifier Market
- •Certification Bias
- •Matchmaking
- •When Sellers Search for Buyers
- •Chapter Summary
- •Questions
- •Endnotes
- •Market Structure and Threats to Sustainability
- •Threats to Sustainability in Competitive and Monopolistically Competitive Markets
- •Threats to Sustainability under All Market Structures
- •Evidence: The Persistence of Profitability
- •The Resource-Based Theory of the Firm
- •Imperfect Mobility and Cospecialization
- •Isolating Mechanisms
- •Impediments to Imitation
- •Legal Restrictions
- •Superior Access to Inputs or Customers
- •The Winner’s Curse
- •Market Size and Scale Economies
- •Intangible Barriers to Imitation
- •Causal Ambiguity
- •Dependence on Historical Circumstances
- •Social Complexity
- •Early-Mover Advantages
- •Learning Curve
- •Reputation and Buyer Uncertainty
- •Buyer Switching Costs
- •Network Effects
- •Networks and Standards
- •Competing “For the Market” versus “In the Market”
- •Knocking off a Dominant Standard
- •Early-Mover Disadvantages
- •Imperfect Imitability and Industry Equilibrium
- •Creating Advantage and Creative Destruction
- •Disruptive Technologies
- •The Productivity Effect
- •The Sunk Cost Effect
- •The Replacement Effect
- •The Efficiency Effect
- •Disruption versus the Resource-Based Theory of the Firm
- •Innovation and the Market for Ideas
- •The Environment
- •Factor Conditions
- •Demand Conditions
- •Related Supplier or Support Industries
- •Strategy, Structure, and Rivalry
- •Chapter Summary
- •Questions
- •Endnotes
- •The Principal–Agent Relationship
- •Combating Agency Problems
- •Performance-Based Incentives
- •Problems with Performance-Based Incentives
- •Preferences over Risky Outcomes
- •Risk Sharing
- •Risk and Incentives
- •Selecting Performance Measures: Managing Trade-offs between Costs
- •Do Pay-for-Performance Incentives Work?
- •Implicit Incentive Contracts
- •Subjective Performance Evaluation
- •Promotion Tournaments
- •Efficiency Wages and the Threat of Termination
- •Incentives in Teams
- •Chapter Summary
- •Questions
- •Endnotes
- •13: Strategy and Structure
- •An Introduction to Structure
- •Individuals, Teams, and Hierarchies
- •Complex Hierarchy
- •Departmentalization
- •Coordination and Control
- •Approaches to Coordination
- •Types of Organizational Structures
- •Functional Structure (U-form)
- •Multidivisional Structure (M-form)
- •Matrix Structure
- •Matrix or Division? A Model of Optimal Structure
- •Network Structure
- •Why Are There So Few Structural Types?
- •Structure—Environment Coherence
- •Technology and Task Interdependence
- •Efficient Information Processing
- •Structure Follows Strategy
- •Strategy, Structure, and the Multinational Firm
- •Chapter Summary
- •Questions
- •Endnotes
- •The Social Context of Firm Behavior
- •Internal Context
- •Power
- •The Sources of Power
- •Structural Views of Power
- •Do Successful Organizations Need Powerful Managers?
- •The Decision to Allocate Formal Power to Individuals
- •Culture
- •Culture Complements Formal Controls
- •Culture Facilitates Cooperation and Reduces Bargaining Costs
- •Culture, Inertia, and Performance
- •A Word of Caution about Culture
- •External Context, Institutions, and Strategies
- •Institutions and Regulation
- •Interfirm Resource Dependence Relationships
- •Industry Logics: Beliefs, Values, and Behavioral Norms
- •Chapter Summary
- •Questions
- •Endnotes
- •Glossary
- •Name Index
- •Subject Index
Selecting Performance Measures: Managing Trade-offs between Costs • 417
is another. Disagreements over such charges grew heated, and much of top management’s time was consumed in mediating these disputes. Some divisions also began engaging in “channel stuffing”—a practice in which orders from other parts of the company were rushed to be filled at the end of each month. This allowed the division filling the order to recognize the revenue from this sale (and thus increase profits), but it led to problems with excess inventory. Internal strife became so severe that the firm eventually elected to discontinue the division-based bonuses and instead make profitsharing payments based on the performance of the entire firm.
A manager designing pay plans should identify the activities an employee can undertake to improve measured performance and then ask how well this set of activities overlaps with those the firm would like the employee to pursue. Are there activities important to the firm that are not reflected in measured performance? Are there activities that improve measured performance that the firm does not want employees to pursue? The larger these sets of activities, the lower the efficacy of pay-for-performance based on the measure in question.
Organizations may respond to this problem in a number of ways. First, they may elect not to use any pay-for-performance incentives. If the performance measures are of poor quality, the firm may be better off paying fixed salaries and instructing employees in how to allocate their efforts toward various activities. While this approach does not motivate employees to exert extra effort on the job, it has the virtue of not motivating them to ignore tasks that are important but difficult to measure. Most U.S. public schools offer little or no compensation based on teacher performance. The problems of identifying good performance measures may be so severe that the best outcome may be to rely on teachers’ inherent concern for students’ educational progress.
Firms may also deal with limitations in measurement through careful job design. Grouping tasks according to ease of measurement can mitigate the multitasking problem. Suppose, for example, that activities A and B are easy to measure but activities C and D are hard to measure. If, on one hand, tasks A and C are assigned to one employee and B and D are assigned to another, the firm faces the multitask problems identified earlier. If, on the other hand, the firm assigns tasks A and B to one employee and C and D to the other, it can provide strong incentives for tasks A and B without pulling focus away from C and D.
Finally, firms can augment explicit incentive contracts with direct monitoring and subjective performance evaluation. Recalling the Lantech example, it may be difficult to base an explicit contract on whether a manager is fighting “too hard” to allocate overhead to another division. But it may be relatively easy for a CEO or other top manager to subjectively assess whether this is going on. If such assessments can be incorporated into the determination of overall compensation, subjective evaluations of performance can mitigate the problems we have described. Critics of U.S. public schools contend that the tenure system and rigid pay and promotion criteria limit the ability of school administrators to reward teachers based on subjective evaluations.
SELECTING PERFORMANCE MEASURES:
MANAGING TRADE-OFFS BETWEEN COSTS
The foregoing discussion identifies three features of a good performance measure:
•A performance measure that is less affected by random factors will allow the firm to tie pay closely to performance without introducing much variability into the employee’s pay.
418 • Chapter 12 • Performance Measurement and Incentives
TABLE 12.3
Jobs with Varying Ease of Performance Measurement
Job for Which Performance Is Relatively Easy to Measure
Job for Which Performance Is Relatively Difficult to Measure
Harvesting grapes |
Vintner |
Bicycle messenger |
Flight attendant |
Pharmaceutical sales representative |
Pharmaceutical research scientist |
Manager of advertising campaign |
Manager of customer service center |
|
|
•A measure that reflects all the activities the firm wants undertaken will allow the firm to use strong incentives without pulling the employee’s attention away from important tasks.
•A performance measure that cannot be improved by actions the firm does not want undertaken will allow the firm to offer strong incentives without also motivating counterproductive actions.
Unfortunately, performance measures meeting all three of these criteria are rare. In Table 12.3, we highlight a selection of jobs for which performance is relatively easy to measure, and compare this to another selection for which performance is relatively more difficult to measure. In Table 12.4, we list performance measures that might be used for various jobs, and identify some problems associated with each.
A firm’s search for the best performance measure involves trade-offs among the costs identified earlier. Consider the question of whether to use “absolute” or “relative” measures of an employee’s performance. A relative measure is constructed by comparing one employee’s performance to another’s. If the sources of randomness affecting the two employees’ individual performance exhibit a positive correlation, basing each employee’s pay on the difference between the individual performance measures will shield employees from risk.17 Hence, a firm using relative performance measures may be able to pay a smaller risk premium and therefore use stronger incentives. But relative measures can exacerbate multitask problems. Consider the possibility that one employee might be able to take actions that reduce the productivity of another employee. Clearly, the firm does not want to encourage this activity; however, relative performance evaluation directly rewards it. In determining whether to use relative rather than absolute performance measures, firms must weigh the possible reductions in risk against potential increases in the incentive to undertake counterproductive actions.
In a well-known application of this idea, the fictional real estate salesmen in the play Glengarry Glen Ross were compensated based on relative performance. Their boss announced that the first prize in a sales contest was a Cadillac El Dorado. Second prize was a set of steak knives. Third prize was “you’re fired.” This scheme was excessively harsh, but it did reward hard work and ability and shielded the salesmen from correlated risks (such as macroeconomic fluctuations). It also led one salesman to do something entirely counterproductive—he stole the list of promising leads.
Similar considerations enter into the choice between narrow or broad performance measures. An example of a narrow measure might be the number of pieces of output produced by an individual employee. A broad measure might be the accounting profits of the plant where the employee works. The broad measure has the advantage of rewarding the employee for helping coworkers or for making suggestions that
Selecting Performance Measures: Managing Trade-offs between Costs • 419
TABLE 12.4
Performance Measures of Varying Quality for Different Jobs
Job Description |
Performance Measure |
Discussion |
Baseball pitcher |
Number of games won |
Depends on how team’s batters |
|
|
perform when pitcher is pitching; |
|
|
this measure is therefore affected |
|
|
by random factors beyond the |
|
|
pitcher’s control. |
|
Opponents’ batting |
May motivate pitcher to pitch too |
|
average |
cautiously. Pitcher would rather |
|
|
issue a walk (which does not count |
|
|
against batting average) than |
|
|
possibly surrender a hit. |
|
Earned run average |
Less noisy than number of wins, and |
|
|
motivates pitcher to take any action |
|
|
that will prevent other team from |
|
|
scoring runs. |
Police officer |
Crime rate on beat |
Crime rates vary considerably by |
|
|
neighborhood; this measure |
|
|
therefore depends on factors |
|
|
beyond the police officer’s control. |
|
Number of arrests |
Officer can make an arrest only if a |
|
|
crime has been committed; this |
|
|
measure therefore limits incentive |
|
|
to prevent crimes from being |
|
|
committed. |
|
Change in crime rate |
Less noisy than the level of crime, |
|
|
and motivates officer to take |
|
|
actions that reduce crime even if |
|
|
no arrest results. |
Local TV news |
Profits of station |
Profits depend crucially on the |
producer |
|
quality of network programming |
|
|
shown on the station; this measure |
|
|
may therefore be noisy. |
|
Number of journalism |
May motivate producer to overspend |
|
awards won |
on high-profile stories. |
|
Share of viewing |
Motivates actions that retain the |
|
audience retained |
potential audience; less noisy than |
|
when news comes on |
profits. |
|
|
|
improve the plant’s overall efficiency. However, the broad measure is also likely to be subject to more random factors. The broad measure depends on the actions of many workers and many sources of randomness; hence, linking an individual employee’s pay to this measure exposes that employee to considerable risk. A firm may therefore find it very costly (in terms of employee risk premiums) to use high-powered incentives based on broad measures. In determining whether to use the broad measure or the narrow one, the firm must weigh the benefit associated with “help” activities and extra suggestions against the cost of weaker incentives for individual effort. The firm
420 • Chapter 12 • Performance Measurement and Incentives
EXAMPLE 12.4 HERDING, RPE AND THE 2007–2008
CREDIT CRISIS
One of the clearest examples of how random factors can affect measured performance comes from the very top of most organizations. Pay for top executives like chief executive officers, chief operating officers, and chief financial officers is frequently tied directly to the price of the firm’s stock through grants of equity or equity-based instruments such as stock options.
The theory of financial markets suggests that the price of a firm’s stock will move up or down for a variety of reasons. Share prices are clearly affected by any news bearing directly on the firm’s future cash flow, but they are also affected by overall movements in the market. For example, during the late 1990s a major bull market pulled all U.S. share prices up by 25 percent or more annually. Even mediocre firms saw great gains in their share prices over this period. Similarly, the declining stock market during 2001 to 2002 saw nearly all firms’ share prices fall—even those of firms with good operating performance over the period. As a result, some analysts believe that a better performance measure might be the firm’s performance relative to competitors or market indices.
Jeff Zwiebel argues that while relative performance evaluation has some benefits, substantial costs are present as well.18 Zwiebel notes that relative performance evaluation could encourage “herding.” Herding is a phenomenon whereby individuals ignore their own information about the best course of action and instead simply do what everyone else is doing.
Zwiebel’s argument is this: Suppose a manager is likely to be fired when her firm’s performance is poor relative to industry rivals but will keep her job otherwise. Suppose also that the manager faces the following strategic choice: she can “follow the herd” by making strategic choices that are similar to those made by competitors, or she can adopt a new, promising, but untested strategy. Following the
herd means the manager’s performance is unlikely to be much different from that of rivals, and so she is unlikely to be fired. The contrarian strategy has a higher expected payoff than the herd strategy, but its newness means that there is at least some chance it will fail. If the contrarian strategy fails, the firm’s performance will lag the industry, and the manager will be fired. Under these conditions, the manager may well stick with the herd, even if she knows the potential returns to the contrarian strategy are high. As we noted at the start of this chapter, Merrill Lynch CEO O’Neal seems to have been comparing his firm’s trading performance to that of rivals. One reason for his insistence on matching Goldman may have been that his continued job security depended on achieving earnings similar to Goldman’s. Could this form of relative performance evaluation have led to herding on Wall Street? It is difficult to say for sure, but it is clearly the case that many, many financial institutions were actively involved in financing risky subprime mortgages. As housing prices rose through the late 1990s and early 2000s, mortgage default rates stayed low and these risky investments paid off handsomely. Any firm choosing not to play this risky game would show poor relative performance. Managers of such firms might begin to feel the heat from shareholders, as Zwiebel suggests.
Any manager with the contrarian strategy— taking, say, a short position in the subprime mortgage securities, betting that default rates will rise—would have incurred losses through the 2000 to 2006 period. But this strategy would have earned huge profits as house prices fell and the subprime mortgage market imploded in 2007. Would a contrarian manager have kept her job long enough to earn those profits? Or would the poor relative performance between 2000 and 2006 have led to that manager’s firing?